NONLINEAR PARTIAL DIFFERENTIAL EQUATIONS

© 2004 by Chapman & Hall/CRC

HANDBOOK OF

**NONLINEAR PARTIAL DIFFERENTIAL EQUATIONS
**

Andrei D. Polyanin Valentin F. Zaitsev

**CHAPMAN & HALL/CRC
**

A CRC Press Company Boca Raton London New York Washington, D.C.

© 2004 by Chapman & Hall/CRC

**Library of Congress Cataloging-in-Publication Data
**

Polyanin, A.D. (Andrei Dmitrievich) Handbook of nonlinear partial differential equations / by Andrei D. Polyanin, Valentin F. Zaitsev. p. cm. Includes bibliographical references and index. ISBN 1-58488-355-3 (alk. paper) 1. Differential equations, Nonlinear — Numerical solutions. 2. Nonlinear mechanics — Mathematics. I. Zaitsev, V.F. (Valentin F.) II. Title. QA372.P726 2003 515¢.355 — dc22

2003058473

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, micro lming, and recording, or by an y information storage or retrieval system, without prior permission in writing from the publisher. The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Speci c permission must be obtained in writing f rom CRC Press LLC for such copying. Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identi cation and e xplanation, without intent to infringe.

**Visit the CRC Press Web site at www.crcpress.com
**

© 2004 by Chapman & Hall/CRC No claim to original U.S. Government works International Standard Book Number 1-58488-355-3 Library of Congress Card Number 2003058473 Printed in the United States of America 1 2 3 4 5 6 7 8 9 0 Printed on acid-free paper

© 2004 by Chapman & Hall/CRC

CONTENTS

Authors Foreword Some Notations and Remarks 1. Parabolic Equations with One Space Variable 1.1. Equations with Power-Law Nonlinearities 2 1.1.1. Equations of the Form ∂w = a ∂ w + bw + cw2 ∂t ∂x2 1.1.2. Equations of the Form 1.1.3. 1.1.4. 1.1.5. 1.1.6. 1.1.7. 1.1.8.

1.1.9. 1.1.10. 1.1.11. 1.1.12. 1.1.13. 1.1.14. 1.1.15. Other Equations

∂2w ∂w 2 3 ∂t = a ∂x2 + b0 + b1 w + b2 w + b3 w 2 Equations of the Form ∂w = a ∂ w + f (w) ∂t ∂x2 ∂2w ∂w Equations of the Form ∂t = a ∂x2 + f (x, t, w) 2 Equations of the Form ∂w = a ∂ w + f (w) ∂w + g(w) ∂t ∂x ∂x2 2 Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂w + g(x, t, w) 2 ∂t ∂x ∂x 2 2 Equations of the Form ∂w = a ∂ w + b ∂w + f (x, t, w) 2 ∂t ∂x ∂x 2 Equations of the Form ∂w = a ∂ w + f x, t, w, ∂w 2 ∂t ∂x ∂x 2 Equations of the Form ∂w = awk ∂ w + f x, t, w, ∂w ∂t ∂x ∂x2 ∂ Equations of the Form ∂w = a ∂x wm ∂w ∂t ∂x ∂ Equations of the Form ∂w = a ∂x wm ∂w + bwk ∂t ∂x ∂ Equations of the Form ∂w = a ∂x wm ∂w + bw + c1 wk1 + c2 wk2 ∂t ∂x ∂ Equations of the Form ∂w = ∂x f (w) ∂w + g(w) ∂t ∂x ∂ Equations of the Form ∂w = ∂x f (w) ∂w + g x, t, w, ∂w ∂t ∂x ∂x

+ c 3 w k3

1.2. Equations with Exponential Nonlinearities 2 1.2.1. Equations of the Form ∂w = a ∂ w + b0 + b1 eλw + b2 e2λw ∂t ∂x2 ∂ 1.2.2. Equations of the Form ∂w = a ∂x eλw ∂w + f (w) ∂t ∂x ∂ ∂w 1.2.3. Equations of the Form ∂t = ∂x f (w) ∂w + g(w) ∂x 1.2.4. Other Equations Explicitly Independent of x and t 1.2.5. Equations Explicitly Dependent on x and/or t 1.3. Equations with Hyperbolic Nonlinearities 1.3.1. Equations Involving Hyperbolic Cosine 1.3.2. Equations Involving Hyperbolic Sine 1.3.3. Equations Involving Hyperbolic Tangent 1.3.4. Equations Involving Hyperbolic Cotangent 1.4. Equations with Logarithmic Nonlinearities 2 1.4.1. Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂t ∂x2 1.4.2. Other Equations 1.5. Equations with Trigonometric Nonlinearities 1.5.1. Equations Involving Cosine 1.5.2. Equations Involving Sine 1.5.3. Equations Involving Tangent 1.5.4. Equations Involving Cotangent 1.5.5. Equations Involving Inverse Trigonometric Functions

© 2004 by Chapman & Hall/CRC

1.6. Equations Involving Arbitrary Functions 2 1.6.1. Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂t ∂x2 1.6.2. Equations of the Form 1.6.3. 1.6.4. 1.6.5. 1.6.6. 1.6.7. 1.6.8. 1.6.9. 1.6.10. 1.6.11. 1.6.12. 1.6.13. 1.6.14. 1.6.15.

∂w ∂2w ∂w ∂t = a ∂x2 + f (x, t) ∂x + g(x, t, w) 2 Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂t ∂x2 ∂x 2 2 Equations of the Form ∂w = a ∂ w + b ∂w + f (x, t, w) ∂t ∂x ∂x2 2 2 Equations of the Form ∂w = a ∂ w + b ∂w + f (x, t, w) ∂w + g(x, t, w) ∂t ∂x ∂x ∂x2 2 2 Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂w + h(x, t, w) ∂t ∂x ∂x ∂x2 2 Equations of the Form ∂w = a ∂ w + f x, t, w, ∂w ∂t ∂x2 ∂x 2 Equations of the Form ∂w = f (x, t) ∂ w + g x, t, w, ∂w ∂t ∂x ∂x2 2 Equations of the Form ∂w = aw ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂t ∂x2 ∂x 2 2 Equations of the Form ∂w = (aw + b) ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂w + ∂t ∂x2 ∂x ∂x

h(x, t, w) Equations of the Form Equations of the Form Equations of the Form Equations of the Form Equations of the Form

1.6.16. Equations of the Form 1.6.17. Equations of the Form

= f x, w, ∂w ∂ w + g x, t, w, ∂w 1.6.18. Equations of the Form ∂x ∂x2 ∂x 1.6.19. Nonlinear Equations of the Thermal (Diffusion) Boundary Layer 1.7. Nonlinear Schr¨ dinger Equations and Related Equations o 2 1.7.1. Equations of the Form i ∂w + ∂ w + f (|w|)w = 0 Involving Arbitrary Parameters ∂t ∂x2 ∂ 1.7.2. Equations of the Form i ∂w + x1n ∂x xn ∂w + f (|w|)w = 0 Involving Arbitrary ∂t ∂x Parameters 1.7.3. Other Equations Involving Arbitrary Parameters 1.7.4. Equations with Cubic Nonlinearities Involving Arbitrary Functions 1.7.5. Equations of General Form Involving Arbitrary Functions of a Single Argument 1.7.6. Equations of General Form Involving Arbitrary Functions of Two Arguments 2. Parabolic Equations with Two or More Space Variables 2.1. Equations with Two Space Variables Involving Power-Law Nonlinearities ∂ ∂ 2.1.1. Equations of the Form ∂w = ∂x f (x) ∂w + ∂y g(y) ∂w + awp ∂t ∂x ∂y 2.1.2. Equations of the Form 2.1.3. Equations of the Form 2.1.4. Other Equations

∂w ∂t ∂w ∂t ∂ ∂ = a ∂x wn ∂w + b ∂y wk ∂w ∂x ∂y

∂w ∂t ∂w ∂t ∂w ∂t ∂w ∂t ∂w ∂t ∂w ∂t ∂w ∂t ∂w ∂t

**= awm ∂ w + f (x, t) ∂w + g(x, t, w) ∂x ∂x2 ∂ = a ∂x w ∂w + f (x, t) ∂w + g(x, t, w) ∂x ∂x ∂ = a ∂x wm ∂w + f (x, t) ∂w + g(x, t, w) ∂x ∂x ∂ = a ∂x eλw ∂w +f (x, t, w) ∂x ∂ = ∂x f (w) ∂w +g x, t, w, ∂w ∂x ∂x = f (x, w) ∂ w ∂x2
**

2

2

= f (x, t, w) ∂ w + g x, t, w, ∂x2

2

2

∂w ∂x

=

∂ ∂x

f (w) ∂w + ∂x

∂ ∂y

g(w) ∂w + h(w) ∂y

2.2. Equations with Two Space Variables Involving Exponential Nonlinearities ∂ ∂ 2.2.1. Equations of the Form ∂w = ∂x f (x) ∂w + ∂y g(y) ∂w + aeλw ∂t ∂x ∂y 2.2.2. Equations of the Form

∂w ∂t ∂ ∂ = a ∂x eβw ∂w + b ∂y eλw ∂w + f (w) ∂x ∂y

2.3. Other Equations with Two Space Variables Involving Arbitrary Parameters 2.3.1. Equations with Logarithmic Nonlinearities 2.3.2. Equations with Trigonometrical Nonlinearities

© 2004 by Chapman & Hall/CRC

2.4. Equations Involving Arbitrary Functions 2.4.1. Heat and Mass Transfer Equations in Quiescent or Moving Media with Chemical Reactions ∂ ∂ 2.4.2. Equations of the Form ∂w = ∂x f (x) ∂w + ∂y g(y) ∂w + h(w) ∂t ∂x ∂y ∂ ∂ 2.4.3. Equations of the Form ∂w = ∂x f (w) ∂w + ∂y g(w) ∂w + h(t, w) ∂t ∂x ∂y 2.4.4. Other Equations Linear in the Highest Derivatives 2.4.5. Nonlinear Diffusion Boundary Layer Equations 2.5. Equations with Three or More Space Variables 2.5.1. Equations of Mass Transfer in Quiescent or Moving Media with Chemical Reactions 2.5.2. Heat Equations with Power-Law or Exponential Temperature-Dependent Thermal Diffusivity 2.5.3. Equations of Heat and Mass Transfer in Anisotropic Media 2.5.4. Other Equations with Three Space Variables 2.5.5. Equations with n Space Variables 2.6. Nonlinear Schr¨ dinger Equations o 2.6.1. Two-Dimensional Equations 2.6.2. Three and n-Dimensional Equations 3. Hyperbolic Equations with One Space Variable 3.1. Equations with Power-Law Nonlinearities 2 2 3.1.1. Equations of the Form ∂ w = ∂ w + aw + bwn + cw2n−1 ∂t2 ∂x2 3.1.2. Equations of the Form 3.1.3. Equations of the Form 3.1.4. Equations of the Form 3.1.5. Equations of the Form 3.1.6. Equations of the Form 3.1.7. Other Equations

∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2

= a ∂ w + f (x, t, w) ∂x2

2 2

2

= a ∂ w + f x, t, w, ∂x2 = =

= f (x) ∂ w + g x, t, w, ∂x2

2 awn ∂ w + f (x, w) ∂x2 ∂ a ∂x wn ∂w +f (w) ∂x

∂w ∂x ∂w ∂x

3.2. Equations with Exponential Nonlinearities 2 2 3.2.1. Equations of the Form ∂ w = a ∂ w + beβw + ceγw ∂t2 ∂x2 3.2.2. Equations of the Form 3.2.3. Equations of the Form 3.2.4. Other Equations

∂2w ∂t2 ∂2w ∂t2

= a ∂ w + f (x, t, w) ∂x2

2

2

= f (x) ∂ w + g x, t, w, ∂x2

∂w ∂x

3.3. Other Equations Involving Arbitrary Parameters 3.3.1. Equations with Hyperbolic Nonlinearities 3.3.2. Equations with Logarithmic Nonlinearities 3.3.3. Sine-Gordon Equation and Other Equations with Trigonometric Nonlinearities 2 ∂ 3.3.4. Equations of the Form ∂ w + a ∂w = ∂x f (w) ∂w ∂t ∂x ∂t2 3.3.5. Equations of the Form

∂2w ∂t2

+ f (w) ∂w = ∂t

∂ ∂x

g(w) ∂w ∂x

3.4. Equations Involving Arbitrary Functions 2 2 3.4.1. Equations of the Form ∂ w = a ∂ w + f (x, t, w) ∂t2 ∂x2 3.4.2. Equations of the Form 3.4.3. Equations of the Form 3.4.4. Equations of the Form 3.4.5. Equations of the Form

© 2004 by Chapman & Hall/CRC

∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2

2 2

= a ∂ w + f x, t, w, ∂x2 = =

2

= f (x) ∂ w + g x, t, w, ∂x2

∂w ∂x

∂w ∂x f (w) ∂ w + g x, t, w, ∂w 2 ∂x ∂x 2 f (x, w) ∂ w + g x, t, w, ∂w ∂x2 ∂x

3.4.6. Equations of the Form ∂ w = f (t, w) ∂ w + g x, t, w, ∂t2 ∂x2 3.4.7. Other Equations Linear in the Highest Derivatives 3.5. Equations of the Form

∂2w ∂x∂y

2

2

∂w ∂x

= F x, y, w,

∂w ∂w ∂x , ∂y

3.5.1. Equations Involving Arbitrary Parameters of the Form 3.5.2. Other Equations Involving Arbitrary Parameters 3.5.3. Equations Involving Arbitrary Functions 4. Hyperbolic Equations with Two or Three Space Variables

∂2w ∂x∂y

= f (w)

4.1. Equations with Two Space Variables Involving Power-Law Nonlinearities 2 ∂ ∂ 4.1.1. Equations of the Form ∂ w = ∂x f (x) ∂w + ∂y g(y) ∂w + awp ∂x ∂y ∂t2 4.1.2. Equations of the Form 4.1.3. Equations of the Form 4.1.4. Other Equations

∂2w ∂t2 ∂2w ∂t2 ∂ ∂ = a ∂x wn ∂w + b ∂y wk ∂w ∂x ∂y

=

∂ ∂x

f (w) ∂w + ∂x

∂ ∂y

g(w) ∂w ∂y

4.2. Equations with Two Space Variables Involving Exponential Nonlinearities 2 ∂ ∂ 4.2.1. Equations of the Form ∂ w = ∂x f (x) ∂w + ∂y g(y) ∂w + aeλw ∂t2 ∂x ∂y 4.2.2. Equations of the Form 4.2.3. Other Equations

∂2w ∂t2 ∂ ∂ = a ∂x eβw ∂w + b ∂y eλw ∂w ∂x ∂y

4.3. Nonlinear Telegraph Equations with Two Space Variables 4.3.1. Equations Involving Power-Law Nonlinearities 4.3.2. Equations Involving Exponential Nonlinearities 4.4. Equations with Two Space Variables Involving Arbitrary Functions 2 ∂ ∂ 4.4.1. Equations of the Form ∂ w = ∂x f (x) ∂w + ∂y g(y) ∂w + h(w) ∂x ∂y ∂t2 4.4.2. Equations of the Form 4.4.3. Other Equations

∂2w ∂t2

=

∂ ∂x

f (w) ∂w + ∂x

∂ ∂y

g(w) ∂w + h(w) ∂y

4.5. Equations with Three Space Variables Involving Arbitrary Parameters 2 ∂ ∂ ∂ 4.5.1. Equations of the Form ∂ w = ∂x f (x) ∂w + ∂y g(y) ∂w + ∂z h(z) ∂w + awp ∂x ∂y ∂z ∂t2 4.5.2. Equations of the Form 4.5.3. Equations of the Form 4.5.4. Equations of the Form seβw

∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2

=

∂ ∂w ∂ ∂w ∂ ∂w λw ∂x f (x) ∂x + ∂y g(y) ∂y + ∂z h(z) ∂z + ae ∂ ∂ ∂ = a ∂x wn ∂w + b ∂y wm ∂w + c ∂z wk ∂w + swp ∂x ∂y ∂z ∂ ∂ ∂ = a ∂x eλ1 w ∂w + b ∂y eλ2 w ∂w + c ∂z eλ3 w ∂w + ∂x ∂y ∂z

4.6. Equations with Three Space Variables Involving Arbitrary Functions 2 ∂ ∂ ∂ 4.6.1. Equations of the Form ∂ w = ∂x f1 (x) ∂w + ∂y f2 (y) ∂w + ∂z f3 (z) ∂w +g(w) ∂t2 ∂x ∂y ∂z 4.6.2. Equations of the Form ∂ w = ∂t2 4.6.3. Other Equations

2

∂ ∂x

∂ ∂ f1 (w) ∂w + ∂y f2 (w) ∂w + ∂z f3 (w) ∂w +g(w) ∂x ∂y ∂z

5. Elliptic Equations with Two Space Variables 5.1. Equations with Power-Law Nonlinearities 2 2 5.1.1. Equations of the Form ∂ w + ∂ w = aw + bwn + cw2n−1 ∂x2 ∂y 2 5.1.2. Equations of the Form 5.1.3. 5.1.4. 5.1.5. 5.1.6. Equations of the Form Equations of the Form Equations of the Form Other Equations Involving Arbitrary Parameters

2 ∂2w + ∂ w = f (x, y, w) ∂x2 ∂y 2 ∂ 2w ∂w ∂w ∂2w ∂x2 + a ∂y 2 = F x, y, w, ∂x , ∂y ∂w ∂ ∂w ∂ ∂x f1 (x, y) ∂x + ∂y f2 (x, y) ∂y = g(w) ∂w ∂ ∂w ∂ ∂x f1 (w) ∂x + ∂y f2 (w) ∂y = g(w)

© 2004 by Chapman & Hall/CRC

5.2. Equations with Exponential Nonlinearities 2 2 5.2.1. Equations of the Form ∂ w + ∂ w = a + beβw + ceγw ∂x2 ∂y 2 5.2.2. 5.2.3. 5.2.4. 5.2.5. Equations of the Form ∂ w + ∂ w = f (x, y, w) ∂x2 ∂y 2 ∂ ∂ Equations of the Form ∂x f1 (x, y) ∂w + ∂y f2 (x, y) ∂w = g(w) ∂x ∂y ∂ ∂ Equations of the Form ∂x f1 (w) ∂w + ∂y f2 (w) ∂w = g(w) ∂x ∂y Other Equations Involving Arbitrary Parameters

2 2

5.3. Equations Involving Other Nonlinearities 5.3.1. Equations with Hyperbolic Nonlinearities 5.3.2. Equations with Logarithmic Nonlinearities 5.3.3. Equations with Trigonometric Nonlinearities 5.4. Equations Involving Arbitrary Functions 2 2 5.4.1. Equations of the Form ∂ w + ∂ w = F (x, y, w) ∂x2 ∂y 2 5.4.2. 5.4.3. 5.4.4. 5.4.5. Equations of the Form a ∂ w + b ∂ w = F x, y, w, ∂w , ∂w ∂x ∂y ∂x2 ∂y 2 ∂ ∂ Heat and Mass Transfer Equations of the Form ∂x f (x) ∂w + ∂y g(y) ∂w = h(w) ∂x ∂y ∂ ∂ Equations of the Form ∂x f (x, y, w) ∂w + ∂y g(x, y, w) ∂w = h(x, y, w) ∂x ∂y Other Equations

2 2

6. Elliptic Equations with Three or More Space Variables 6.1. Equations with Three Space Variables Involving Power-Law Nonlinearities ∂ ∂ ∂ 6.1.1. Equations of the Form ∂x f (x) ∂w + ∂y g(y) ∂w + ∂z h(z) ∂w = awp ∂x ∂y ∂z ∂ ∂ ∂ 6.1.2. Equations of the Form ∂x f (w) ∂w + ∂y g(w) ∂w + ∂z g(w) ∂w = 0 ∂x ∂y ∂z 6.2. Equations with Three Space Variables Involving Exponential Nonlinearities ∂ ∂ ∂ 6.2.1. Equations of the Form ∂x f (x) ∂w + ∂y g(y) ∂w + ∂z h(z) ∂w = aeλw ∂x ∂y ∂z ∂ ∂ ∂ 6.2.2. Equations of the Form a1 ∂x eλ1 w ∂w +a2 ∂y eλ2 w ∂w +a3 ∂y eλ2 w ∂w = beβw 416 ∂x ∂y ∂y 6.3. Three-Dimensional Equations Involving Arbitrary Functions ∂ 6.3.1. Heat and Mass Transfer Equations of the Form ∂x f1 (x) ∂w + ∂x ∂w ∂ ∂z f3 (z) ∂z = g(w) 6.3.2. Heat and Mass Transfer Equations with Complicating Factors 6.3.3. Other Equations

∂ ∂y

f2 (y) ∂w + ∂y

6.4. Equations with n Independent Variables ∂w ∂ ∂w ∂ 6.4.1. Equations of the Form ∂x1 f1 (x1 ) ∂x1 +· · ·+ ∂xn fn (xn ) ∂xn = g(x1 , . . . , xn , w) 6.4.2. Other Equations 7. Equations Involving Mixed Derivatives and Some Other Equations 7.1. Equations Linear in the Mixed Derivative 7.1.1. Calogero Equation 7.1.2. Khokhlov–Zabolotskaya Equation 7.1.3. Equation of Unsteady Transonic Gas Flows 2 ∂2w 7.1.4. Equations of the Form ∂w ∂x∂y − ∂w ∂ w = F x, y, ∂y ∂x ∂y 2 7.1.5. Other Equations with Two Independent Variables 7.1.6. Other Equations with Three Independent Variables 7.2. Equations Quadratic in the Highest Derivatives 2 2 7.2.1. Equations of the Form ∂ w ∂ w = F (x, y) ∂x2 ∂y 2 7.2.2. Monge–Amp` re equation e 7.2.3. Equations of the Form

© 2004 by Chapman & Hall/CRC

∂w ∂w ∂x , ∂y

2 2 ∂2w 2 −∂w∂w = ∂x∂y ∂x2 ∂y 2 2 2 ∂2w 2 − ∂w∂w =F ∂x∂y ∂x2 ∂y 2

F (x, y) x, y, w,

∂w ∂w ∂x , ∂y

7.2.4. Equations of the Form 7.2.5. Other Equations

∂2w 2 ∂x∂y

= f (x, y) ∂ w ∂x2

2

∂2w ∂y 2

+ g(x, y)

7.3. Bellman Type Equations and Related Equations 7.3.1. Equations with Quadratic Nonlinearities 7.3.2. Equations with Power-Law Nonlinearities 8. Second›Order Equations of General Form 8.1. Equations Involving the First Derivative in t 2 8.1.1. Equations of the Form ∂w = F w, ∂w , ∂ w ∂t ∂x ∂x2 2 8.1.2. Equations of the Form ∂w = F t, w, ∂w , ∂ w ∂t ∂x ∂x2 2 8.1.3. Equations of the Form ∂w = F x, w, ∂w , ∂ w ∂t ∂x ∂x2 2 8.1.4. Equations of the Form ∂w = F x, t, w, ∂w , ∂ w ∂t ∂x ∂x2 2 8.1.5. Equations of the Form F x, t, w, ∂w , ∂w , ∂ w = 0 ∂t ∂x ∂x2 8.1.6. Equations with Three Independent Variables 8.2. Equations Involving Two or More Second Derivatives 2 2 8.2.1. Equations of the Form ∂ w = F w, ∂w , ∂ w ∂x ∂x2 ∂t2 2 2 8.2.2. Equations of the Form ∂ w = F x, t, w, ∂w , ∂w , ∂ w ∂x ∂t ∂x2 ∂t2 8.2.3. Equations Linear in the Mixed Derivative 8.2.4. Equations with Two Independent Variables, Nonlinear in Two or More Highest Derivatives 8.2.5. Equations with n Independent Variables 9. Third›Order Equations 9.1. Equations Involving the First Derivative in t 3 9.1.1. Korteweg–de Vries Equation ∂w + a ∂ w + bw ∂w = 0 ∂t ∂x3 ∂x 9.1.2. Cylindrical, Spherical, and Modiﬁed Korteweg–de Vries Equations 3 9.1.3. Generalized Korteweg–de Vries Equation ∂w + a ∂ w + f (w) ∂w = 0 ∂t ∂x ∂x3 9.1.4. Equations Reducible to the Korteweg–de Vries Equation 3 9.1.5. Equations of the Form ∂w + a ∂ w + f w, ∂w = 0 ∂t ∂x ∂x3 3 9.1.6. Equations of the Form ∂w + a ∂ w + F x, t, w, ∂w = 0 ∂t ∂x ∂x3 9.1.7. Burgers–Korteweg–de Vries Equation and Other Equations 9.2. Equations Involving the Second Derivative in t 9.2.1. Equations with Quadratic Nonlinearities 9.2.2. Other Equations 9.3. Hydrodynamic Boundary Layer Equations 9.3.1. Steady Hydrodynamic Boundary Layer Equations for a Newtonian Fluid 9.3.2. Steady Boundary Layer Equations for Non-Newtonian Fluids 9.3.3. Unsteady Boundary Layer Equations for a Newtonian Fluid 9.3.4. Unsteady Boundary Layer Equations for Non-Newtonian Fluids 9.3.5. Related Equations 9.4. Equations of Motion of Ideal Fluid (Euler Equations) 9.4.1. Stationary Equations 9.4.2. Nonstationary Equations 9.5. Other Third-Order Nonlinear Equations 9.5.1. Equations Involving Second-Order Mixed Derivatives 9.5.2. Equations Involving Third-Order Mixed Derivatives 3 3 9.5.3. Equations Involving ∂ w and ∂ w ∂x3 ∂y 3

© 2004 by Chapman & Hall/CRC

10. Fourth›Order Equations 10.1. Equations Involving the First Derivative in t 4 10.1.1. Equations of the Form ∂w = a ∂ w + F x, t, w, ∂t ∂x4 10.1.2. Other Equations 10.2. Equations Involving the Second Derivative in t 10.2.1. Boussinesq Equation and Its Modiﬁcations 10.2.2. Equations with Quadratic Nonlinearities 10.2.3. Other Equations 10.3. Equations Involving Mixed Derivatives 10.3.1. Kadomtsev–Petviashvili Equation 10.3.2. Stationary Hydrodynamic Equations (Navier–Stokes Equations) 10.3.3. Nonstationary Hydrodynamic Equations (Navier–Stokes equations) 10.3.4. Other Equations 11. Equations of Higher Orders 11.1. Equations Involving the First Derivative in t and Linear in the Highest Derivative 11.1.1. Fifth-Order Equations n 11.1.2. Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂t ∂xn n 11.1.3. Equations of the Form ∂w = a ∂ w + f (w) ∂w ∂t ∂xn ∂x n 11.1.4. Equations of the Form ∂w = a ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂t ∂xn ∂x n 11.1.5. Equations of the Form ∂w = a ∂ w + F x, t, w, ∂w ∂t ∂xn ∂x n n−1 w 11.1.6. Equations of the Form ∂w = a ∂ w + F x, t, w, ∂w , . . . , ∂ n−1 ∂t ∂xn ∂x ∂x n 11.1.7. Equations of the Form ∂w = aw ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂t ∂xn ∂x 11.1.8. Other Equations 11.2. General Form Equations Involving the First Derivative in t n 11.2.1. Equations of the Form ∂w = F w, ∂w , . . . , ∂ w ∂t ∂x ∂xn n 11.2.2. Equations of the Form ∂w = F t, w, ∂w , . . . , ∂ w ∂t ∂x ∂xn n 11.2.3. Equations of the Form ∂w = F x, w, ∂w , . . . , ∂ w ∂t ∂x ∂xn n 11.2.4. Equations of the Form ∂w = F x, t, w, ∂w , . . . , ∂ w ∂t ∂x ∂xn 11.3. Equations Involving the Second Derivative in t n 2 11.3.1. Equations of the Form ∂ w = a ∂ w + f (x, t, w) ∂t2 ∂xn 11.3.2. Equations of the Form 11.3.3. Equations of the Form 11.3.4. Equations of the Form 11.3.5. Equations of the Form

∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2 ∂2w ∂t2

n

∂w ∂x

= a ∂ w + F x, t, w, ∂xn = = =

∂w ∂x n−1 ∂ nw w a ∂xn + F x, t, w, ∂w , . . . , ∂ n−1 ∂x ∂x n aw ∂ w + f (x, t, w) ∂w + g(x, t, w) ∂xn ∂x n F x, t, w, ∂w , . . . , ∂ w ∂x ∂xn

11.4. Other Equations 11.4.1. Equations Involving Mixed Derivatives m n w 11.4.2. Equations Involving ∂ w and ∂ m ∂xn ∂y Supplements. Exact Methods for Solving Nonlinear Partial Differential Equations S.1. Classiﬁcation of Second-Order Semilinear Partial Differential Equations in Two Independent Variables S.1.1. Types of Equations. Characteristic Equation S.1.2. Canonical Form of Parabolic Equations S.1.3. Canonical Form of Hyperbolic Equations S.1.4. Canonical Form of Elliptic Equations

© 2004 by Chapman & Hall/CRC

S.2. Transformations of Equations of Mathematical Physics S.2.1. Point Transformations S.2.2. Hodograph Transformation S.2.3. Contact Transformations. Legendre and Euler Transformations S.2.4. B¨ cklund Transformations. Differential Substitutions a S.3. Traveling-Wave Solutions and Self-Similar Solutions. Similarity Methods S.3.1. Preliminary Remarks S.3.2. Traveling-Wave Solutions. Invariance of Equations Under Translations S.3.3. Self-Similar Solutions. Invariance of Equations Under Scaling Transformations S.3.4. Exponential Self-Similar Solutions. Equations Invariant Under Combined Translation and Scaling S.4. Method of Generalized Separation of Variables S.4.1. Introduction S.4.2. Structure of Generalized Separable Solutions S.4.3. Solution of Functional-Differential Equations by Differentiation S.4.4. Solution of Functional-Differential Equations by Splitting S.4.5. Simpliﬁed Scheme for Constructing Generalized Separable Solutions S.4.6. Titov–Galaktionov Method S.5. Method of Functional Separation of Variables S.5.1. Structure of Functional Separable Solutions S.5.2. Special Functional Separable Solutions S.5.3. Differentiation Method S.5.4. Splitting Method. Reduction to a Functional Equation with Two Variables S.5.5. Solutions of Some Nonlinear Functional Equations and Their Applications S.6. Generalized Similarity Reductions of Nonlinear Equations S.6.1. Clarkson–Kruskal Direct Method: a Special Form for Similarity Reduction S.6.2. Clarkson–Kruskal Direct Method: the General Form for Similarity Reduction S.6.3. Some Modiﬁcations and Generalizations S.7. Group Analysis Methods S.7.1. Classical Method for Symmetry Reductions S.7.2. Nonclassical Method for Symmetry Reductions S.8. Differential Constraints Method S.8.1. Description of the Method S.8.2. First-Order Differential Constraints S.8.3. Second- and Higher-Order Differential Constraints S.8.4. Connection Between the Differential Constraints Method and Other Methods S.9. Painlev´ Test for Nonlinear Equations of Mathematical Physics e S.9.1. Movable Singularities of Solutions of Ordinary Differential Equations S.9.2. Solutions of Partial Differential Equations with a Movable Pole. Description of the Method S.9.3. Examples of the Painlev´ Test Applications e S.10. Inverse Scattering Method S.10.1. Lax Pair Method S.10.2. Method Based on the Compatibility Condition for Two Linear Equations S.10.3. Method Based on Linear Integral Equations S.11. Conservation Laws S.11.1. Basic Deﬁnitions and Examples S.11.2. Equations Admitting Variational Formulation. Noetherian Symmetries

© 2004 by Chapman & Hall/CRC

S.12. Hyperbolic Systems of Quasilinear Equations S.12.1. Conservation Laws. Some Examples S.12.2. Cauchy Problem, Riemann Problem, and Initial-Boundary Value Problem S.12.3. Characteristic Lines. Hyperbolic Systems. Riemann Invariants S.12.4. Self-Similar Continuous Solutions. Rarefaction Waves S.12.5. Shock Waves. Rankine–Hugoniot Jump Conditions S.12.6. Evolutionary Shocks. Lax Condition (Various Formulations) S.12.7. Solutions for the Riemann Problem S.12.8. Initial-Boundary Value Problems of Special Form S.12.9. Examples of Nonstrict Hyperbolic Systems References

© 2004 by Chapman & Hall/CRC

AUTHORS

Andrei D. Polyanin, Ph.D., D.Sc., is a noted scientist of broad interests, who works in various areas of mathematics, mechanics, and chemical engineering sciences. A. D. Polyanin graduated from the Department of Mechanics and Mathematics of the Moscow State University in 1974. He received his Ph.D. degree in 1981 and D.Sc. degree in 1986 at the Institute for Problems in Mechanics of the Russian (former USSR) Academy of Sciences. Since 1975, A. D. Polyanin has been a member of the staff of the Institute for Problems in Mechanics of the Russian Academy of Sciences. He is a member of the Russian National Committee on Theoretical and Applied Mechanics. Professor Polyanin is an author of 33 books in English, Russian, German, and Bulgarian, as well as over 120 research papers and three patents. He has written a number of fundamental handbooks, including A. D. Polyanin and V. F. Zaitsev, Handbook of Exact Solutions for Ordinary Differential Equations, CRC Press, 1995 and 2002; A. D. Polyanin and A. V. Manzhirov, Handbook of Integral Equations, CRC Press, 1998; A. D. Polyanin, Handbook of Linear Partial Differential Equations for Engineers and Scientists, Chapman & Hall/CRC Press, 2002; A. D. Polyanin, V. F. Zaitsev, and A. Moussiaux, Handbook of First Order Partial Differential Equations, Taylor & Francis, 2002; and A. D. Polyanin and V. F. Zaitsev, Handbook of Nonlinear Mathematical Physics Equations, Fizmatlit, 2002. Professor Polyanin is Editor of the book series Differential and Integral Equations and Their Applications, Taylor & Francis, London, and Physical and Mathematical Reference Literature, Fizmatlit, Moscow. In 1991, A. D. Polyanin was awarded a Chaplygin Prize of the Russian Academy of Sciences for his research in mechanics. In 2001, he received an award from the Ministry of Education of the Russian Federation.

Address: Institute for Problems in Mechanics, RAS, 101 Vernadsky Avenue, Building 1, 119526 Moscow, Russia E›mail: polyanin@ipmnet.ru

Valentin F. Zaitsev, Ph.D., D.Sc., is a noted scientist in the ﬁelds of ordinary differential equations, mathematical physics, and nonlinear mechanics. V. F. Zaitsev graduated from the Radio Electronics Faculty of the Leningrad Polytechnical Institute (now Saint-Petersburg Technical University) in 1969 and received his Ph.D. degree in 1983 at the Leningrad State University. His Ph.D. thesis was devoted to the group approach to the study of some classes of ordinary differential equations. In 1992, Professor Zaitsev received his Doctor of Sciences degree; his D.Sc. thesis was dedicated to the discrete-group analysis of ordinary differential equations. In 1971–1996, V. F. Zaitsev worked in the Research Institute for Computational Mathematics and Control Processes of the St. Petersburg State University. Since 1996, Professor Zaitsev has been a member of the staff of the Russian State Pedagogical University (St. Petersburg). Professor Zaitsev has made important contributions to new methods in the theory of ordinary and partial differential equations. He is an author of more than 130 scientiﬁc publications, including 18 books and one patent.

Address: Russian State Pedagogical University, 48 Naberezhnaya reki Moiki, 191186 Saint-Petersburg, Russia E›mail: valentin zaitsev@mail.ru

© 2004 by Chapman & Hall/CRC

FOREWORD

Nonlinear partial differential equations are encountered in various ﬁelds of mathematics, physics, chemistry, and biology, and numerous applications. Exact (closed-form) solutions of differential equations play an important role in the proper understanding of qualitative features of many phenomena and processes in various areas of natural science. Exact solutions of nonlinear equations graphically demonstrate and allow unraveling the mechanisms of many complex nonlinear phenomena such as spatial localization of transfer processes, multiplicity or absence steady states under various conditions, existence of peaking regimes and many others. Furthermore, simple solutions are often used in teaching many courses as speciﬁc examples illustrating basic tenets of a theory that admit mathematical formulation. Even those special exact solutions that do not have a clear physical meaning can be used as “test problems” to verify the consistency and estimate errors of various numerical, asymptotic, and approximate analytical methods. Exact solutions can serve as a basis for perfecting and testing computer algebra software packages for solving differential equations. It is signiﬁcant that many equations of physics, chemistry, and biology contain empirical parameters or empirical functions. Exact solutions allow researchers to design and run experiments, by creating appropriate natural conditions, to determine these parameters or functions. This book contains more than 1600 nonlinear mathematical physics equations and nonlinear partial differential equations and their solutions. A large number of new exact solutions to nonlinear equations are described. Equations of parabolic, hyperbolic, elliptic, mixed, and general types are discussed. Second-, third-, fourth-, and higher-order nonlinear equations are considered. The book presents exact solutions to equations of heat and mass transfer, wave theory, nonlinear mechanics, hydrodynamics, gas dynamics, plasticity theory, nonlinear acoustics, combustion theory, nonlinear optics, theoretical physics, differential geometry, control theory, chemical engineering sciences, biology, and other ﬁelds. Special attention is paid to general-form equations that depend on arbitrary functions; exact solutions of such equations are of principal value for testing numerical and approximate methods. Almost all other equations contain one or more arbitrary parameters (in fact, this book deals with whole families of partial differential equations), which can be ﬁxed by the reader at will. In total, the handbook contains signiﬁcantly more nonlinear PDE’s and exact solutions than any other book currently available. The supplement of the book presents exact analytical methods for solving nonlinear mathematical physics equations. When selecting the material, the authors have given a pronounced preference to practical aspects of the matter; that is, to methods that allow effectively “constructing” exact solutions. Apart from the classical methods, the book also describes wide-range methods that have been greatly developed over the last decade (the nonclassical and direct methods for symmetry reductions, the differential constraints method, the method of generalized separation of variables, and others). For the reader’s better understanding of the methods, numerous examples of solving speciﬁc differential equations and systems of differential equations are given throughout the book. For the convenience of a wide audience with different mathematical backgrounds, the authors tried to do their best, wherever possible, to avoid special terminology. Therefore, some of the methods are outlined in a schematic and somewhat simpliﬁed manner, with necessary references made to books where these methods are considered in more detail. Many sections were written so that they could be read independently from each other. This allows the reader to quickly get to the heart of the matter. The handbook consists of chapters, sections, and subsections. Equations and formulas are numbered separately in each subsection. The equations within subsections are arranged in increasing order of complexity. The extensive table of contents provides rapid access to the desired equations.

© 2004 by Chapman & Hall/CRC

Separate parts of the book may be used by lecturers of universities and colleges for practical courses and lectures on nonlinear mathematical physics equations for graduate and postgraduate students. Furthermore, the books may be used as a database of test problems for numerical and approximate methods for solving nonlinear partial differential equations. We would like to express our deep gratitude to Alexei Zhurov for fruitful discussions and valuable remarks. The authors hope that this book will be helpful for a wide range of scientists, university teachers, engineers, and students engaged in the ﬁelds of mathematics, physics, mechanics, control, chemistry, and engineering sciences. Andrei D. Polyanin Valentin F. Zaitsev

© 2004 by Chapman & Hall/CRC

**SOME NOTATIONS AND REMARKS
**

Latin Characters

C1 , C 2 , . . . r, ϕ, z r, θ, ϕ t w x, y, z x1 , . . . , x n are arbitrary constants; cylindrical coordinates, r = x2 + y 2 and x = r cos ϕ, y = r sin ϕ; spherical coordinates, r = x2 + y 2 + z 2 and x = r sin θ cos ϕ, y = sin θ sin ϕ, z = r cos θ; time (t ≥ 0); unknown function (dependent variable); space (Cartesian) coordinates; Cartesian coordinates in n-dimensional space.

Greek Characters

∆ Laplace operator; in two-dimensional case, ∆ = in three-dimensional case, ∆ = ∆∆

∂2 ∂x2 ∂2 ∂x2

+

∂2 ∂y 2 ;

+

∂2 ∂y 2

+

∂2 ∂z 2 ;

**in n-dimensional case, ∆ =
**

∂ ∂x4

4

n k=1

∂2 ∂x2 k

;

biharmonic operator; in two-dimensional case, ∆∆ =

+ 2 ∂x∂∂y2 + 2

4

∂ ∂y 4 .

4

**Brief Notation for Derivatives
**

wx = ∂w , wt = ∂x df , fxx = fx = dx ∂w ∂2w ∂2w ∂2w , wxx = , wxt = , wtt = , ... (partial derivatives); ∂t ∂x2 ∂x∂t ∂t2 d2 f d3 f d4 f dn f (n) , fxxx = , fxxxx = , fx = (derivatives for f = f (x)). 2 3 4 dx dx dx dxn

**Brief Notation for Differential Operators
**

∂ ∂ ∂ ∂ , ∂y = , ∂t = , ∂w = (differential operators in x, y, t, and w); ∂x ∂y ∂t ∂w ∂ ∂ ∂ ∂ + wx + wxx + wxt + · · · (total differential operator in x); Dx = ∂x ∂w ∂wx ∂wt ∂ ∂ ∂ ∂ + wt + wxt + wtt +··· (total differential operator in t). Dt = ∂t ∂w ∂wx ∂wt In the last two relations, w is assumed to be dependent on x and t, w = w(x, t). ∂x =

Remarks

1. The book presents solutions of the following types: (a) expressible in terms of elementary functions explicitly, implicitly, or parametrically; (b) expressible in terms of elementary functions and integrals of elementary functions; (c) expressible in terms of elementary functions, functions involved in the equation (if the equation contains arbitrary functions), and integrals of the equation functions and/or other elementary functions; (d) expressible in terms of ordinary differential equations or ﬁnite systems of ordinary differential equations; (e) expressible in terms of solutions to nonlinear equations that can be reduced to linear partial differential equations or linear integral equations.

© 2004 by Chapman & Hall/CRC

αβ ≠ 0 w = ϕ(x) + ψ(t) w = ϕ(x)ψ(t) w = tα F (z). 8. and one unknown. When referencing a particular equation. Aris. ODE and PDE are conventional abbreviations for ordinary differential equation and partial differential equation. The present book does not consider ﬁrst-order nonlinear partial differential equations. will also be called a self-similar ¯¯ ¯ solution. As a rule. a solution of the form w = tα F (z). respectively. For equations in two independent variables. we use a notation like 3. and Moussiaux (2002).1. The book also deals with solutions described by equations with fewer new variables than those in the original equations. 6. If a formula or a solution contains an expression like a−2 assumption a ≠ 2 is implied. z = ϕ(t)x + ψ(t) For uniformity of presentation.2. z = αx + βt. ¢ ¡ © 2004 by Chapman & Hall/CRC . z = xtβ . z = xtβ w = ϕ(t)F (z). x and t. 3. the book does not present simple solutions that depend on only one of the variables involved in the original equation. For these equations. where x = x + C1 and t = t + C2 . Equations are numbered separately in each subsection. Solution name 1 2 3 4 5 6 7 8 ∗ ∗∗ Solution structure ∗ Traveling-wave solution w = F (z).2. (b) the cited source provides further information on the equation in question and their solutions. w. z = ϕ1 (x)ψ1 (t) + · · · + ϕn (x)ψn (t) Additive separable solution Multiplicative separable solution Self-similar solution ∗∗ Generalized self-similar solution Generalized separable solution Functional separable solution Generalized traveling-wave solution w = F (z).5.1. Though incomplete. and Amundson (1986. An expression that solves an equations in three independent variables and is determined by an equation in two independent variables will be called a two-dimensional solution. very simple and graphical classiﬁcation of solutions by their appearance is used in the book. z = xψ(t) w = ϕ1 (x)ψ1 (t) + · · · + ϕn (x)ψn (t) w = F (z). see Kamke (1965). which implies equation 5 from Subsection 3. Zaitsev. 4. f (x) . and Polyanin. ¯ ¯ Sometimes. 1989). the solution name and structure are as follows (x and t in the solutions below can be swapped): No. Rhee.2. 7. we also use this term in the cases where the variable t plays the role of a spatial coordinate. it is often not stated that the 5. This symbol indicates references to literature sources whenever: (a) at least one of the solutions was obtained in the cited source (even though the solution contained “correctable” misprints in signs or coefﬁcients).

1.Chapter 1 Parabolic Equations with One Space Variable 1. 2. z = x + λt. Then the functions w1 = w( x + C1 . ∂t ∂x2 Fisher equation. where the function u(ξ) is determined by the ordinary differential equation 1 auξξ + 2 ξuξ + u + bu2 = 0. Traveling-wave solution (λ is an arbitrary constant): w = w(z). and ecology. where C1 . t) is a solution of this equation. 3◦ . Equations with Power-Law Nonlinearities 2 1. Then the function 2 2 w1 = C1 w(C1 x + C2 . Equations of the Form ∂w = a ∂ w + bw + cw 2 ∂t ∂x2 1. Self-similar solution: w = t−1 u(ξ). are also solutions of the equation.1. biology.1. This is a special case of equation 1. Suppose w(x. For example. The kinetic function f (w) = aw(1 − w) models also an autocatalytic chain reaction in combustion theory. ∂t ∂x2 1◦ .3. ξ = xt−1/2 . combustion theory. C1 t + C3 ). This equation arises in heat and mass transfer. it describes the mass transfer in a two-component medium at rest with a volume chemical reaction of quasi-ﬁrst order. is also a solution of the equation. where C1 and C2 are arbitrary constants. Suppose w(x. t) is a solution of the equation in question.2 with m = 2. t + C2 ). C2 . ∂2w ∂w =a + bw2 . ∂2w ∂w = + aw(1 – w). © 2004 by Chapman & Hall/CRC . 1◦ . where the function w(z) is determined by the autonomous ordinary differential equation awzz − λwz + bw2 = 0. 2◦ .1. and C3 are arbitrary constants.

ϕ(ξ) = ℘(ξ + C3 . ∂t ∂x2 This is a special case of equation 1. ξ = xt−1/2 . b C1 x2 + C2 x + 6aC1 t + C3 3◦ . Ablowitz and A. 4◦ . w(x.1. – bw3 . =a 1◦ . £ ¤¢ Reference: P. 2◦ . w(x. √ −2 1 5 . are also solutions of the equation. Traveling-wave solution (λ is an arbitrary constant): ¡ w(x. The substitution U = 1 − w leads to an equation of the similar form ∂ 2U ∂U = − aU (1 − U ). t) = w = w(z). L. For the upper sign. Equations of the Form ∂w = a ∂ w + b0 + b1 w + b2 w 2 + b3 w 3 ∂t ∂x2 1. 1). Solutions: 2C1 x + C2 2a . t) = ξ 2 ϕ(ξ). t) = 1 + C exp − 6 at 1 6 √ −2 ξ = C1 exp where the function ϕ(ξ) is deﬁned implicitly by ξ= ¡ dϕ (4ϕ3 − 1) and C1 and C2 are arbitrary constants. 6 − C2 . Self-similar solution: w = t−1/2 u(ξ). 0. ¡ ¡ 5 w(x. Clarkson and E. t) is a solution of the equation in question. 2 1. .1. Danilov.2 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . Zeppetella (1978). t) = √ 2 1 5 1 + C exp − 6 at 6 −6a x 3◦ . Mansﬁeld (1994). ∂t ∂x2 £ ¤¢ References: M. J. Volosov (1995). A. t) = −1 + C exp − 6 at 6 6a x √ 1 + 2C exp − 5 at 1 −6a x 6 6 . C1 t + C3 ). Then the functions 2 w1 = C1 w( C1 x + C2 . 4◦ . G. V. V. Maslov. Traveling-wave solutions: w(x. C2 . the inversion of this relation corresponds to the classical Weierstrass elliptic function. where the function u(ξ) is determined by the ordinary differential equation 1 1 auξξ + 2 ξuξ + 2 u − bu3 = 0. ¡ ∂w ∂2w where C1 .2. P. A. Suppose w(x. and C3 are arbitrary constants. Traveling-wave solutions (C is an arbitrary constant): 6a x .2.5 with b0 = b1 = b2 = 0. © 2004 by Chapman & Hall/CRC ¡ ¡ ¡ 1 6 √ 6a x + 5 at . z = x + λt. and K. where the function w(z) is determined by the autonomous ordinary differential equation awzz − λwz − bw3 = 0.

C2 .1. Tabor (1989). Nucci and P.2 with m = 3. ∂2w ∂w + aw – bw3 .3. t + C2 ). C. § ¤¦ References: F. A. C2 . Solutions with a > 0 and b > 0: w= w= √ √ 1 1 C1 exp 2 2a x − C2 exp − 2 2a x a √ √ . 1 3 b C1 exp 1 2a x + C2 exp − 2 2a x + C3 exp − 2 at 2 √ √ 1 3 2C1 exp 2a x + C2 exp 2 2a x − 2 at a √ √ −1 . Clarkson (1992). and C3 are arbitrary constants. t) is a solution of the equation in question. ∂t ∂x2 This is a special case of equation 1. Cariello and M. ¥ ¥ where C1 and C2 are arbitrary constants. Solution with a < 0 (generalizes the solution of Item 3◦ ): w = exp ξ = exp 3 2 at 3 2 at 1 2 1 2 sin cos 2|a| x + C1 V (ξ). 6◦ . Solution with a < 0 and b > 0: w= |a| b cos sin 1 2 1 2 √ 2|a| x + C1 √ 3 2|a| x + C1 + C2 exp − 2 at .1. t) = where C is an arbitrary constant. © 2004 by Chapman & Hall/CRC ct b x+C 2a −1 . 1 3 b C1 exp 2a x + C2 exp 2 2a x − 2 at + C3 where C1 . 2 √ √ 1 3 3 z = C1 exp 1 2a x + 2 at + C2 exp − 2 2a x + 2 at + C3 . 2 where C1 . 3◦ . Traveling-wave solutions: ¥ ∂w ∂2w w(x. 5◦ . are also solutions of the equation (the signs are chosen arbitrarily). Suppose w(x.2. 3. 4◦ . Then the functions w1 = w( x + C1 . 2|a| x + C1 + C2 .2. 2◦ .5 with b1 = b0 = 0. Solution with a > 0 (generalizes the ﬁrst solution of Item 2◦ ): √ √ 1 3 3 w = C1 exp 1 2a x + 2 at − C2 exp − 2 2a x + 2 at U (z). =a 1◦ . where C1 and C2 are arbitrary constants. EQUATIONS WITH POWER-LAW NONLINEARITIES 3 2. where C1 and C2 are arbitrary constants.1. and C3 are arbitrary constants. = ∂t ∂x2 This is a special case of equation 1.1. 1◦ . and the function V = V (ξ) is determined by the autonomous ordinary differential equation aVξξ = −2bV 3 (whose solution can be written out in implicit form).5 with b0 = b2 = 0. – bw3 – cw2 .1. . See also equation 1. and the function U = U (z) is determined by the autonomous ordinary differential equation aUzz = 2bU 3 (whose solution can be written out in implicit form). M.

4 2◦ . ∂w = ∂2w 1◦ . Solutions: PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2a 1 ∂F . 1 2 (1 − a)x + 4 (1 − a2 )t + A ¨ c2 c2 c2 kc kc kc x− x + t exp √ x− t + C2 √ t + C3 . t) = w(x. ¨ w(x. This equation arises in population genetics and models the transmission of nerve impulses. t) = w(x. 1 1 2 2 a x + 2 a(2 − a)t 1 2 − a2 )t + A . √ 1 1 1 1 2 a + 2 a coth 4 2 ax + 4 a(a − 2)t + A . C2 . t) = w(x. w = 1 are stable. √ 1 1 1 1 2 (1 + a) + 2 (1 − a) coth 4 2 (1 − a)x + 4 (1 © 2004 by Chapman & Hall/CRC ¨ (1 + a) − (1 − a) coth 1 4 ¨ ¨ (1 + a) − (1 − a) tanh √ ¨ ¨ 1 4 2a √ . w = a is unstable. √ 1 1 1 1 2 (1 + a) + 2 (1 − a) tanh 4 2 (1 − a)x + 4 (1 A exp √ 1 √ . b F ∂x b 4. t) = w(x. 1 1 1 1 2 + 2 coth 4 2 x + 4 (1 − 2a)t + A . w = 1 are stable. Solutions: w(x. ¨ ¨ ¨ ¨ ¨ . Traveling-wave solutions (A. There are three stationary solutions: w = wk . A exp 1 2 (1 − a)x + 1 (1 − a2 )t + 1 2 2 √ 1 1 1 1 2 + 2 tanh 4 2 x + 4 (1 − 2a)t + A . if 0 < a < 1: the solutions w = 0. F = C1 exp √ 2b b 2b 2ab 2ab 2ab k = 1. t) = k where 2a 1 ∂F c − . 3◦ . t) = w(x. ∂t ∂x2 Fitzhugh–Nagumo equation. The linear stability analysis shows that if −1 ≤ a < 0: the solutions w = a. ¨ 2◦ . 2a √ . and w3 = a. √ 1 1 1 1 2 a + 2 a tanh 4 2 ax + 4 a(a − 2)t + A . and C3 are arbitrary constants. t + C2 exp − √ b 2b 2ab and C1 . t) = k where F = C1 x + kc c2 kc 2a x + t + C3 . t) = 1 + A exp ¨ 1 + A exp ¨ 1 1 2 2 2 (1 − a)x + 2 (1 − a )t + a √ . w = 0 is unstable. t) = w(x. B. where w1 = 0. There is a stationary nonhomogeneous solution that can be represented in implicit form (A and B are arbitrary constants): dw 1 4 4w 1 1 − 3 (a + 1)w3 + 2 aw2 + A = x + B. w2 = 1. b F ∂x k = 1. 1 2 (1 − a)x + 4 (1 − a2 )t + A − a2 )t + A . t) = w(x. t) = w(x. 1 2 x + 2 (2a − 1)t a √ . t) = w(x. – w(1 – w)(a – w). and C are arbitrary constants): w(x. t) = w(x.

ξ = x + λt. w2 = a − aw ax. t). EQUATIONS WITH POWER-LAW NONLINEARITIES 5 3◦ . Kawahara and M. Solutions are given by 5. V. The solutions of Item 2◦ are special cases of the traveling-wave solution w(x. See also Example 1 in Subsection S. t. z = ax a leads to an equation of the similar form with parameter a2 = 1 − a−1 : v(z. ∂t ∂x2 1 w(x. Polyanin (1996). 6◦ . τ ) = 1 − ∂2v 1 ∂v = − v(1 − v) 1 − − v . Let us give two transformations that preserve the form of the original equation. and the function w(ξ) is determined by the autonomous ordinary differential equation wξξ − λwξ = w(1 − w)(a − w). The substitution wξ = λy(w) leads to an Abel equation of the second kind: yyw − y = λ−2 aw − (a + 1)w2 + w3 . Clarkson (1992). z2 = √ 2 2 x+ 1 2 ax + a 1 2a − 1 t.1. 3 √ 2 can be found in Polyanin and Zaitsev β= − 2a . “Two-phase” solution: w(x. ¤ The transformation References for equation 1. Ibragimov (1994). Nucci and P. The abovesaid allows us to “multiply” exact solutions. where λ is an arbitrary constant. © 2004 by Chapman & Hall/CRC © z1 = √ 2 2 A exp(z1 ) + aB exp(z2 ) . C. 1 − a . t. 4◦ . τ = a2 t. M. a) is a solution of the equation in question. B. t) depends on the equation coefﬁcients. ∂t ∂x2 1◦ . F. D. Zaitsev and A.2. A exp(z1 ) + B exp(z2 ) + C − a t. H. t) = w(ξ). t) = β ∂F + λ. t) = © where A. a2 t. b3 (1) (2) . F ∂x where λ is any of the roots of the cubic equation b 3 λ3 + b 2 λ2 + b 1 λ + b 0 = 0 and the speciﬁc form of F = F (x. A. 5◦ . =a © ∂w ∂2w w(x. if w = w(x. N. and C are arbitrary constants. 1 − a−1 are also solutions of the equation. The substitution u = 1 − w leads to an equation of the similar form with parameter a 1 = 1 − a: ∂u ∂ 2 u = − u(1 − u)(1 − a − u). + b0 + b1 w + b2 w 2 + b3 w 3 . © The general solution of this equation with a = −1 and λ = (2003).2.4: T. ∂τ ∂z 2 a Therefore.1.1. then the functions w1 = 1 − w x.7. Tanaka (1983).

For q2 = q1 = 0. n = 1.2. ∂w =a ∂2w where C1 . where C1 . Sybochev (1991). q1 = − 1 (3b3 λ2 + 2b2 λ + b1 ). 2.4. 2 1. Traveling-wave solution: 3◦ . Mansﬁeld (1994). Yu. C1 t + C3 ). w = w(z). For q2 = 0 and q1 ≠ 0. © 2004 by Chapman & Hall/CRC 2◦ . A. 2a q2 = − (3) F (x. z = x + λt.3. and C3 are arbitrary constants. w(x. P.6 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE Introduce the notation p1 = −3a. w = w(x + γt). s 2 = − 2 p 1 q1 + 1 p 2 q1 . 2 2 1 1 1 1 k = − 2 q1 .1. Self-similar solution: 1 . F (x. Example. t) is a solution of this equation. we have β (b2 + 3b3 λ). C2 . Then the functions k−1 2k−2 2 w1 = C1 w( C1 x + C2 . Suppose w(x. C2 . t) = C1 exp(k1 x + s1 t) + C2 exp(k2 x + s2 t) + C3 . G. b0 = 0. 1. ¤ References: V. a = 1. 2 √ 1 1 z2 = 2 2b λ2 x + 2 bλ2 (λ2 − 2λ1 )t. and C3 are arbitrary constants. t where the function u(ξ) is determined by the ordinary differential equation 1 1 u + buk = 0. 2a Four cases are possible. Danilov and P. 2 sn = −kn p1 − kn p2 . t) = C1 (x − p2 t) + C2 exp[−q1 x + q1 (p2 − p1 q1 )t] + C3 . There is a traveling-wave solution. x ξ= √ . b1 w + b2 w 2 + b3 w 3 = −bw(w − λ1 )(w − λ2 ). Let 2◦ . A. For q2 ≠ 0 and q1 ≠ 4q2 . where λ is an arbitrary constant and the function w(z) is determined by the autonomous ordinary differential equation awzz − λwz + bwk = 0. auξξ + ξuξ + 2 k−1 w = t 1−k u(ξ). (4) F (x. are also solutions of the equation. + bwk . L. Kudryashov (1993). 2 1. t) = C1 exp(kx + s1 t) + C2 (kx + s2 t) exp(kx + s1 t) + C3 .1. one can obtain the solution C1 λ1 exp(z1 ) + C2 λ2 exp(z2 ) . By formulas (1)–(4) with λ = 0. 2 1. we have 1 k n = − 2 q1 1 2 2 q1 − 4q2 . p2 = β(b2 + 3b3 λ). N. 2 1. ∂t ∂x2 1◦ . For q2 ≠ 0 and q1 = 4q2 . t) = C1 (x − p2 t)2 + C2 (x − p2 t) − 2C1 p1 t + C3 . Clarkson and E. Equations of the Form ∂w = a ∂ w + f (w) ∂t ∂x2 1. t) = C1 exp(z1 ) + C2 exp(z2 ) + C3 where √ 1 z1 = 1 2b λ1 x + 2 bλ1 (λ1 − 2λ2 )t.3. s 1 = − 4 p 1 q1 + 2 p 2 q1 . F (x.

3. 3◦ . and β are given by λ= a(1 − m)(m + 3) . and ecology. biology. V. V. For (3) µ= a(m + 3)2 . V. m−1 w = ζ f (ζ) 2 m−1 . Polyanin (1996). combustion theory. EQUATIONS WITH POWER-LAW NONLINEARITIES 7 2. This equation arises in heat and mass transfer.1.1. The books by Polyanin and Zaitsev (1995. D. © 2004 by Chapman & Hall/CRC . 2(m + 1) β= b − . Volosov (1995). P. the class of traveling-wave solutions: w = w(z). b1 = −bµ2 λ−2 . t) = −β + C exp(λt 2 1−m . µ. and K. Kaliappan (1984). Traveling-wave solutions: w(x. G. (1) (2) w(x. where the function f (ζ) is given by f (ζ) = C1 − 4b ζ m+1 a(m − 1)2 −1/2 dζ + C2 . t) = β + C exp(λt µx) µx) 2 1−m . and ecology. F. where C is an arbitrary constant and the parameters λ. Solutions (1) and (2) are special cases of a wider class of solutions. Danilov. ∂2w ∂w = + aw + bwm + cw2m–1 . ∂2w ∂w + aw + bwm . and C1 and C2 are arbitrary constants. ¤ References: P. 1◦ . By the change of variable U (w) = µ2 λ−1 wz . ∂t ∂x2 This equation arises in heat and mass transfer. These are determined by the autonomous equation µ2 wzz − λwz + aw + bwm = 0. combustion theory. 2(m + 1) µ= a(1 − m)2 . z = µx + λt. 2(m + 1) λ = µ2 (m ≠ 1. m ≠ −3) the solution of equation (3) can be represented in parametric form as z= m+3 ln f (ζ). = ∂t ∂x2 Kolmogorov–Petrovskii–Piskunov equation. 2003) present exact solutions of this equation for some values of m and a1 (b1 is any). Zaitsev and A. equation (3) can be reduced to an Abel equation of the second kind: U Uw − U = a 1 w + b1 w m . Maslov. a 2◦ . biology. A. a1 = −aµ2 λ−2 .

t) = β + C exp(λt + µx) 1 1−m . equation (6) has also other traveling-wave solutions: ! z = x + σt. (5) 3◦ . t) = (1 − m) bt 4. and µ are determined by the system of algebraic equations aβ 2 + bβ + c = 0. w = w(z). F. In the general case. (1) where C is an arbitrary constant and the parameters β. The substitution u(x. (3) a + mbw − mb2 wm mb In the special case a = 0. Solution (1) is a special case of a wider class of traveling-wave solutions. which generate four exact solutions of the original equation. For a = 0. whose general solutions for some m (no constraints are imposed on a.8 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1◦ . (6) u ∂t ∂x 1 − m ∂x Solution (1) corresponds to a particular solution of (6) that has the form u = β + C exp(ωt + µx). that are determined by the autonomous equation wzz − σwz + aw + bwm + cw2m−1 = 0. Traveling-wave solutions: w(x. 2003). Polyanin (1996). where C is an arbitrary constant. µ − λ + (1 − m)[2a + (b/β) = 0. µ − (1 − m)λ + a(1 − m) = 0. D. ◦ 2 . © 2004 by Chapman & Hall/CRC . u = w1−m leads to an equation with quadratic nonlinearity: ∂2u m ∂u 2 ∂u =u 2 + + a(1 − m)u2 + b(1 − m)u + c(1 − m). The substitution U (w) = wz brings (5) to the Abel equation U Uw − σU + aw + bwm + cw2m−1 = 0. ∂t ∂x2 Traveling-wave solution: w = w(z). ∂w = ∂2w − c x + C. ¤ Reference: V. where λ is an arbitrary constant and the function w(z) is determined by the autonomous ordinary differential equation wzz − λwz + awm−1 + bmwm − mb2 w2m−1 = 0. z = x + λt. it can be shown that a one-parameter family of solutions to equation (1) satisﬁes the ﬁrst-order equation a . b. Zaitsev and A. (1) For λ = 1. 2 2 2 (2) (3) (4) The quadratic equation (2) for β can be solved independently. λ. (2) wz = w − bwm + mb Integrating (2) yields a solution in implicit form (A is any): 1 dw = z + A. m + awm–1 + bmwm – mb2 w2m–1 . system (2)–(4) gives four sets of the parameters. it follows from (3) that w(z) = C exp (1 − m)z + b 1 1−m . and c) can be found in the books by Polyanin and Zaitsev (1995.

∂w =a ∂2w 2n + m + 2 1 ξuξ + u + bξ m uk = 0. ∂w =a ∂2w . w) = sz k wn . Equations of the Form ∂w = a ∂ w + f (x.2 with f (z. one arrives at the simpler equation ∂2w ∂w = a 2 + cw + k1 wn1 + k2 wn2 . w) = sez wn . and acoustics. Self-similar solution: 2n+m+2 2(1−k) u(ξ). ∂w =a ∂2w " 2.5. . is also a solution of the equation. It is used for describing wave processes in gas dynamics.3 with f (t) = b. w) = s1 z k + s2 wn . z = x + bt.1. hydrodynamics.2 with f (z. ∂t ∂x2 This is a special case of equation 1.6.2 with f (z.1.3. ∂w =a ∂2w " " 1.1. ∂x2 ∂w = ∂2w +w ∂w © 2004 by Chapman & Hall/CRC " 3.4. t. t where the function u = u(ξ) is determined by the ordinary differential equation w=t auξξ + 5.2 with f (z. 2 ∂t ∂x ∂x This is a special case of equation 1. 4. + 1 (bx + ct)k + 2 wn .1.1. + btn xm wk . ∂t ∂x Burgers equation. On passing from t. Equations of the Form ∂w = a ∂ w + f (w) ∂w + g(w) ∂t ∂x ∂x2 1. t) is a solution of this equation. ∂t ∂x2 This is a special case of equation 1.1. 2 2(k − 1) 2◦ . x ξ= √ .6.1 to 1. Suppose w(x. ∂t ∂x2 1◦ . ∂t ∂x2 This is a special case of equation 1. ∂w =a ∂2w + ebx+ct wn . ∂t ∂z special cases of which are discussed in Subsections 1.6.1. EQUATIONS WITH POWER-LAW NONLINEARITIES 9 2 1. C 2k−2 t).2. Then the function w1 = C 2n+m+2 w(C k−1 x.6. + (w + bx + ct)k . x to the new variables t.1. " 2 1.1. 2. ∂t ∂x2 This is a special case of equation 1.1.6. where C is an arbitrary constant. w) = s(w + z)k . w) ∂t ∂x2 + (bx + ct)k wn . ∂2w ∂w ∂w =a +b + cw + k1 wn1 + k2 wn2 .

and erf z ≡ √ π (also called the probability integral). . w(x. exp(−ξ 2 ) dξ is the error function 3◦ . t) = √ 4πt ∞ −∞ −∞ < x < ∞. 2 where A. t) = A−x . Hopf (1950).10 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1◦ . w(x. The transformation (1) and equation 1. t) = 2 where 1 F (x.3. t) = λ + w(x. Hopf (1950). Solution: w(x. t) = u ∂x where u = u(x. (2) ∂t ∂x2 For details about this equation. x + λt + A 4x + 2A . t). Then the function 2 w1 = C1 w(C1 x + C1 C2 t + C3 . C1 t + C4 ) + C2 . Solutions: w(x. C4 are arbitrary constants. J. w(x. © 2004 by Chapman & Hall/CRC . t) = √ 4(t + λ) π(t + λ) A erf x+B √ 2 t+λ z 0 +C −1 . t) = 3 x + 6xt + 3Ax + B 2λ . t) = λ2 t λx + B λ 2 tanh 2 +A λ t+A − λx − B . Cole (1951). t) = −λ + 2A tan A(λt − x) + B . C.2. t) = 1 + A exp(−λ2 t − λx) exp A(x − λt)] − B . B. which is a generalized Burgers equation. t) is a solution of the linear heat equation ∂u ∂ 2 u = . B+t 2 .6. t) = −λ + A exp A(x − λt) + B w(x. Other solutions can be obtained using the following formula (Hopf–Cole transformation): 2 ∂u . where C1 . Reference: E. 2λ cos(λx + A) . $ ¤# $ ¤# References: E. t) = −λ + 2A tanh A(x − λt) + B . Remark. Suppose w(x. were encountered much earlier in Fortsyth (1906). ∂x (x − ξ)2 1 − 4t 2 ξ 0 exp − f (ξ ) dξ dξ. and λ are arbitrary constants. see the books Tikhonov and Samarskii (1990) and Polyanin (2002). . w(x. t) = 2 x + Ax + 2t + B 6(x2 + 2t + A) . Cauchy problem. 2◦ . t) = B exp(λ2 t) + sin(λx + A) (x + B)2 2A exp − w(x. w(x. 4◦ . w(x. (1) w(x. . . Initial condition: w = f (x) at t = 0. t) is a solution of the equation in question. w(x. ∂ ln F (x. is also a solution of the equation.

3: ∂ 2u ∂u ∂u = a 2 + bu . where C1 . Suppose w(x. F. N. t) − 2◦ .5. Rudenko and C. ∂t ∂x2 ∂x w = u(z. Traveling-wave solution: ∂w + σw ∂w =a ∂2w ∂w ∂2w ∂w + σw =a + b1 w + b0 . t) + ct.5.1.1.1. z = x + λt. The Burgers equation is connected with the linear heat equation (2) by the B¨ cklund transfora mation ∂u 1 − uw = 0. b1 leads to a simpler equation of the form 1. and C3 are arbitrary constants. ∂x 2 ∂u 1 ∂(uw) − = 0. σ(1 + C2 e−bt ) 2◦ . Suppose w(x. The scaling of the independent variables x = z.1. is also a solution of the equation. ∂t ∂x ∂x2 1◦ .1. ∂t ∂x ∂x2 1◦ . w = w(z).5. t + C3 + Cbebt .5. V. 1 z = x + 2 bct2 . ∂2w ∂w ∂w =a + bw . is also a solution of the equation. Degenerate solution: w(x. t) is a solution of this equation. V. Then the function w1 = w x − C1 σebt + C2 . EQUATIONS WITH POWER-LAW NONLINEARITIES 11 5◦ . Soluyan (1975). Burgers (1948). leads to the Burgers equation 1. M. 3◦ .2: J. C2 . ∂t 2 ∂x References for equation 1. where C1 . t) is a solution of this equation. © 2004 by Chapman & Hall/CRC & ¤% 3. t) = 6. C2 . b(x + C1 ) .1. H. and C3 are arbitrary constants.2: ∂w ∂w ∂ 2 w = +w . Zaitsev and A. Ibragimov (1994). ∂τ ∂z 2 ∂z 4. a a Unnormalized Burgers equation. t = 2 τ leads to b b an equation of the form 1. Then the function w1 = w x − C1 σeb1 t + C2 . b1 ∂u ∂ 2u ∂u + σu = a 2 + b1 u. ∂t ∂z ∂z . t + C3 + Cb1 eb1 t . b0 . ∂w ∂t ∂x2 The transformation =a ∂2w + bw ∂w ∂x + c. + bw. ∂t ∂z ∂z 5. Polyanin (1996).5: w = u(z. where λ is an arbitrary constant and the function w(z) is determined by the autonomous ordinary differential equation awzz − σwwz − λwz + bw = 0. I. O. The transformation z =x+σ b0 t. D.

s 1 = − 4 p 1 q1 + 2 p 2 q1 . For q2 ≠ 0 and q1 ≠ 4q2 . 2 sn = −kn p1 − kn p2 . © 2004 by Chapman & Hall/CRC . ( ¤' λ= bk . ∂w ∂2w ∂w + σw =a + b0 + b1 w + b2 w 2 + b3 w 3 .3. 2 1. For q2 ≠ 0 and q1 = 4q2 . ∂t ∂x2 ∂x 9a Solution: k(−1 + C1 e4λx ) . Introduce the notation: p1 = −βσ − 3a.5. t) = C1 + C2 exp Ax + A b1 σ 2ab2 + t . σ z(x. A. C2 .4. t) depends on the equation coefﬁcients. we have z(x. z ∂x (1) Here. βσ + 2a βσ + 2a Four cases are possible. Case b3 = b2 = 0. t) = C1 exp(kx + s1 t) + C2 (kx + s2 t) exp(kx + s1 t) + C3 . 2ab2 where λ is a root of the quadratic equation b2 λ2 + b1 λ + b0 = 0.1. 1. z(x. Volosov (2000). q2 = − . 2. ( ¤' Reference: N. w= 1 + C1 e4λx + C2 e2λx+bkλt where C1 and C2 are arbitrary constants. For q2 = q1 = 0. For q2 = 0 and q1 ≠ 0.1. Solutions are given by (1) with β=− 2a . Case b3 = 0 and b2 ≠ 0. q1 = − p2 = λσ + βb2 + 3βλb3 . 2 2 1 1 1 1 k = − 2 q1 . 2 1.4–1. where C1 . and the speciﬁc form of z = z(x. z(x. t) = C1 (x − p2 t) + C2 exp[−q1 x + q1 (p2 − p1 q1 )t] + C3 . t) = β ∂z + λ. 3◦ . t) = C1 (x − p2 t)2 + C2 (x − p2 t) − 2C1 p1 t + C3 . Case b3 ≠ 0. 1◦ . Kudryashov (1993). b3 λ3 + b2 λ2 + b1 λ + b0 = 0. 3b3 λ2 + 2b2 λ + b1 βb2 + 3βλb3 . s 2 = − 2 p 1 q1 + 1 p 2 q1 .6. n = 1. t) = C1 exp(k1 x + s1 t) + C2 exp(k2 x + s2 t) + C3 . ) 1 k n = − 2 q1 1 2 2 q1 − 4q2 . 12a Private communication: K. A.5. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w b2 ∂w + bw =a + w(w – k)(w + k). 2b2 σ A= σ(b1 + 2b2 λ) . See equations 1.1. and C3 are arbitrary constants. 2◦ . ∂t ∂x ∂x2 Solutions of the equation are given by 8. z(x.2. β and λ are any of the roots of the respective quadratic and cubic equations b3 β 2 + σβ + 2a = 0.12 7. w(x. 2 1.

z = x + C1 e−bt . 2 ∂t ∂x ∂x 1◦ . where the function w(z) is determined by the ordinary differential equation awzz + bzwz + swk = 0. t) + C3 t −1 if k ≠ 1. 1◦ . w(x.1 with f (w) = sw k . ∂w =a ∂2w + btn ∂w 1◦ . x to the new variables b n+1 t . z = x + λt. C1 t).2. one arrives at the simpler equation t. and C3 are arbitrary constants. t) is a solution of this equation. Traveling-wave solution: b λm z + . t) is a solution of this equation. w) ∂w + g(x. ξ = x|t|− 2 .7 for f (w) = bw m . t) + C3 (1 − k)t−k w3 = w(x − bC3 ln |t|. w) ∂t ∂x2 ∂x 0 1. Suppose w(x. ∂t ∂x This is a special case of equation 1.1. C 1 t + C 3 . ∂x2 3. 2◦ . are also solutions of the equation. ∂t ∂z special cases of which are discussed in Subsections 1. EQUATIONS WITH POWER-LAW NONLINEARITIES 13 ∂w ∂2w ∂w + bwm =a . ∂w ∂t t ∂x Modiﬁed Burgers equation.1.1. t. A wider family of traveling-wave solutions is presented in 1. t. C2 . 2 ∂t ∂x ∂x This is a special case of equation 1. Then the functions 2 w1 = C1 w(C1 x + C2 .6. On passing from t. Equations of the Form ∂w = a ∂ w + f (x. t + C2 ). where C1 . w1 = w(x + C1 e−bt . + k w + bw ∂w =a ∂2w ∂x2 . m 2m w1 = C1 w C1 x + C2 .1. where C1 . 1 −1/m 2 1. t) = C exp − 3◦ . 2◦ .6. Suppose w(x. and C3 are arbitrary constants.3. Then the function 9. a λ(m + 1) where C and λ are arbitrary constants. There is a self-similar solution of the form 1 w(ξ. t) = |t|− 2m ϕ(ξ). + cw + k1 wm1 + k2 wm2 .3. is also a solution of the equation. ∂2w ∂w ∂w =a + (bx + c) + wk . Then the function where C1 and C2 are arbitrary constants. z = x + n+1 ∂ 2w ∂w = a 2 + cw + k1 wm1 + k2 wm2 .2. w2 = w(x − bC3 t1−k . is also a solution of the equation.1. © 2004 by Chapman & Hall/CRC . Generalized traveling-wave solution: w = w(z).6. C2 .6. 2. Suppose w(x.1 to 1. t) is a solution of the equation in question.3 with f (t) = btn . if k = 1.

3. if k = 1. t) = cxk + stn . . Generalized traveling-wave solution: w = w(z). ∂t x ∂x ∂x ∂w + bw ∂w =a 1 ∂ ∂w 8.6. . Degenerate solution linear in x: (1 − k)x + C1 C2 tk + bt x + C1 w(x. Then the function ∂x2 w1 = w(x + C1 e−bt . ∂t ∂x ∂x 6. ∂2θ ∂θ ∂θ = a 2 + cxk . t) = − b θ ∂x where the function θ = θ(x. ∂t ∂x 1◦ . z = x + C1 e−bt . w(x. w(x. Self-similar solution: if k ≠ 1. J. t) = 3◦ . ∂x2 ∂w =a ∂2w + bw ∂w . ∂w ∂2w ∂w ∂w + bw =a + cxk + ckxk–1 w. Solution: 2a 1 ∂θ . is also a solution of the equation. E. The variable x plays the role of the radial coordinate. w(x. 2 ∂t ∂x ∂x ∂x Solution: 2a 1 ∂θ . + c(x + t)k . 2◦ . 4. ∂2w ∂w ∂w =a + bw + cxk + tn . Musielak (1996). 3 2 ¤1 Reference: S. ∂t ∂x x ∂x ∂x x2 Cylindrical Burgers equation. and Z.6.14 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . Suppose w(x. t) satisﬁes the linear equation 5. t) is a solution of this equation. ∂w =a ∂2w + (bx + cwk ) ∂w © 2004 by Chapman & Hall/CRC 3 7. Nerney.2 with f (x. ∂t ∂x This is a special case of equation 1. where the function u = u(z) is determined by the ordinary differential equation auzz + 1 2z − bu uz + – w 1 2 − k u = 0. t) = u(z)t−1/2 . Schmahl. t + C2 ). t) satisﬁes the linear heat equation with axial symmetry x a ∂ ∂θ ∂θ = x .3. t) = c(x + st)k .2 with f (x. w(x. t) = − b θ ∂x where the function θ = θ(x. E. z = xt−1/2 . 2 ∂t ∂x ∂x This is a special case of equation 1. t) = t(C2 + b ln |t|) where C1 and C2 are arbitrary constants. where C1 and C2 are arbitrary constants. where the function w(z) is determined by the ordinary differential equation awzz + (bz + cwk )wz = 0.

Solutions: a ln |Ax + B| + C. x to the new variables t.6. B. b a a 2 λ2 t + ln |cos(λx + A)| + B. ∂w ∂t =a ∂2w ∂x2 + (bwm + ctk ) 4 9. c k+1 t . t) = A bt Ax + B. and λ are arbitrary constants.1. EQUATIONS WITH POWER-LAW NONLINEARITIES 15 ∂2w ∂w ∂w + (bwm + ct + ) =a . b 2 w(x. g(t) = ctk .9: ∂2w ∂w ∂w = a 2 + bwm . ∂t ∂x For details about this equation. we obtain an equation of the On passing from t. g(t) = ct + s. and h(w) = 0.13 with f (z.11 with f (w) = bw m . we obtain an equation of the form 1. Equations of the Form ∂w = a ∂ w + b ∂w ∂t ∂x2 ∂x 2 + f (x. t) = a ln |u(x. t. t) = ln |x4 + 12ax2 t + 12a2 t2 + A| + B.9: ∂w ∂2w ∂w = a 2 + bwm . ∂w ∂x .1. x to the new variables t.3. t) = ln |x3 + 6axt + Ax + B| + C.6. ∂t ∂z ∂z 4 4 11. z = x + 2 ct2 + st.5. w) = s1 z k wn and g(z. z = x + k+1 form 1. t) = − b b where A. t)| b ∂2u ∂u =a 2.1.5. C. The substitution leads to the linear heat equation w(x. w(x. 4bt 2b 5 a ln |x2 + 2at + Ax + B| + C. w(x) = w(x. b a w(x. t) = − w(x. and h(w) = 0. t) = a (x + A)2 − ln t + B. b a w(x.6. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). This is a special case of equation 1. 1◦ . ∂t ∂z ∂z 10.3. 2 ∂t ∂x ∂x This is a special case of equation 1. ∂w ∂t =a ∂2w ∂x2 + 1 (bx + ct)k wn ∂w ∂x + 2 (bx + ct)p wq .1. w) = s2 z p wq . ∂w ∂t =a ∂2w ∂x2 +b ∂w ∂x 2 . w) 1.1. © 2004 by Chapman & Hall/CRC .3. 2 1. This is a special case of equation 1. 2◦ .11 with f (w) = bw m . 1 On passing from t.7.

t + C2 ) + C3 ect . Solutions for a < 0: √ w(x. ∂w ∂t =a ∂2w ∂x2 +a PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂x 2 + b. t) = ϕ(t) + ψ(t) A exp x −a + B exp −x −a . V. where C1 . The substitution u = ew leads to the constant coefﬁcient linear equation ∂2u ∂u = a 2 + bu. Then the functions w1 = w( x + C1 . D. and C3 are arbitrary constants. ∂t ∂x2 ∂x 1◦ . are also solutions of the equation. © 2004 by Chapman & Hall/CRC . 9 ¤8 (1) (2) Dividing equation (1) by (2) termwise yields the ﬁrst-order equation (2ϕ − 1)ψϕ ψ = ϕ2 + 4ABψ 2 . t) = 4. and the functions ϕ(t) and ψ(t) are determined by the autonomous system of ﬁrst-order ordinary differential equations ϕt = a ϕ2 + ψ 2 . Posashkov (1989). √ C2 1 + exp −at x −a . Generalized separable solution for a < 0: √ √ w(x. ψt = a(2ϕ − 1)ψ. t) = 2 C1 − at (C1 − at) where C1 and C2 are arbitrary constants. Galaktionov and S. ∂t ∂x2 ∂x ◦ 1 . The ﬁrst formula represents a multiplicative separable solution and the second one. t) = ϕ(t) + ψ(t) cos x a + C . ∂w = ∂2w + ∂w 2 s 2a ct c(x + C2 )2 e ln |C1 e−ct − 4b| + C3 ect − . Polyanin (1996). t) is a solution of this equation. − −ct − 4b C1 e C1 c + aw2 . and the functions ϕ(t) and ψ(t) are determined by the autonomous system of ﬁrst-order ordinary differential equations ϕt = a ϕ2 + 4ABψ 2 . Generalized separable solution for a > 0: √ w(x.16 2. ψt = a(2ϕ − 1)ψ. 7 7 where A and B are arbitrary constants. where C is an arbitrary constant. 7 ∂w =a ∂2w +b ∂w 2 2◦ . F. A. Suppose w(x. 2◦ . Dividing equation (3) by (4) termwise yields a ﬁrst-order equation. w(x. Zaitsev and A. C2 . A. Generalized separable solution quadratic in x: w(x. + cw + . a generalized separable solution. (3) (4) References: V. 3◦ . t) = C1 exp −at x −a . ∂t ∂x 6 3.

∂t ∂x ∂x This is a special case of equation 1. This is a special case of equation 1.4 with f (x. Generalized separable solution for c < 0: A w(x. ψt = (2bcϕ + s − ac)ψ.8. ∂t ∂x ∂x ∂2w ∂w 2 2. ∂t ∂x2 ∂x 1◦ . The solution of system (2). A. V. ∂t ∂x2 ∂x This is a special case of equation 1. . + cw2 + tn w + ktm . and h(t) = stk . t) = ctm .6. h = const. The substitution u = ew leads to the linear equation ∂x2 ∂2u ∂u ∂u = a 2 + btn + ctm u.5. Equations of the Form ∂w = a ∂ w + f x. see equation 1. w. D. ∂t ∂x2 ∂x ∂x The substitution u = ew leads to the constant coefﬁcient linear equation ∂w =a +a ∂2u ∂u ∂u =a 2 +b + cu. EQUATIONS WITH POWER-LAW NONLINEARITIES 17 + bcw2 + w + k.6. ∂w ∂t =a ∂2w ∂x2 + btn ∂w ∂x 2 + ctm w + tk .6. For more complicated generalized separable solutions that involve hyperbolic and trigonometric functions of x. Galaktionov and S. Zaitsev and A.1 with f (t) = btn . F.6. and h = ktm .1. t) = btn and g(x.6.6. (3) is given by ϕ(t) = λ + ψ(t) = 2bcλ + s . Posashkov (1989). ∂w ∂t ∂x2 ∂x @ 6.6. Polyanin (1996). g. ∂t ∂x ∂x @ ∂w =a ∂2w +a ∂w 2 + btn ∂w 3. (1) (2) (3) . g = stn . © 2004 by Chapman & Hall/CRC @ 5. 2 1. ∂w +b + c. g(t) = ctm . A. where C1 and C2 are arbitrary constants and λ = λ1 and λ = λ2 are roots of the quadratic equation 2◦ .2 with f = const.2 with f . C1 exp −(2bcλ + s)t − bc 2 C1 C2 exp −(2bcλ + s + ac)t C1 exp −(2bcλ + s)t − bc bcλ2 + sλ + k = 0. ∂w =a ∂2w +b ∂w 2 √ x −c .1.1. + ctm . C ¤B References: V. t) = ϕ(t) + ψ(t) exp where the functions ϕ(t) and ψ(t) are determined by the autonomous system of ﬁrst-order ordinary differential equations ϕt = bcϕ2 + sϕ + k. ∂w =a ∂2w +b ∂w 2 1. t.

and λ are arbitrary constants.9. C2 t + C4 ). where C1 .6. © 2004 by Chapman & Hall/CRC . k. 7. C − 2at where C1 .6. . Then the function −2 w1 = C1 C2 w(C1 x + C3 . t. 4◦ . This is a special case of equation 1.1 with f (t) = beλt .6.1.6.6.9. g(t) = b. Equations of the Form ∂w = aw k ∂ w + f x. t) is a solution of this equation. The substitution 1 wa+1 if a ≠ −1. C2 . B. . Traveling-wave solution in implicit form: ak 2 dw = kx + λt + C2 . 6. see equation 1.18 4.1.6.1. Suppose w(x. ∂w ∂t = ∂2w ∂x2 + awk ∂w ∂x 2 . . This is a special case of equation 1. ∂w ∂t ∂x2 ∂x 1.6.8 with f (w) = a/w. and C are arbitrary constants. see equation 1. Items 5 ◦ to 8◦ . w. ∂2w ∂w = aw . 3◦ . ∂2w a ∂w = + 2 ∂t ∂x w ∂w ∂x 2 . and h(t) = seνt . This is a special case of equation 1. C4 are arbitrary constants. 2◦ . . For other exact solutions.6.7. ∂w ∂t = ∂2w ∂x2 + awm ∂w ∂x 2 + (bx + ct + ) D ∂w ∂x . t) = where A. g(t) = ceµt .1.8. ∂w ∂t =a ∂2w ∂x2 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE + beλt ∂w ∂x 2 + ceµt w + eνt . 5. The substitution a wk+1 dw u = exp k+1 leads to the constant coefﬁcient linear equation ∂t u = ∂xx u. D This is a special case of equation 1. and h(t) = ct + s. For k = 0. Multiplicative separable solution: w(x. see equation 1.10 with f (w) = aw m . ∂t ∂x2 1◦ . u= a+1 ln |w| if a = −1 leads to the constant coefﬁcient linear equation ∂t u = ∂xx u. is also a solution of the equation. and for k = −1. λ ln |w| + C1 x2 + Ax + B . 2 1.5.8 with f (w) = aw k .1.18 with m = 1.

t) = Ax + B + bt. G ¤F + cw2 + kw + . w(x.1. and the function w(z) is determined by the autonomous ordinary differential equation ak 2 wwzz − λwz + bw + c = 0. ∂2w ∂w + b. Galaktionov and S. Solutions: w(x. where C1 . ∂t ∂x2 This is a special case of equation 1. where k and λ are arbitrary constants. C − 2at 4a where A.3 with f (t) = ct + d and g(t) = st + k. 2◦ . B. = aw ∂t ∂x2 1◦ . ∂w = aw ∂2w . ∂t ∂x2 This is a special case of equation 1. z = kx + λt. C2 . ∂t ∂x2 1◦ . w(x. and the function w(z) is determined by the autonomous ordinary differential equation ak 2 wwzz − λwz + b = 0. = aw Reference: V. A. E 4. 2◦ . Generalized separable solutions: c w(x. The ﬁrst solution is degenerate and the second one is a generalized separable solution. EQUATIONS WITH POWER-LAW NONLINEARITIES 19 2. and C3 are arbitrary constants. t) = Bbe−bt − 2a where A. Then the function −1 w1 = C1 w(C1 x + C2 . ξ = x/t. ∂w ∂2w + bw2 + (ct + d)w + t + k. is also a solution of the equation. C1 t + C3 ). 3. 3◦ .9. Posashkov (1989). and C are arbitrary constants. Self-similar solution: w = tU (ξ). Traveling-wave solution: w = w(z).1. b b(x + A)2 − Bce−bt − 2act + C . where the function U (ξ) is determined by the autonomous ordinary differential equation aU Uξξ + ξUξ − U + b = 0.6. Suppose w(x.9. t) is a solution of this equation. ∂w = aw ∂2w where k and λ are arbitrary constants. Traveling-wave solution: w = w(z). © 2004 by Chapman & Hall/CRC E 5. A. and C are arbitrary constants (the ﬁrst solution is degenerate). + bw + c.9 with b = 0.1. z = kx + λt. t) = b x2 + Ax + B − (C − 2at). 4◦ . t) = Aebt x + Bebt − . B.

C1 . t) = ϕ(t) + ψ(t) exp( λx). Generalized separable solutions involving an exponential of x: −c . H P ¤I + pw + q. λ= 1/2 ϕt = cϕ2 + kϕ + s. © 2004 by Chapman & Hall/CRC (aλ2 ϕ + 2cϕ + k) dt . and C2 are arbitrary constants.10. 3 The general solution of this system with A ≠ 0 has the form f2 (t) = 3 f0 (t) = ϕ(t) dt + 3B. .6.20 6. and f0 (t) are determined by the system of ordinary differential equations 2 f1 = 18Af0 − 3 f1 f2 . and s(t) = q.10. (1) a+b where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations w(x. V. A. ∂t ∂x2 ∂x 1◦ . h(t) = p. Galaktionov (1995).6. Reference: J. ∂t ∂x2 3 ∂x Generalized separable solution: w(x. and s(t) = pt + k. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w ∂w +b = aw + (ct + d)w + pt + k. ∂x2 ∂2w ∂w 2 ∂w = aw +b + cw2 + kw + . H Q ∂w = aw ∂2w +b ∂w 2 +c ∂w 9. R. 36A2 t 1 1 ϕ(t) ϕ(t) dt + B + 2 9A 6A2 where the function ϕ(t) is deﬁned implicitly by ϕ(t) dt + B + (C1 + 72A2 abϕ − 8ϕ3 )−1/2 dϕ = t + C2 . 1 3Ax3 + f2 (t)x2 + f1 (t)x + f0 (t) . f1 (t). 8. 2 ∂t ∂x ∂x This is a special case of equation 1. f1 (t) = 3 1 A ϕ(t) dt + B 2 + 1 ϕ(t). one can ﬁnd ϕ = ϕ(t) in explicit form. + kϕ + s On computing the integral. Integrating (2) yields (2) (3) dϕ = t + C1 .2 with f (t) ≡ 0. ψt = (aλ2 ϕ + 2cϕ + k)ψ.2 with f (t) = b. The solution of equation (3) is expressed in terms of ϕ(t) as cϕ2 ψ(t) = C2 exp where C1 and C2 are arbitrary constants. 2 f2 = 6Af1 − 2 f2 . g(t) = c. t) = 7. h(t) = ct + d. King (1993). and B. 3 2 f0 = 2f0 f2 − 2 f1 + ab. 2A 1 ϕ (t). g(t) = b. a where A is an arbitrary constant and the functions f2 (t). ∂t ∂x ∂x This is a special case of equation 1. ∂w = aw ∂2w – 2 a ∂w 2 + b.

1. t) = ϕ(t) + ψ(t) cos(λx + A). H.1. where C1 . C1 t + C3 ).1. and C3 are arbitrary constants. w(x.1. Suppose w(x. Then the function 2 −1 2 w1 = C1 w(C1 x + C2 . 11. Then the functions 2 2 −1 w1 = C1 w( C1 x + C2 . S ¤R ∂w = aw2 ∂2w Reference: N. A. =a ∂t ∂x v 2 ∂x Therefore the solutions of the original equation are expressed via solutions of the linear heat equation ∂2u ∂u =a 2 ∂t ∂y by the relations ∂u . The transformation w = 1/u. ∂t ∂x2 1◦ . t) is a solution of this equation. ∂t ∂x2 The substitution w = 1/v leads to an equation of the form 1. and C3 are arbitrary constants.1 with f (t) = k and g(t) = s. t) = ϕ(t) + ψ(t) sinh(λx + A). see Items 2◦ to 4◦ of equation 1. + bw–1 . C1 t + C3 ).3: 1 ∂v ∂ ∂v .11. C2 . 10. λ = w(x.19 with m = 2 and b = −1. are also solutions of the equation. x = u.10. where C1 .10. ∂t ∂x2 This is a special case of equation 1. a+b The functions ϕ = ϕ(t) and ψ = ψ(t) are determined by autonomous systems of ﬁrst-order ordinary differential equations (these systems can be reduced to a single ﬁrst-order equation each). 2◦ . Posashkov (1989). ∂w = aw2 ∂2w ∂w = aw2 ∂2w 1◦ . + bw2 . A. . ∂t ∂x u2 ∂x a 12. t) = C1 (x + C2 )2 + C3 exp(2aC1 t) − where C1 . Galaktionov and S.1. For details about these solutions. t) = ϕ(t) + ψ(t) cosh(λx + A).2: ∂ 1 ∂u b ∂u = − . t) in explicit form. is also a solution of the equation. . t) is a solution of the equation in question. S ¤R Reference: V. There are also generalized separable solutions that involve hyperbolic and trigonometric functions (A is an arbitrary constant): w(x. EQUATIONS WITH POWER-LAW NONLINEARITIES 21 2◦ . Ibragimov (1985). Suppose w(x. C2 . C2 . λ= λ= −c a+b −c a+b 1/2 . Functional separable solutions: T w(x. w= ∂y The variable y should be eliminated to obtain w = w(x. τ = at leads to an equation of the form 1. 1/2 c . 2◦ .9. © 2004 by Chapman & Hall/CRC T b aC1 1/2 . 1/2 .6. and C3 are arbitrary constants.

Setting C1 = 0 in the last relation. where C is an arbitrary constant. ∂t ∂x2 This is a special case of equation 1. = aw2 ∂t ∂x2 Functional separable solutions: U w(x. ∂t ∂x2 Functional separable solutions: w(x. 1 A ϕ(t) dt + B 2 + 1 ϕ(t). .1. t) = a−1/3 3Ax3 + f2 (t)x2 + f1 (t)x + f0 (t) 16. ∂2u 2 ∂u ∂u = au 2 − a ∂t ∂x 3 ∂x 2 ∂w = aw3 ∂2w + 3b.9.9. Reference: G. one obtains the 1 function ϕ in explicit form: ϕ = − 2 (t + C2 )−2 . The substitution w = u1/3 leads to an equation of the form 1. Semenov (1999). I. © 2004 by Chapman & Hall/CRC U W ¤V + bw–2 .18 with m = 3. 1 ψt = 2 aψ(4ϕχ − ψ 2 ) + 2bψ.22 13. and C2 are arbitrary constants. and A. f2 (t) = 3 f0 (t) = ϕ(t) dt + 3B. C1 . C2 .1. f1 (t) = 3 1/3 ∂w = aw3 ∂2w . It follows from the ﬁrst two equations that ϕ = Cψ. Functional separable solution: w(x. ∂2w ∂w = aw4 + bw + cw–1 . t) = ϕ(t)x2 + ψ(t)x + χ(t). 14. B. and C3 are arbitrary constants. ∂t ∂x2 This is a special case of equation 1. t) = a−1/3 3Ax3 + f2 (t)x2 + f1 (t)x + f0 (t) Here.1. Therefore the equation in question has a generalized separable solution of the form w(x. A.19 with m = 3 and b = −2. 2A 1 ϕ (t). Rudykh and E. 36A2 t 1 1 ϕ(t) ϕ(t) dt + B + 2 9A 6A2 where the function ϕ(t) is deﬁned implicitly by U ϕ(t) dt + B + (C1 − 8ϕ3 )−1/2 dϕ = t + C2 . where the functions ϕ(t). 1/3 . PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w + bw + cw–1 . where C1 . 2 1 χt = 2 aχ(4ϕχ − ψ 2 ) + 2bχ + 2c.9. . and χ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = 1 aϕ(4ϕχ − ψ 2 ) + 2bϕ.7: 15. ψ(t). F (t) = exp aC1 e2bt + 2bt . t) = bC1 e2bt (x + C2 )2 + C3 F (t) + 2cF (t) dt F (t) 1/2 .

ϕ(t) = C + 2a(m − 2) t. w = w(z).11. and. β= . Traveling-wave solution in implicit form: βx + λt + C2 dw = . and the function f = f (x) is determined by the autonomous ordinary differential equation amfxx + λf 1−m = 0 (its solution can be written out in implicit form).6. t) = ( βx + βλt + A)1/m . is also a solution of the equation. and the function U = U (ζ) is determined by the ordinary differential equation aU m Uζζ = βU − 1 (mβ + 1)ζUζ . a(1 − m) m(x − A)2 2a(2 − m)(B − t) m 1/m w(x. t) = A|t + B| m−2 + w(x. . Solutions: w(x) = Ax + B. Self-similar solution: x z= √ . ∂2w ∂w + bxm w5 . λw1−m + C1 aβ 2 (1 − m) where C1 . C. where β is an arbitrary constant. mλ . 1/m m(m−3) (x + A)2 + B(x + A)m |ϕ(t)| 2 ϕ(t) . C4 are arbitrary constants.1.1. C2 . 4◦ . its order can be reduced. and λ are arbitrary constants (the ﬁrst solution is degenerate). where C1 . Then the function w1 = C1 2◦ . ∂2w ∂w = awm . . Suppose w(x. and λ are arbitrary constants. . ∂t ∂x2 1◦ . t) = (λt + A)−1/m f (x). Multiplicative separable solution: w(x. w(x. = aw4 ∂t ∂x2 This is a special case of equation 1. C2 t + C4 ).1 with f (x) = bxm . m where A. © 2004 by Chapman & Hall/CRC . 18. t) is a solution of this equation. t) = X −2/m C2 1/m w(C1 x + C3 . 2 This equation is generalized homogeneous. hence. t) = (x + C)2 m 2a(m − 2) t + B 1/m . ζ = xt− mβ+1 2 . w(x. EQUATIONS WITH POWER-LAW NONLINEARITIES 23 17. To λ = 0 there corresponds a stationary solution. . Self-similar solution of a more general form: w = tβ U (ζ). β. 5◦ . B. 3◦ . where λ is an arbitrary constant. 6◦ . t where the function w(z) is determined by the ordinary differential equation 2aw m wzz zwz = 0. and to C1 = 0 there corresponds the second solution in Item 2◦ .

where λ is an arbitrary constant.1. w = (At + B)−1/m ψ(u). Traveling-wave solutions: w = w(z). w1 = C1 w( C1 Y 8◦ . its order can be reduced. ξ = xeλmt . B.6. hence. and the function ψ = ψ(u) is determined by the autonomous ordinary differential equation A aψ m ψuu = Akψu − ψ. 19.18: τ= 1 bmt e + const. bm ∂ 2U ∂U = aU m 2 . ∂2w ∂w + bwk .11. ∂t ∂x ∂x which is considered Subsection 1. leads to an equation of the form 1.10. the transformation w(x. Generalized self-similar solution: w = e−2λt ϕ(ξ).11. For m ≠ 1. This equation is generalized homogeneous. 1 ξ = xt k−m−1 2(1−k) . t) = ebt U (x. Solution: where C1 . The substitution u = w 1−m leads to the equation m ∂u ∂ ∂u =a u 1−m . the substitution u = w 1−m leads to the equation m ∂u k−m ∂ ∂u =a u 1−m + b(1 − m)u 1−m . and C3 are arbitrary constants. u = x + k ln(At + B). C1 t + C3 ). = awm ∂t ∂x2 1◦ .4 with f (t) = btn and g(t) = ctk . τ ). z = kx + λt. For k = 1. m 9◦ .24 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 7◦ . 3◦ .9. where k and λ are arbitrary constants. where A. Suppose w(x. are also solutions of the equation. C2 . ∂t ∂x This is a special case of equation 1. t) is a solution of this equation. Self-similar solution for k ≠ 1: w = t 1−k u(ξ).1. and the function ϕ = ϕ(ξ) is determined by the ordinary differential equation aϕm ϕξξ = λmξϕξ − 2λϕ. and. aum uξξ + 2(1 − k) 1−k 4◦ . and k are arbitrary constants. ∂x2 ∂w = awm ∂2w + bxtn ∂w © 2004 by Chapman & Hall/CRC .1. 2◦ . 5◦ . and the function w(z) is determined by the autonomous ordinary differential equation awm wzz − λwz + bwk = 0. ∂τ ∂x 20. ∂t ∂x ∂x which is considered in Subsection 1. Then the functions k−m−1 2k−2 2 x + C2 . + ctk w. where the function u(ξ) is determined by the ordinary differential equation 1 m−k+1 ξuξ + buk − u = 0.

where the functions f = f (ξ) and g(ξ) are determined by the system of ordinary differential equations (fξ )2 − f fξξ = afξξξ . Solution in parametric form: ◦ x = (6at + C1 )ξ + C2 ξ 2 + C3 . The above solution was obtained. taking into account (1) and (3). t) = C1 x + aC1 t + C2 . Solution in parametric form: x = tf (ξ) + g(ξ). t) = − References: D. Suppose a solution of equation (1) is known. from a solution of the hydrodynamic boundary layer equation (see 9. + 6a(t + C2 ) |t + C2 |1/3 (x + C1 )2 + C3 |x + C1 |1/2 |C2 − 6at|−5/8 .10. ◦ f (ξ) = 6a(ξ + C)−1 . © 2004 by Chapman & Hall/CRC a ¤` a ¤` 2 . D. Reference: A. Traveling-wave solution in implicit form: 3 . w(x.1. and C3 are arbitrary constants. . (3) f dξ . 2003): g(ξ) = C1 + C2 f + C3 f f = f (ξ). w(x. Polyanin and V.1. C2 . (1) (2) The order of equation (1) can be reduced by two.10. Solutions: 2 w(x.1. Zaitsev (2002). w = tfξ (ξ) + gξ (ξ). Polyanin and V. EQUATIONS WITH POWER-LAW NONLINEARITIES 25 1. w 1◦ .1. with the help of the Mises transformation. we obtain another solution. t) = C2 − 6at where C1 .1. that the ﬁrst solution in (4) leads to the solution of Item 3◦ . For other solutions. ∂t ∂x ∂x This is a special case of equation 1. ψ= ψ dξ − f ψ dξ . ∂w =a ∂ ∂w C3 (x + C1 )2 . Equations of the Form ∂w = a ∂ w m ∂w ∂t ∂x ∂x 1.10. g2 = f (ξ).1. Substituting the second relation of (4) into (1). One can see. Items 5 ◦ and 7◦ ). Remark. The general solution of equation (1) can be represented in the form (see Polyanin and Zaitsev. D. The second particular solution follows from the comparison of (1) and (2). F. where C and λ are arbitrary constants. Equations of this form admit traveling-wave solutions w = w(kx + λt). ◦ 2 w − C2 ln |w + C2 | = C1 x + aC1 t + C3 .7 with m = 1. .1. Zwillinger (1989). Zaitsev (2002). 1 1 exp − 2 (fξ ) a It is not difﬁcult to verify that equation (1) has the following particular solutions: (4) f (ξ) = Ceλξ − aλ. 4◦ .7 with m = 1. F. 5 .3. w = −(6at + C1 )ξ 2 − 2C2 ξ 3 . fξ gξ − f gξξ = agξξξ . see Items 4◦ to 9◦ of equation 1. A. Equation (2) is linear in g and has two linearly independent particular solutions g1 = 1.

y) = 2 C1 x C1 C3 exp − at + C2 at + C2 −1+ C1 x at + C2 −1 . 2 sinh (C1 x + C3 ) 2 C2 − 2aC1 t . t) = Transformation (2) means that the dependent variable z is taken to be the independent variable. ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ 1 ∂w =a . Solutions: 2 w(x. t): ∂2u ∂u =a 2. S. (1) ∂t ∂x ∂x2 By the hodograph transformation x = u. the dependent one. © 2004 by Chapman & Hall/CRC d −1/2 . y) = cos2 (C1 x + C3 ) where C1 . and the independent variable x. and C3 are arbitrary constants.26 2. √ −1/2 C1 2at ln 2 . ∂t ∂y (3) d w(x. w(x. Aristov (1999). . Pukhnachov (1987). w(x. y) = C2 + C3 exp(aC2 t − C1 x) w(x. 3. 2a(t + C1 ) .7 with m = −2. y) = (x + C2 )2 + C3 (t + C1 )2 2 C1 . ∂z and then integrate the resulting equation with 2◦ . w(x. Introduce a new unknown z = z(x. w(x. c ¤b References: V. C2 . C2 . Solutions: 2 w(x. the second is self-similar.1.10. (2) equation (1) can be reduced to a linear heat equation for u = u(y.7 with m = −1. ∂t ∂x w ∂x This is a special case of equation 1. w(x.1. The ﬁrst solution is of the traveling-wave type. and C3 are arbitrary constants. and the third is a functional separable solution. y) = w(x. 2a where C1 . y) = (C1 x − aC1 t + C2 )−1 . ∂t ∂x ∂x This is a special case of equation 1. V. N. y) = cosh2 (C1 x + C3 ) 2 2aC1 t + C2 . y) = (2at + C1 )(x + C2 )−2 . w(x.10. 1◦ . t) = (2C1 x − 2aC1 t + C2 )−1/2 . z = y. t) by w = ∂x respect to x to obtain −2 2 ∂z ∂ z ∂z =a . t) = x x t d ∂w =a ∂ w–2 ∂w C1 (x + C2 )2 + C3 exp(C1 t) . 2 2aC1 t + C2 .

t) is a solution of the equation in question. t ) ¯ ¯ ∂x ¯ dt . W. t) = w(x. B2 . Feix (1981). t) = 1 . ∂w ∂t =a ∂ ∂x w–4/3 ∂w ∂x . This is a special case of equation 1. Functional separable solution: w(x. . J. H. B1 . Suppose w(x. and M. C2 . The transformation x= ¯ x x0 t t0 w(y. (4) w= ∂y The variable y should be eliminated from (4) to obtain w = w(x. t) = where C1 . τ ) ∂w (x. t) = ϕ4 (t)x4 + ϕ3 (t)x3 + ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) © 2004 by Chapman & Hall/CRC −3/4 g (x + C1 )2 + C3 (t + C2 )2 a(t + C2 ) −3/4 . . The ﬁrst solution is of the traveling-wave type. A. ¯ ∂t ∂x ¯ The inversion of transformation (5) is given by x= f ¤e x ¯ x0 ¯ ¯ w(x . Kumei (1980). leads to the linear equation ∂2w ¯ ∂w ¯ =a 2. R. L. t) of the original equation are expressed via solutions u = u(y. f ¤e References: L. A2 x + B2 where A1 . and C3 are arbitrary constants. 2 w(x. C1 . 1982). Fijalkow. Munier. and C2 are arbitrary constants (A1 B2 − A2 B1 ≠ 0). E. Solutions: w(x. and the other are functional separable solutions. V. 3◦ . g w(x. the second is a solution in multiplicative separable form. ◦ 2 w(x. t).1. τ ) ∂x dτ . x=x0 ¯ t = t − t0 . Gutierres. EQUATIONS WITH POWER-LAW NONLINEARITIES 27 Solutions w = w(x. t) = (at + C1 )3/4 [(x + C2 )(C3 x + C2 C3 + 1)]−3/2 . w(x. −3/4 (x + C1 )2 + C3 (t + C2 )2 (x + C1 )4 a(t + C2 ) .1. t) in explicit form. is also a solution of the equation.7 with m = −4/3 (the equation admits more invariant solutions than for m ≠ −4/3). Then the function w1 = |A1 B2 − A2 B1 |3/2 C1 (A2 x + B2 )3 −3/4 w A1 x + B1 . 1◦ . t) dy + a w−2 (x. w(x. t) dx + ¯ ¯ t ¯ t0 ∂ w(x. x = u(y. t) of the linear equation (3) according to −1 ∂u . t) ¯ ¯ ¯ References: M. C 1 t + C2 . w(x. 4. t) = ¯ ¯ ¯ 1 . Storm (1951). t) = ( 2C1 x3 + C2 x4 − 3aC1 x4 t)−3/4 . J. 2 .10.1. t) (5) where x0 and t0 are any numbers. N. A2 . Ovsiannikov (1959. G. Bluman and S. x=x0 ¯ ¯ ¯ ¯ t = t − t0 . R. 3◦ . Burgan. t) = ( 2C1 x − 3aC1 t + C2 )−3/4 . w(x. Ibragimov (1985).

7 with m = −3/2. Semenov (1998). R. C1 . Gazizov. x = v. w(x. ∂τ ∂ξ ∂ξ and A. A. I. 4◦ . y = t − i ¤h Reference: N. Functional separable solution: 6. There are exact solutions of the following forms: 1 . H. t) = x−3 G(z).7 with m = −2/3. 2A 1 ϕ (t).10.1. Burgan. where ∂v ∂ξ t = τ . H. x tx2 . Fijalkow (1984). ϕ3 = −aϕ2 ϕ3 + 6aϕ1 ϕ4 .1. Munier. Ibragimov (1994). 1 ϕ2 = −aϕ2 + 3 aϕ1 ϕ3 + 12aϕ0 ϕ4 . Feix (1981). © 2004 by Chapman & Hall/CRC p i ¤h References: A. R. t) = a2/3 3Ax3 + f2 (t)x2 + f1 (t)x + f0 (t) Here. and E. ∂t ∂x ∂x 1◦ . Solution: 2 . Feix. R. J. For other solutions.1. Fijalkow. and N. Sh. Galaktionov (1995).1. I. Ibragimov (1994). N. f1 (t) = 3 −2/3 . R. G. J. . Gutierres. Setting C1 = 0 in this relation. we ﬁnd ϕ in explicit form: ϕ = − 1 (t + C2 )−2 . B. ∂ ∂w ∂w =a w–3/2 . ∂t ∂x ∂x This is a special case of equation 1. ◦ 5 . Rudykh and E. z = (x + 1)2 w(x. The transformation ◦ ∂w =a ∂ w–2/3 ∂w w = (C − 4at)3/2 (C − 4at)3/2 − x2 −3/2 . = u. 5. Akhatov. f2 (t) = 3 f0 (t) = ϕ(t) dt + 3B.4: ∂ ∂u ∂u =a u−4/3 . see equation 1. ϕ4 = − 3 aϕ2 + 2aϕ2 ϕ4 . 3 4 The prime denotes a derivative with respect to t. H. A. Burgan. 36A2 t 1 1 ϕ(t) dt + B + ϕ(t) 9A2 6A2 where the function ϕ(t) is deﬁned implicitly by ϕ(t) dt + B + (C1 − 8ϕ3 )−1/2 dϕ = t + C2 . Munier.7 with m = −4/3. 2 2◦ . . 1 A ϕ(t) dt + B 2 + 1 ϕ(t). and M.28 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE where the functions ϕk = ϕk (t) are determined by the system of ordinary differential equations ϕ1 = −aϕ1 ϕ2 + 6aϕ0 ϕ3 . see equation 1. J. w(x.10. 2 2 References: V. R. M. 1◦ .10. K. t) = x−3 F (y). Ibragimov (1989). A. and C2 are arbitrary constants. leads to an equation of the form 1. For other solutions. w = 1/u. i ¤h 3 ϕ0 = −a 4 ϕ2 + 2aϕ0 ϕ2 . E.10.

and λ are arbitrary constants. and ﬂows in porous media. 1 m m(x − A)2 2a(m + 2)(B − t) . A. w(x. t © 2004 by Chapman & Hall/CRC . . S. Kurdyumov. Mikhailov (1995). λw + C1 q w = w(z). for x > λt. exact solutions of nonlinear equations are given only for the domain of their spatial localization. * For the sake of brevity. B. x = 0 (t > 0). Solutions:* w(x) = (Ax + B) m+1 .1. P. k(t − x/λ)1/m 0 ak m /m.1. The third solution for B > 0 and the fourth solution for B < 0 correspond to blow-up regimes (the solution increases without bound on a ﬁnite time interval). see also equations 1. and A. Barenblatt (1952). 2◦ . Then the function m 2 w1 = C1 w(C2 x + C3 . C. w(x. −2/3. t) = q 1 k = λm/a. P. Semenov (1998). . Example. where C1 . B. −3/2. −1.7. I. 1 m m m(x + A)2 − m(2m+3) + B|x + A| m+1 |ϕ(t)| 2(m+1)2 w(x. z = x + λt. A. and λ. References: Ya.1 to 1. C1 C2 t + C4 ). For example. t) = ϕ(t) . Suppose w(x. t) = ( kx + kλt + A)1/m .1. Samarskii. A solution satisfying the initial and boundary conditions w=0 w = kt1/m is given by w(x. here and henceforth. −4/3. Galaktionov. t) is a solution of the equation in question. 1 m m w(x. C4 are arbitrary constants. G. C1 . for 0 ≤ x ≤ λt. ∂w ∂t =a ∂ ∂x wm ∂w ∂x . . it describes unsteady heat transfer in a quiescent medium with the heat diffusivity being a power-law function of temperature. A. EQUATIONS WITH POWER-LAW NONLINEARITIES 29 7. 3◦ . V. Kompaneets (1950). 1◦ . Traveling-wave solutions: where the function w(z) is deﬁned implicitly by a wm dw = C2 + z.1. ϕ(t) = C − 2a(m + 2)t. This equation often occurs in nonlinear problems of heat and mass transfer. and to C1 = 0 there corresponds the second solution in Item 2◦ . and C2 are arbitrary constants. is also a solution of the equation.7. t) = A|t + B|− m+2 − (x + C)2 m 2a(m + 2) t + B . I. For m = 1. combustion theory. Zel’dovich and A.6. S. . To λ = 0 there corresponds a stationary solution. Rudykh and E. −2. where w 0. A. t) = where λ = s ¤r at at t=0 (x > 0). where A. G.

Multiplicative separable solution: w(x. (4) m mC where α = a(m+2) and β = a(m+2) . With the boundary conditions w = 1 at z = 0. am(m + 2) where the function w(z) is determined by the ordinary differential equation 2a wm wz z + zwz = 0. Performing the change of variable ϕ = F m in equation (3). m) = bk ξ k . x z= √ t (0 ≤ x < ∞). . w = t− m+2 F (ξ). . m) for 0 ≤ Z ≤ 1.30 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 4◦ . (3) where C is an arbitrary constant. Here. . 2003) present general solutions of equation (4) for m = −1 and m = 1. bf m+2 and λ. ∞ 2a . and C2 are arbitrary constants. m) P (ξ. To the particular solution of equation (2) with w(z) = k2 z 2/m there corresponds the third solution in Item 2◦ . © 2004 by Chapman & Hall/CRC . 1 −1 6 . see the book by Zel’dovich and Raiser (1966). b1 = ◦ (1 − Z)1/m 0 z . m) P (1. For details. The books by Polyanin and Zaitsev (1995. C1 . u (1) b= 2λ . which describes the propagation of a thermal wave coming from a plane source. To C = 0 in (3) there corresponds the fourth solution in Item 2◦ . mP (1. see also the book by Lykov (1967). = x + C2 . z0 2 z0 = P (1 − Z. Fujita (1952) obtained the general solution of equation (2) for m = −1 and m = −2. w = 0 at z = ∞ the solution of equation (2) is localized and has the structure w= where Z= b0 = 1. . Self-similar solution: w = w(z). for 1 ≤ Z < ∞. k=0 − 1 [m(m 2 + 1)] . (2) Solution of this sort usually describe situations where the unknown function assumes constant values at the initial and boundary conditions. see Samarskii and Sobol’ (1963). the function F = F (ξ) is determined by the ﬁrst-order ordinary differential equation a(m + 2)F m Fξ + ξF = C. one obtains ϕξ = αϕ−1/m − βξ. t) = (λt + A)−1/m f (x). Self-similar solution: ξ = xt− m+2 1 (0 ≤ x < ∞). where the function f = f (x) is deﬁned implicitly by f m df C1 − 5◦ .

t) = At t 0 where A= m 2(m + 2) 1/m 1/m for |x| ≤ η0 t1/(m+2) . (5) where A1 = −(mβ + 1)/(2a) and A2 = β(m + 1)/a.1. b are any. and. A. and A. Kurdyumov. hence. therefore. its order can be reduced (and then it can be transformed to an Abel equation of the second kind). β is any. P. P. −2. −1 can be found in Polyanin and Zaitsev (2003). 10◦ +. with (5). up to notation. the function g = g(ζ) is determined by the ordinary differential equation Gζζ = A1 ζG− m+1 Gζ + A2 G m+1 . . Samarskii. A. and the condition of conservation of energy ∞ −∞ w ¤v w(x. u = x + b ln(t + A). with Γ(z) being the gamma function. Generalized self-similar solution: w = e−2λt ϕ(u). where δ(x) is the Dirac delta function. and. Exact analytical solutions of equation (5) for various values of m can be found in Polyanin and Zaitsev (2003). (6) where the function ϕ = ϕ(u) is determined by the ordinary differential equation a(ϕm ϕu )u = λmuϕu − 2λϕ. (7) where the function ψ = ψ(u) is determined by the autonomous ordinary differential equation a(ψ m ψu )u = bψu − ψ/m. 8◦ . Taking into account the identity b du b −m d ψ p . A. The substitution Φ = ϕm+1 brings (6) to an equation that coincides. we arrive at an Abel equation of the second kind: a dψ ppψ = p − sψ m+1 . Self-similar solution of a more general form: w = tβ g(ζ). its order can be reduced (and then it can be transformed to an Abel equation of the second kind). This equation is homogeneous. V. t) dx = E0 > 0. u = xeλmt . Galaktionov. Mikhailov (1995). − 2 . for |x| > η0 t1/(m+2) . m ζ = xt− mβ+1 2 . © 2004 by Chapman & Hall/CRC . Solution: w = (t + A)−1/m ψ(u). S. EQUATIONS WITH POWER-LAW NONLINEARITIES 31 7◦ . Here. d a = Introduce the new dependent variable p(ψ) = ψ m ψu . λ is any. s = a/(mb2). 9◦ . Unsteady point source solution with a = 1: −1/(m+2) 2 x2 η0 − 2/(m+2) w(x. η0 = Γ(1/m + 3/2) √ E0 A π Γ(1/m + 1) m/(m+2) . 0) = E0 δ(x). This equation is homogeneous. Reference: A. 3 The general solutions of this equation with m = −3. Γ(z) = ∞ 0 e−ξ ξ z−1 dξ.1. The above solution satisﬁes the initial condition w(x. 1 G = g m+1 .

A. Ψ= ∂y ∂y ∂y u ∂y u It follows that any solution u = u(x. C2 . a au2 − C1 u + b x + C1 t + C2 4◦ . Suppose w(x. V. C2 . t) dy + a t wm (x. Ovsiannikov (1959. ∂w =a ∂ ∂w where C1 . w(x. Kurdyumov.7: L. Generalized separable solutions linear and quadratic in x: 2 w(x. S. Then the functions w −1 w1 = C1 w( C1 x + C2 . τ ) ∂x dτ . Equations of the Form ∂w = a ∂ w m ∂w ∂t ∂x ∂x 1.11 with m = 1 and k = 0. 3 (x + C2 )2 + C3 |t + C1 |−1/3 + b(t + C1 ). R. Samarskii. The transformation t = t − t0 . bu ∂y w(x.1. see equation 1. Dorodnitsyn and S. −1 2 ∂u . 6a(t + C1 ) 4 where C1 .1. The ﬁrst solution is degenerate. x= PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE x x0 w(y. and C3 are arbitrary constants. 2◦ . where Φ = . The transformation x=− leads to the equation ∂Ψ ∂ 1 ∂u 1 ∂ Φ = 0. Ibragimov (1994). H. ∂t ∂x ∂x 1◦ .1.11.32 11◦ . P. Mikhailov (1995). are also solutions of the equation. C2 . and C3 are arbitrary constants. + bw k Equations of this form admit traveling-wave solutions w = w(kx + λt). ∂t ∂y 2 (2) © 2004 by Chapman & Hall/CRC . 1. A. t) is a solution of this equation. Traveling-wave solution in implicit form: w u du = − ln | x + C1 t + C2 | + C3 . t) of the linear heat equation ∂u ∂ 2 u − =0 ∂t ∂y 2 generates a solution (1) of the original nonlinear equation. + b. ∂t ∂x ∂x 1◦ . Galaktionov. C1 t + C3 ). t) of the original equation to a solution w(x.10. A. where C1 . t) = 1 w(x. Svirshchevskii (1983). 2. 1962. u = . τ ) t0 ∂w (x. P. A. Suppose w(x. Then the functions −1 w1 = C1 w( C1 x + C2 . t) = − 3◦ . ∂ ∂w ∂w = w–2 + b. and C3 are arbitrary constants. x=x0 w(x. V. are also solutions of the equation. 2◦ .11. 1982). Reference: V. C1 t + C3 ). t) = − b ∂ 2 ∂y 1 ∂u u ∂y −1 (1) ∂u ∂ 2 u − . For other solutions. and A. N. t) takes a nonzero solution w(x. t) of a similar equation ∂ ∂w ∂w =a w−m−2 . t) is a solution of this equation. ∂x ∂x ∂t y ¤x y ¤x References for equation 1. t) = C1 x + (aC1 + b)t + C2 .

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

33

3.

**∂ ∂w =a w–2 – bw3 . ∂t ∂x ∂x Functional separable solutions: ∂w
**

w(x, t) =

C1 (x + C2 )2 + C3 exp(2aC1 t) −

b aC1

−1/2

,

**where C1 , C2 , and C3 are arbitrary constants. 4. + bw–1/3 . ∂t ∂x ∂x 1◦ . For ab > 0, the transformation =a
**

∂w

∂

w–4/3

∂w

w(x, t) = exp( 3λx)z(ξ, t),

ξ=

1 exp( 2λx), 2λ

λ=

b 3a

1/2

,

**leads to a simpler equation of the form 1.1.10.4: ∂z ∂ ∂z z −4/3 . =a ∂t ∂ξ ∂ξ
**

¤

(1)

References: N. H. Ibragimov (1994), A. A. Samarskii, V. A. Galaktionov, S. P. Kurdyumov, and A. P. Mikhailov (1995).

2◦ . For ab < 0, the transformation z(ξ, t) , cos3 (λx) also leads to equation 1.1.10.4. w(x, t) = 3◦ . Multiplicative separable solution: w(x, t) = (t + C)3/4 u(x), where C is an arbitrary constant, and the function u = u(x) is determined by the autonomous ordinary differential equation a(u−4/3 ux )x + bu−1/3 − 3 u = 0. 4 4◦ . See also equation 1.1.12.6 with b = c = 0. 5. + bw5/2 . ∂t ∂x ∂x 1◦ . Functional separable solution: =a w(x, t) = a2/3 3Ax3 + f2 (t)x2 + f1 (t)x + f0 (t) Here, f2 (t) = 3 f0 (t) = ϕ(t) dt + 3B, f1 (t) =

3 −2/3

ξ=

1 tan(λx), λ

λ=

−

b 3a

1/2

,

∂w

∂

w–3/2

∂w

.

1 A

ϕ(t) dt + B

2

+

1 ϕ(t), 2A 1 ϕ (t), 36A2 t

1 1 ϕ(t) dt + B + ϕ(t) 9A2 6A2 where the function ϕ(t) is deﬁned implicitly by

ϕ(t) dt + B +

(C1 − 108A2 abϕ − 8ϕ3 )−1/2 dϕ = t + C2 , and A, B, C1 , and C2 are arbitrary constants. 2◦ . For other solutions, see equations 1.1.11.9 and 1.1.11.11 with m = −3/2. 3◦ . The substitution w = u−2/3 leads to an equation of the form 1.1.9.7: ∂2u 2 ∂u ∂u = au 2 − a ∂t ∂x 3 ∂x

2

−

3 b. 2

© 2004 by Chapman & Hall/CRC

34 6. ∂w

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

**∂ ∂w =a w–4/3 + cw7/3 . ∂t ∂x ∂x This is a special case of equation 1.1.12.5 with b = 0. See also equation 1.1.11.11 with m = −4/3 and k = 7/3. 7. ∂w ∂t ∂x 1◦ . Solutions: =a ∂ wm ∂w ∂x + bw.
**

1

w(x, t) = ebt (Ax + B) m+1 , w(x, t) = e w(x, t) = e

bt

λ2 bmt λm x+ e +A a ab

1 m

,

bt

bm2 (x − A)2 2a(m + 2)(B − ebmt )

1 m

,

1 m

**2bmt w(x, t) = A exp m+2 w(x, t) = e 2◦ . By the transformation
**

bt

bm2 (x + B)2 − 2a(m + 2)

,

1 m

A|e

bmt

m + B|− m+2 −

(x + C)2 bm2 2a(m + 2) ebmt + B

,

where A, B, C, and λ are arbitrary constants. 1 bmt e + const bm the original equation can be reduced to an equation of the form 1.1.10.7: ∂ ∂v ∂v =a vm . ∂τ ∂x ∂x 3◦ . See also equation 1.1.11.11 with k = 1. w(x, t) = ebt v(x, τ ), τ=

¤

Reference: L. K. Martinson and K. B. Pavlov (1972).

L , 2 (1) L for |x| > , 2 1/2 where L = 2π(m + 1) /m. Solution (1) describes a blow-up regime that exists on a limited time interval t [0, t0 ). The solution is localized in the interval |x| < L/2. for |x| ≤

References: N. V. Zmitrenko, S. P. Kurdyumov, A. P. Mikhailov and A. A. Samarskii (1976), A. A. Samarskii, V. A. Galaktionov, S. P. Kurdyumov, and A. P. Mikhailov (1995).

**+ bwm+1 . ∂t ∂x ∂x 1◦ . Multiplicative separable solution (a = b = 1, m > 0): 2(m + 1) cos2 (πx/L) m(m + 2) (t0 − t) w(x, t) = 0 8. wm
**

∂w

=a

∂

∂w

1/m

© 2004 by Chapman & Hall/CRC

¤

**2◦ . Multiplicative separable solution: w(x, t) = B= λ2 (m + 1)2 , 4b2 A(m + 2)2 Aeµx + Be−µx + D mλt + C D=− λ(m + 1) , b(m + 2)
**

1/m

, − b , a(m + 1)

µ=m

where A, C, and λ are arbitrary constants, ab(m + 1) < 0.

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

35

3◦ . Multiplicative separable solution (C and λ are arbitrary constants): w(x, t) = (mλt + C)−1/m ϕ(x), where the function ϕ = ϕ(x) is determined by the ordinary differential equation a(ϕm ϕx )x + bϕm+1 + λϕ = 0. Equation (2) has the following solution in implicit form:

(2)

ϕm A −

b 2λ ϕm+2 − ϕ2m+2 a(m + 2) a(m + 1)

−1/2

dϕ = x + B,

**where A and B are arbitrary constants. 4◦ . Functional separable solution [it is assumed that ab(m + 1) < 0]: w(x, t) = f (t) + g(t)eλx
**

1/m

where the functions f = f (t) and g = g(t) are determined by the autonomous system of ordinary differential equations bm(m + 2) f g. ft = bmf 2 , gt = m+1 Integrating yields m+2 f (t) = (C − bmt)−1 , g(t) = C (C − bmt)− m+1 ,

1 2 1

where C1 and C2 are arbitrary constants. 5◦ . Functional separable solution (A and B are arbitrary constants): w(x, t) = f (t) + g(t)(Aeλx + Be−λx )

1/m

,

where the functions f = f (t) and g = g(t) are determined by the autonomous system of ordinary differential equations ft = bmf 2 + 4bmAB 2 g , m+1 gt = bm(m + 2) f g. m+1 (4)

On eliminating t from this system, one obtains a homogeneous ﬁrst-order equation: fg = 4AB g m+1 f + . m+2 g m+2 f (5)

The substitution ζ = f /g leads to a separable equation. Integrating yields the solution of equation (5) in the form 1 2 C1 is any. f = g 4AB + C1 g − m+2 2 , Substituting this expression into the second equation of system (4), one obtains a separable equation for g = g(t). 6◦ . The functional separable solutions w(x, t) = f (t) + g(t) cosh(λx) w(x, t) = f (t) + g(t) sinh(λx) are special cases of formula (3) with A = 1 , B = 2

© 2004 by Chapman & Hall/CRC

1 2 1/m 1/m

1 1 and A = 2 , B = − 2 , respectively.

,

λ= m

−b , a(m + 1)

λ=m

−b , a(m + 1)

(3)

,

36

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

**7◦ . Functional separable solution [it is assumed that ab(m + 1) > 0]: w(x, t) = f (t) + g(t) cos(λx + C)
**

1/m

,

λ=m

b , a(m + 1)

where the functions f = f (t) and g = g(t) are determined by the autonomous system of ordinary differential equations bm(m + 2) bm 2 g , gt = f g, ft = bmf 2 + m+1 m+1 which coincides with system (4) for AB = 1 . 4

References for equation 1.1.11.8: M. Bertsch, R. Kersner, and L. A. Peletier (1985), V. A. Galaktionov and S. A. Posashkov (1989), V. F. Zaitsev and A. D. Polyanin (1996).

w(x, t) =

where C1 and C2 are arbitrary constants. 11. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + bwk .

This is a special case of equation 1.6.15.2 with f (w) = aw m and g(w) = bwk . For b = 0, see Subsection 1.1.10. 1◦ . Suppose w(x, t) is a solution of the equation in question. Then the functions

k−m−1 2k−2 2 x + C2 , C1 t + C3 ), w1 = C1 w( C1

where C1 , C2 , and C3 are arbitrary constants, are also solutions of the equation. 2◦ . A space-homogeneous solution and a stationary solution are given by (the latter is written out in implicit form): w(t) = wm A − (1 − k)bt + C Cebt

1 1−k

dw = x + B,

**where A, B, and C are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

¤ ¤

**+ bw1–m . ∂t ∂x ∂x This is a special case of equation 1.1.11.11 with k = 1 − m. Functional separable solution: 9. =a wm
**

m bm2 1 (x + A)2 + B|F |− m+2 − F F 4a(m + 1) 2a(m + 2) t, F = F (t) = C − m where A, B, and C are arbitrary constants.

∂w

∂

∂w

w(x, t) =

1/m

,

Reference: R. Kersner (1978).

**+ bw1–n . ∂t ∂x ∂x This is a special case of equation 1.1.11.11 with m = 2n and k = 1 − n. Generalized traveling-wave solution: 10. =a bn2 x + C1 √ − (C2 − kt) C2 − kt 3a(n + 1)
**

1/n

∂w

∂

w2n

∂w

,

k=

2a(n + 1) , n

if k ≠ 1, if k = 1,

−1/2

2b wm+k+1 a(m + k + 1)

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

37

3◦ . Traveling-wave solutions: where the function w(z) is determined by the autonomous ordinary differential equation a(wm wz )z − λwz + bwk = 0. The substitution u(w) = brings (1) to the Abel equation a m w wz λ (2) (1)

w = w(z),

z = x + λt,

uuw − u = −abλ−2 wm+k .

The book by Polyanin and Zaitsev (2003) presents exact solutions of equation (2) with m + k = 1 −2, −1, − 2 , 0, 1. 4◦ . Self-similar solution for k ≠ 1: w = t 1−k u(ξ),

1 k−m−1 2(1−k)

ξ = xt

,

**where the function u(ξ) is determined by the ordinary differential equation a(um uξ )ξ +
**

¤

1 m−k+1 ξuξ + buk − u = 0. 2(1 − k) 1−k

Reference: V. A. Dorodnitsyn (1982).

**1.1.12. Equations of the Form ∂w = a ∂ w m ∂w + bw + c w k1 + c w k2 + c w k3 1 2 3 ∂t ∂x ∂x
**

Equations of this form admit traveling-wave solutions w = w(kx + λt). 1. ∂ ∂w ∂w =a w ∂t ∂x ∂x + bw + c.

**For c = 0, see equation 1.1.11.7. Generalized separable solutions linear and quadratic in x:
**

2 aC1 2bt c e − , b b cebt ebt bebt (x + C2 )2 + C3 1/3 + 1/3 w(x, t) = ϕ(t) ϕ (t) ϕ (t)

w(x, t) = C1 xebt + C2 ebt +

e−bt ϕ1/3 (t)dt,

ϕ(t) = C1 − 6aebt ,

where C1 , C2 , and C3 are arbitrary constants. The ﬁrst solution is degenerate. 2. ∂w ∂t =a ∂ ∂x w–2 ∂w ∂x – bw – cw3 .

**Functional separable solutions for b ≠ 0: w(x, t) =
**

bC1 e2bt (x + C2 )2 + C3 F (t) + 2cF (t)

dt F (t)

−1/2

,

F (t) = exp aC1 e2bt + 2bt ,

**where C1 , C2 , and C3 are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

38 3. ∂w

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

**∂ ∂w =a w–4/3 + bw–1/3 + cw. ∂t ∂x ∂x 1◦ . Multiplicative separable solutions: w(x, t) = ect Aekx + Be−kx
**

−3 −3

if b/(3a) = k 2 > 0, if b/(3a) = −k 2 < 0,

w(x, t) = ect A cos(kx) + B sin(kx) where A and B are arbitrary constants. 2◦ . The transformation w = ect u(x, τ ), τ =−

leads to a simpler equation of the form 1.1.11.4: ∂ ∂u ∂u =a u−4/3 ∂τ ∂x ∂x 4. ∂ ∂w ∂w =a w–4/3 ∂t ∂x ∂x Functional separable solution: + bw + cw5/3 .

3 − 4 ct e 3 + const 4c + bu−1/3 .

w(x, t) = ϕ(t)x2 + ψ(t)x + χ(t) , where the functions ϕ(t), ψ(t), and χ(t) are determined by the system of ﬁrst-order ordinary differential equations 2 1 ϕt = 2 aϕ(4ϕχ − ψ 2 ) − 3 bϕ,

2 χt = 1 aχ(4ϕχ − ψ 2 ) − 2 bχ − 3 c. 2 3 It follows from the ﬁrst two equations that ϕ = Cψ, where C is an arbitrary constant. 2 ψt = 1 aψ(4ϕχ − ψ 2 ) − 3 bψ, 2

−3/2

5.

∂w

∂t ∂x ∂x Functional separable solution:

=a

∂

w–4/3

∂w

+ bw + cw7/3 .

−3/4

w(x, t) = ϕ4 (t)x4 + ϕ3 (t)x3 + ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) , where the functions ϕk = ϕk (t) are determined by the system of ordinary differential equations 4 4 3 ϕ0 = − 4 aϕ2 + 2aϕ0 ϕ2 − 3 bϕ0 − 3 c, 1 ϕ1 = −aϕ1 ϕ2 + 6aϕ0 ϕ3 − 4 bϕ1 , 3

4 ϕ2 = −aϕ2 + 3 aϕ1 ϕ3 + 12aϕ0 ϕ4 − 3 bϕ2 , 2 2

**4 ϕ4 = − 3 aϕ2 + 2aϕ2 ϕ4 − 3 bϕ4 . 3 4 The prime denotes a derivative with respect to t.
**

¤

ϕ3 = −aϕ2 ϕ3 + 6aϕ1 ϕ4 − 4 bϕ3 , 3

Reference: V. A. Galaktionov (1995).

– aw–1/3 + bw7/3 + cw. ∂t ∂x ∂x The substitution u = w −4/3 leads to an equation with quadratic nonlinearity: 6. = ∂ 2 u 3 ∂u 4 ∂u =u 2 − + au2 − cu − b . ∂t ∂x 4 ∂x 3 1◦ . For a = 1, there is a solution of the form u = ϕ1 (t) + ϕ2 (t) cos(kx) + ϕ3 (t) sin(kx) + ϕ4 (t) cos(2kx) + ϕ5 (t) sin(2kx), k = 2 × 3−1/2 , where the functions ϕn = ϕn (t) are determined by the system of ﬁrst-order ordinary differential equations (not written out here).

© 2004 by Chapman & Hall/CRC

2

∂w

∂

w–4/3

∂w

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

39

**2◦ . For a = −1, there is a solution of the form u = ϕ1 (t) + ϕ2 (t) cosh(kx) + ϕ3 (t) sinh(kx) + ϕ4 (t) cosh(2kx) + ϕ5 (t) sinh(2kx),
**

¤

k = 2 × 3−1/2 .

Reference: V. A. Galaktionov (1995).

**+ bwm+1 + cw. ∂t ∂x ∂x 1◦ . Multiplicative separable solutions: 7. =a wm w(x, t) = ect A cos(kx) + B sin(kx)
**

1 m+1 1 m+1

∂w

∂

∂w

if b(m + 1)/a = k 2 > 0, if b(m + 1)/a = −k 2 < 0,

w(x, t) = ect A exp(kx) + B exp(−kx) where A and B are arbitrary constants. 2◦ . The transformation w = ect u(x, τ ), τ= leads to a simpler equation of the form 1.1.11.8:

1 cmt e + const cm + bum+1 .

∂ ∂u ∂u =a um ∂τ ∂x ∂x

Special case. Multiplicative separable solution for m = −1: w = A exp ct − where A and B are arbitrary constants.

b 2 x + Bx , 2a

8.

∂w

∂t ∂x ∂x Functional separable solution:

=a

∂

wm

∂w

+ b + cw–m .

1 m+1

b(m + 1) 2 x + C1 x + C2 w = c(m + 1)t − 2a where C1 and C2 are arbitrary constants. 9. + bw + cw1–m . ∂t ∂x ∂x 1◦ . Generalized traveling-wave solution: =a wm w(x, t) = C1 ebmt x + ∂w ∂ ∂w

,

c a 2bmt e + C2 ebmt − bm2 b

1/m

,

**where C1 and C2 are arbitrary constants. 2◦ . For a more complicated solution, see 1.6.13.4 with f (t) = b and g(t) = c.
**

¤

Reference: V. A. Galaktionov and S. A. Posashkov (1989).

+ bw1+m + cw + w1–m . ∂t ∂x ∂x This is a special case of equation 1.6.13.5 with f (t) = c and g(t) = s. The substitution u = w m leads to an equation of the form 1.1.9.9: ∂2u a ∂u = au 2 + ∂t ∂x m

¤

∂u ∂x

Reference: V. A. Galaktionov and S. A. Posashkov (1989).

© 2004 by Chapman & Hall/CRC

10.

∂w

=a

∂

wm

∂w

2

+ bmu2 + cmu + sm.

40 11. ∂w

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

∂ ∂w =a w2n + bw + cw1–n . ∂t ∂x ∂x Generalized traveling-wave solutions: w(x, t) = ϕ(t)( x + C1 ) + cnϕ(t) where C1 and C2 are arbitrary constants. 12. ∂ ∂w =a wn + bw + ck + bkwn+1 + cw–n . ∂t ∂x ∂x Functional separable solutions: ∂w w(x, t) = w(x, t) = √ √ exp b(n + 1)t C1 cos x λ + C2 sin x λ exp b(n + 1)t C1 cosh x c − b

1 n+1

d

dt ϕ(t)

1/n

,

ϕ(t) = C2 e−2bnt −

a(n + 1) bn2

−1/2

,

if λ > 0,

1 n+1

|λ| + C2 sinh x |λ| bk (n + 1). a

c − b

if λ < 0,

**where C1 and C2 are arbitrary constants and λ =
**

f ¤e

Reference: V. A. Galaktionov (1994).

**1.1.13. Equations of the Form ∂w = ∂ f (w) ∂w + g(w) ∂t ∂x ∂x
**

Equations of this form admit traveling-wave solutions w = w(kx + λt). ∂w ∂t = ∂ ∂x (aw2 + bw) ∂w ∂x . 1.

This is a special case of equation 1.6.15.1 with f (w) = aw 2 + bw. Solutions: b 2 2C1 x + 2aC1 t + C2 − , w(x, t) = a b x + C1 , − w(x, t) = √ 2a C2 − 4at where C1 and C2 are arbitrary constants. The ﬁrst solution is of the traveling-wave type and the second one is self-similar. 2. ∂w ∂t = ∂ ∂x w2 a + b2 ∂w ∂x .

d d

This is a special case of equation 1.6.15.1 with f (w) = a(w 2 + b2 )−1 . 1◦ . Solutions (A and B are arbitrary constants): w(x) = b tan(Ax + B), w(x, t) = bx A − 2ab−2 t − x2 2◦ . Traveling-wave solution in implicit form: λ(kx + λt) + B = A2 1 A w ak 2 ln |w + A| − ln(w2 + b2 ) + arctan , + b2 2 b b

d

−1/2 2

,

−1/2

w(x, t) = Ab exp(ab t − x) 1 − A exp 2(ab−2 t − x)

−2

.

**where A, B, k, and λ are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

41

3◦ . The substitution

bu w= √ 1 − u2 a ∂2u ∂u ∂u = 2 (1 − u2 ) 2 + u ∂t b ∂x ∂x

2

(1)

leads to the equation

,

(2)

which is a special case of 8.1.2.12 with F (t, ξ, η) = ab−2 (ξ − η). Equation (2) has multiplicative separable solutions 2aλ2 Aeλx + Be−λx , k= 2 ; u= √ b 4AB + Ce−kt (3) 2aλ2 A sin(λx) + B cos(λx) , k= 2 , u= √ b A2 + B 2 + Cekt where A, B, C, and λ are arbitrary constants. Formulas (1) and (3) provide two solutions of the original equation.

h ¤g h ¤g

**Reference: P. W. Doyle and P. J. Vassiliou (1998); see also Example 10 in Subsection S.5.3.
**

i

w = b tan

z = x2 cos−2

where C is an arbitrary constant and the function ψ = ψ(z) is determined by the ordinary differential equation ψ 1 . ψz = (1 + ψ 2 ) 1 − 2 z Here the function z = z(x, t) is deﬁned implicitly. 5◦ . Solution: w = b tan ϕ(z) + arctan ψ(z) + z= at C ln 2 , 2 b

i

C at b 2 x2 cos−2 ϕ(z) + ln 2 , at 2 b

where C is an arbitrary constant and the functions ϕ(z) and ψ(z) are determined by the system of ordinary differential equations ϕz = ψ , 2z ψz = 1 C ψ ψ (1 + ψ 2 ) − − . 2 2 2 z

**Here the function z = z(x, t) is deﬁned implicitly.
**

References: I. Sh. Akhatov, R. K. Gazizov, and N. H. Ibragimov (1989), N. H. Ibragimov (1994).

3.

∂w ∂t

=

∂ ∂x

(aw2n + bwn )

∂w ∂x

.

This is a special case of equation 1.6.15.1 with f (w) = aw 2n + bwn . 1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function

2 w1 = w(C1 x + C2 , C1 t + C3 ),

**where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation.
**

© 2004 by Chapman & Hall/CRC

i

i

i

4◦ . Solution:

1 z 2

a t+C , b2 1 z − arctan ψ(z) 2 a t+C , b2

42

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

**2◦ . Traveling-wave solution: w(x, t) =
**

2 2C1 nx + 2aC1 nt + C2 −

b a

1/n

**where C1 and C2 are arbitrary constants. 3◦ . Self-similar solution: w(x, t) = 4. ∂w = ∂ b x + C1 √ − C2 − kt a(n + 1)
**

j

**∂t ∂x ∂x Generalized traveling-wave solution: w(x, t) =
**

j

(aw2n + bwn )

∂w

b cn2 x + C1 √ (C2 − kt) − − a(n + 1) C2 − kt 3a(n + 1)

**where C1 and C2 are arbitrary constants. ∂t ∂x ∂x Generalized traveling-wave solutions:
**

j k

5.

∂w

=

∂

(aw2n + bwn )

∂w

w(x, t) = ϕ(t)( x + C1 ) + ϕ(t) = C2 e−2cnt −

where C1 and C2 are arbitrary constants.

**1.1.14. Equations of the Form ∂w = ∂ f (w) ∂w + g x, t, w, ∂w ∂t ∂x ∂x ∂x
**

1. + bxm w–1/3 . ∂t ∂x ∂x For m = 0, see equation 1.1.11.4. For m ≠ 0, the original equation can be reduced to a simpler equation 1.1.10.4 that corresponds to the case b = 0 [see equation 1.6.13.1 with f (x) = bx m ]. =a 2. ∂w =a ∂ w–2 ∂w +b ∂w ∂w ∂ w–4/3 ∂w

+ cw. ∂t ∂x ∂x ∂x This is a special case of equation 1.6.13.8 with m = −2, f (t) = b, and g(t) = c. The transformation (A and B are arbitrary constants) w(x, t) = ect u(z, τ ), leads to an equation of the form 1.1.10.3: ∂ ∂u ∂u =a u−2 . ∂τ ∂z ∂z

m ¤l

Reference: V. A. Dorodnitsyn and S. R. Svirshchevskii (1983); the case b = 0 was treated.

© 2004 by Chapman & Hall/CRC

j

,

1/n

,

k=

2a(n + 1) . n

+ cw1–n .

1/n

,

k=

2a(n + 1) , n

+ cw + w1–n .

b ϕ(t) n

−1/2

ϕ(t) dt + snϕ(t) ,

dt ϕ(t)

1/n

,

a(n + 1) cn2

z = x + bt + A,

τ =B−

1 −2ct e 2c

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

43

u(z, t) = w(x, t) + b, z = x + ct leads to an equation of the form 1.1.10.3: ∂ ∂u ∂u =a u−2 . ∂t ∂z ∂z 4. . ∂t ∂x ∂x ∂x This is a special case of equation 1.1.14.16 with n = 1. Degenerate solution linear in x: ax + b ln |t + C1 | + C2 , w(x, t) = a2 (t + C1 ) where C1 and C2 are arbitrary constants. w ∂w ∂ ∂w + aw =b w2 . ∂t ∂x ∂x ∂x This is a special case of equation 1.1.14.16 with n = 2. ∂w 2b ∂w + aw ∂w =b ∂ ∂w

∂ a ∂w = ∂t ∂x (w + b)2 ∂x The transformation 3. ∂w

+c

∂w ∂x

.

5.

1◦ . Traveling-wave solution in implicit form: w2 dw = x + λt + C2 , aw2 + 2λw + C1 where C1 , C2 , and λ are arbitrary constants. 2◦ . Degenerate solution linear in x: w(x, t) = (x + C1 )f (t). Here, C1 is an arbitrary constant, and the function f = f (t) is determined by the ordinary differential equation ft + af 2 = 2bf 3 , whose solution can be represented in implicit form: 2b 2bf − a 1 + 2 ln = t + C2 . af a f 6. ∂w ∂ ∂w ∂w + aw = (bw2 + cw) . ∂t ∂x ∂x ∂x 1◦ . Degenerate solution linear in x: w(x, t) = f (t)x + g(t), where the functions f = f (t) and g = g(t) are determined by the system of ordinary differential equations ft + af 2 = 2bf 3 , gt + af g = 2bf 2 g + cf 2 . The solution of the ﬁrst equation can be found in 1.1.14.5, Item 2 ◦ . The second equation is easy to integrate, since it is linear in g. 2◦ . Traveling-wave solution in implicit form: bw2 + cw dw = x + λt + C2 , + 2λw + C1 where C1 , C2 , and λ are arbitrary constants. 2 aw2

© 2004 by Chapman & Hall/CRC

44 7. ∂w

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

∂ ∂w =a wm + btn w. ∂t ∂x ∂x This is a special case of equation 1.6.13.2 with f (t) = btn . 8. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + beλt w.

This is a special case of equation 1.6.13.2 with f (t) = beλt . 9. + btn w1–m . ∂t ∂x ∂x This is a special case of equation 1.6.13.3 with f (t) = btn . =a wm 10. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + beλt w1–m . ∂w ∂ ∂w

This is a special case of equation 1.6.13.3 with f (t) = beλt . 11. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + btn w + ctk w1–m .

This is a special case of equation 1.6.13.4 with f (t) = btn and g(t) = ctk . 12. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + beλt w + ceµt w1–m .

**This is a special case of equation 1.6.13.4 with f (t) = beλt and g(t) = ceµt .
**

n

13.

∂ ∂w ∂w =a wm ∂t ∂x ∂x

+ bw1+m + ctn w + tk w1–m .

This is a special case of equation 1.6.13.5 with f (t) = ctn and g(t) = stk . 14. ∂ ∂w =a wm + bxn w1+m . ∂t ∂x ∂x This is a special case of equation 1.6.13.6 with f (x) = bxn . ∂w 15. ∂w

**∂ ∂w ∂w =a wn – bx . ∂t ∂x ∂x ∂x 1◦ . Suppose w(x, t) is a solution of this equation. Then the functions
**

n −2 w1 = C1 w( C1 x + C2 ebt , t + C3 ),

where C1 , C2 , and C3 are arbitrary constants, are also solutions of the equation. 2◦ . Generalized traveling-wave solutions:

o

w(x, t) =

x + C1 ebt

where C1 and C2 are arbitrary constants. 3◦ . Generalized traveling-wave solutions: where C is an arbitrary constant and the function w(z) is determined by the ordinary differential equation a(wn wz )z − bzwz = 0.

© 2004 by Chapman & Hall/CRC

o

w = w(z),

o

2/n

C2 e2bt +

a(n + 2) bn

−1/n

,

z = x + Cebt ,

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

45

**∂w ∂ ∂w + aw =b wn . ∂t ∂x ∂x ∂x 1◦ . Suppose w(x, t) is a solution of this equation. Then the function 16. ∂w
**

1−n 2−n w1 = C1 w(C1 x + C2 , C1 t + C3 ),

where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. 2◦ . Traveling-wave solution in implicit form: 2b aw2 wn dw = x + λt + C2 , + 2λw + C1

where C1 , C2 , and λ are arbitrary constants. 3◦ . Self-similar solution for n ≠ 2: w(x, t) = u(z)t1/(n−2) , z = xt−(n−1)/(n−2) , 1 n−1 z uz − u = 0. n−2 n−2 where the function u = u(z) is determined by the ordinary differential equation bun uzz + 2bnun−1 (uz )2 − au − 17. ∂w ∂ ∂w =a wn + (bwn + c) . ∂t ∂x ∂x ∂x Generalized traveling-wave solution: ∂w w(x, t) = a ln |t + C1 | c C2 − x + − b(t + C1 ) b2 n(t + C1 ) b ∂2w

2 1/n

,

**where C1 and C2 are arbitrary constants. 18. . ∂t ∂x ∂x ∂x2 ∂x 1◦ . Suppose w(x, t) is a solution of this equation. Then the function wn
**

n 2 w1 = C1 w(C2 x + C3 , C1 C2 t + C4 ),

∂w

=

∂

∂w

+ awn

+ bwn–1

∂w

**where C1 , . . . , C4 are arbitrary constants, is also a solution of the equation. 2◦ . Functional separable solution for b = 1 n(a − 2) − a − 1: 3
**

3 1/n

w(x, t) = n

k=0

ϕk (t)xk

.

Here, ϕ3 (t) = A, ϕ0 (t) = ϕ2 (t) = ψ(t) dt + B,

3

ϕ1 (t) =

1 3A

ψ(t) dt + B ψ(t) dt + B +

2

+

1 ψ(t), 2Aβn

1 27A2

ψ(t) dt + B

+

1 ψ(t) 6A2 βn

−1/2

1 ψ (t), 12A2 β 2 n2 t

**where the function ψ = ψ(t) is deﬁned implicitly by C1 − 8 βnψ 3 3
**

q ¤p

dψ = C2 + t,

**A, B, C1 , and C2 are arbitrary constants, β = a + 1; A ≠ 0, n ≠ 0, a > −1.
**

Reference: G. A. Rudykh and E. I. Semenov (1998).

© 2004 by Chapman & Hall/CRC

46

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

1 3◦ . Functional separable solution for b = 4 n(a − 3) − a − 1: 4

w(x, t) = n

k=0

1/n

ϕk (t)xk

.

Here, the functions ϕk = ϕk (t) are determined by the system of ordinary differential equations ϕ0 = − 3 βϕ2 + 2βϕ2 ϕ0 , 1 4 ϕ1 = −βϕ1 ϕ2 + 6βϕ3 ϕ0 ,

3 ϕ2 = −βϕ2 + 2 βϕ1 ϕ3 + 12βϕ4 ϕ0 , 2

ϕ3 = −βϕ2 ϕ3 + 6βϕ1 ϕ4 , ϕ4 = − 3 βϕ2 + 2βϕ2 ϕ4 , 3 4

**where β = n(a + 1); the prime denotes a derivative with respect to t.
**

s ¤r

**Reference: G. A. Rudykh and E. I. Semenov (1998).
**

◦

**4 . There are exact solutions of the following forms: w(x, t) = F (z), z = Ax + Bt; ξ = xt−
**

βn+1 2 ;

w(x, t) = (At + B)−1/n G(x); w(x, t) = tβ H(ξ), w(x, t) = e−2t U (η), η = xent ; ζ = x + C ln(At + B),

w(x, t) = (At + B)−1/n V (ζ),

where A, B, C, and β are arbitrary constants. The ﬁrst solution is of the traveling-wave type, the second is a solution in multiplicative separable form, and the third is self-similar.

t

∂t ∂x ∂x Generalized traveling-wave solution:

19.

∂w

=

∂

aw2n + bwn

∂w

+ (cwn + )

∂w ∂x

.

w(x, t) = ϕ(t)x + (st + C1 )ϕ(t) +

b ϕ(t) n

1/n

ϕ(t) dt

,

where C1 is an arbitrary constant and the function ϕ(t) is determined by the ﬁrst-order separable ordinary differential equation a(n + 1) 3 ϕ + cϕ2 . ϕt = n

1.1.15. Other Equations

1. . ∂t ∂x2 This is a special case of equation 1.6.16.3 with f (w) = aw 2 + bw4 . 1◦ . Self-similar solutions:

u

∂w

= (aw2 + bw4 )

∂2w

**w(x, t) = where C1 and C2 are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

a (x + C1 )2 − 2a(t + C2 ) b

1/2

,

1.1. EQUATIONS WITH POWER-LAW NONLINEARITIES

47

**2◦ . Functional separable solutions:
**

v

w(x, t) =

ϕ(t)(x2 + C1 x + C2 ) −

a b

1/2

,

where C1 and C2 are arbitrary constants, and the function ϕ(t) is determined by the ﬁrst-order separable equation 2 ϕt = −2aϕ2 + 1 b(4C2 − C1 )ϕ3 , 2 whose solution can be written out in implicit form. 2. ∂w ∂t ∂x2 Functional separable solutions: = (aw2 + bw4 ) ∂2w + cw + kw–1 .

where the functions ϕ(t), ψ(t), and χ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = 1 bϕ(4ϕχ − ψ 2 ) + 2cϕ, 2

1 ψt = 2 bψ(4ϕχ − ψ 2 ) + 2cψ, 1 χt = 2 (bχ + a)(4ϕχ − ψ 2 ) + 2cχ + 2k.

It follows from the ﬁrst two equations that ϕ = Cψ, where C is an arbitrary constant. Remark. The above remains true if the equation coefﬁcients are arbitrary functions of time: a = a(t), b = b(t), c = c(t), and k = k(t). 3. . ∂t ∂x2 This is a special case of equation 1.6.16.4 with f (u) = auk . The transformation w(x, t) = xu(z, t), z = 1/x ∂2u ∂u = auk 2 . ∂t ∂z 4. . ∂t ∂x2 1◦ . Suppose w(x, t) is a solution of this equation. Then the function

k 2−n w1 = C1 w(C2 x, C1 C2 t + C3 ),

∂w

= ax4–k wk

∂2w

leads to a simpler equation of the form 1.1.9.18:

∂w

= axn wk

∂2w

where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. 2◦ . The substitution u = w 1−k leads to an equation of the form 1.1.15.6:

k ∂u ∂ ∂u = axn u 1−k . ∂t ∂x ∂x

3◦ . The transformation

w(x, t) = xu(z, t),

**leads to an equation of the similar form ∂2u ∂u = az 4−n−k uk 2 . ∂t ∂z
**

© 2004 by Chapman & Hall/CRC

v

w(x, t) =

ϕ(t)x2 + ψ(t)x + χ(t),

z = 1/x

48 5. ∂w = ax

3m+4 m+1

PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE

∂w ∂ wm . ∂t ∂x ∂x This is a special case of equation 1.1.15.6. The transformation 1 w(x, t) = x m+1 u(z, t), leads to a simpler equation of the form 1.1.10.7:

z=

1 x

∂ ∂u ∂u =a um . ∂t ∂z ∂z 6. . ∂t ∂x ∂x This equation occurs in nonlinear problems of heat and mass transfer and is a special case of equation 1.6.17.16 with f (w) = aw m . For n = 0, see equation 1.1.10.7. wm

m 2−n w1 = C1 w(C2 x, C1 C2 t + C3 ),

∂w

= axn

∂

∂w

1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. 2◦ . Solutions: w(x) = (Ax + B) m+1 ,

2−n 1 w(x, t) = k(λt + A)− m x m , 1 1 m

mλ k= a(n − 2)(2 + m − n − nm)

2−n

,

w(x, t) = t

(1−n)β

mβ xtβ a(2 − n)

+A

1 m

,

β=

1 , nm + n − m − 2

1 m

m λ w(x, t) = exp(−λt) (m + 1)2 x m+1 exp(λmt) + A a

,

n=

m+2 , m+1

where A, B, and λ are arbitrary constants. 3◦ . Multiplicative separable solution: w(x, t) = (λt + A)−1/m f (x), where the function f = f (x) is expressed via solutions to the Emden–Fowler equation

1 λ(m + 1) −n x F m+1 = 0, F = f m+1 . (1) am To the power-law particular solution of this equation there corresponds the second solution of the original equation in Item 1◦ . The order of equation (1) can be reduced; the equation is analyzed in detail in Polyanin and Zaitsev (2003), where its exact solutions for 26 different pairs of values of the parameters n and m are presented.

Fxx +

**4◦ . Self-similar solution for n ≠ −2: w = w(z), a(2 − n) wm wz z = xt n−2 , + z 1−n wz = 0. (2)
**

1

**where the function w(z) is determined by the ordinary differential equation
**

z

The book by Zaitsev and Polyanin (1993) presents the general solution of equation (2) for m = −1 and any n.

© 2004 by Chapman & Hall/CRC

and C2 are arbitrary constants. D. t) = x m+1 u(z.1. ∂t ∂z ∂z x ¤w 1 x Reference: V. t) = a m+1 where C is an arbitrary constant. which are deﬁned implicitly by wm dw = ξ + C2 . the change of variable Φ = ϕm+1 brings (6) to an equation that coincides. equation (6) has the ﬁrst integral In the special case n = m+1 1 aϕm ϕ = λmu− m+1 ϕ + C. © 2004 by Chapman & Hall/CRC . F. its order can be reduced (thereafter it can be transformed to an Abel equation of the second kind). To the special case C1 = 0 there corresponds the solution a(m + 1) m mλ λ(m + 1) + C|x| m+1 exp − t w(x. . 1 m ξ = ln |x| − λt. there are solutions of the form w = w(ξ). This equation is homogeneous. up to notation. Polyanin (1996). 7◦ . C1 . In the general case. (3) This equation is homogeneous. the change of variable G = g m+1 brings (3) to the equation 1 m G = A ζ 1−n G− m+1 G + A ζ −n G m+1 . so its order can be reduced (thereafter it can be transformed to an Abel equation of the second kind). Generalized self-similar solution: w = eλ(n−2)t ϕ(u). Self-similar solution: = βζgζ + αg. (6) where the function ϕ(u) is determined by the ordinary differential equation aun (ϕm ϕu )u = λmuϕu + λ(n − 2)ϕ. Zaitsev and A. with (5). α is any. The transformation w(x. hence. λ is any. 2003). and. ζ = xtβ . ζζ 1 ζ 2 (5) where A1 = β/a and A2 = α(m + 1)/a. β= . 6◦ . 1 z= leads to an equation of the similar form 4+3m−n−nm ∂ ∂u ∂u m+1 = az um . In the general case. Exact analytical solutions of equation (5) for various values of the parameters n and m can be found in the books by Polyanin and Zaitsev (1995. (4) To C = 0 in (4) there corresponds the third solution in Item 1◦ .1. EQUATIONS WITH POWER-LAW NONLINEARITIES 49 mα + 1 . In the special case 1 1−n . n−2 where the function g(ζ) is determined by the ordinary differential equation w = tα g(ζ). u To C = 0 there corresponds the last solution in Item 1◦ . For n = 2. α= nm + n − m − 2 nm + n − m − 2 a ﬁrst integral of equation (3) is given by ag m gζ = βζ 1−n g + C. u = xeλmt . 8◦ . β= aζ n g m gζ ζ 5◦ . awm+1 − λ(m + 1)w + C1 where λ. m+2 . t).

7: ∂ ∂u ∂u =A um .5 with f (x) = axn .1. and M.10.17. ∂t x ∂x ∂x This equation occurs in nonlinear problems of heat and mass transfer. Feix (1981).1.50 7. 3m + 4 . E. z ¤y Reference: V. Munier. the transformed equation becomes ∂ ∂u ∂u =A u−2 . R. C2 . Burgan.10.1. This is a special case of equation 1. up to 3◦ . t) = x m+1 u(ξ.3 (which can further be reduced to the linear heat equation). ∂w ∂t =a ∂ ∂x xn w m ∂w ∂x . For n = 0. ∂w ∂t = ∂ ∂x PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ax + b cw + k 2 ∂w ∂x . Fijalkow. F. Equation with n = 5 are encountered in the theory of static turbulence. t) is a solution of the equation in question.10. 1◦ . the transformed equation is very simple and coincides. is also a solution of the equation.6. where C1 . D. 2◦ . In the special case of n = 2 and m = −2. t). a ∂ ∂w ∂w = n xn w m . The transformation w(x.7. J. Zaitsev and A. R. . Let m ≠ −1 and 2m − 2n − nm + 3 ≠ 0. ∂t ∂ξ ∂ξ so it coincides with equation 1. ∂t ∂x ∂x © 2004 by Chapman & Hall/CRC z ¤y Reference: A. The substitution u = cw + k (c ≠ 0) leads to an equation of the form 1.3: ax + b ∂ ∂u ∂u = u−2 . with equation 1. see equation 1. leads to an equation of the similar form 3m−3n−2nm+4 ∂ ∂u ∂u =A ξ 2m−2n−nm+3 um . Gutierres. In the special case n = 2m + 3 notation. Then the function m 2−n w1 = C1 w(C2 x. and C3 are arbitrary constants.1. ∂t ∂ξ ∂ξ 1−n 2m−2n−nm+3 m+1 ξ=x (1) where A = a 2m − 2n − nm + 3 m+1 2 . and to n = 2 there correspond spherically symmetric problems. Suppose w(x. J.10. To n = 1 there correspond two-dimensional problems with axial symmetry. Polyanin (1996). 8. 9. C1 C2 t + C3 ). ∂t ∂ξ ∂ξ 4◦ .

I. one obtains from (2) an equation of the form 1. B. b = 0. c = 1. Zel’dovich and A. b = 0.6. 10. ξ= the original equation can be reduced to equation 1. P. t) 1 2m+3 . b = 0. n= 4 2m − 3 which admits a wide class of exact solutions. P. G. 2 Reference: V. V. 1 m . F. By the transformation w(x. S.5: 4−2m ∂v ∂ ∂v = F (ξ)v 2m−3 .1.11. Sedov (1993).17. A. © 2004 by Chapman & Hall/CRC . The transformation √ w(x. . k = 2a(nm + m + 2). c = 1. c = 1. Zaitsev and A. Kurdyumov.5 with f (u) = ku−2m . t) ax2 + bx + c. Kompaneets (1950). m = 1. Samarskii. S. leads to the equation z= dx ax2 + bx + c (1) ∂w = k(ax2 + bx + c)m w4–2m ∂2w ∂2u ∂u = ku4−2m 2 + k(ac − 1 b2 )u5−2m . | ¤{ . B. t) = mx2 A − kt 1 m 1 m+1 .6. Mikhailov (1995).1. m(n+1) mx2 A|kt + B|− nm+m+2 − kt + B 1 m . ∂ ∂ϕ ∂ϕ =k ϕn + pϕn+1 .8. A. (4) ∂t ∂ξ ∂ξ where the function F (ξ) is deﬁned parametrically by dx k . 2◦ . cos ξ ∂ξ m = 1. B. a = −1. Using the change of variable ϕ = u2m−3 . ∂t ∂ξ v 2 ∂ξ ∂ ∂v =k ∂t ∂ξ ∂ ∂v =k ∂t ∂ξ | ¤{ dx . k = 2a(nm + m + 2). n=− References: Ya. p = k(2m − 3)(ac − 1 b2 ). Polyanin (1996). L. t) = u(z. ξ= . A. Galaktionov. m 4aλ t + λx2 w(x.1. a = 1. and λ are arbitrary constants. and A. (2) 4 ∂t ∂z which has a traveling-wave solution u = u(z +λt) and a multiplicative separable solution u = f (t)g(z). Barenblatt (1952. v 2 ∂ξ ξ −3/2 ∂v . t) = w(x. Ya. m= a = −1. ∂t ∂z ∂z 4 − 2m . Raiser (1966). m+2 . Zel’dovich and Yu. EQUATIONS WITH POWER-LAW NONLINEARITIES 51 Solutions: w(x) = Ax1−n + B w(x. t) = A exp − m where A. (ax2 + bx + c)m (3) cosh2 ξ ∂v . I. (5) F (ξ) = 2 + bx + c)m 2 + bx + c)m (ax (ax Note some special cases of equation (4) where the function F = F (ξ) can be written out in explicit form: ∂ cos2 ξ ∂v ∂v =k . ∂t ∂x2 This is a special case of equation 1. P. 1◦ . 1989). D. 1 .16. t) = v(ξ.

Equations with Exponential Nonlinearities 2 1. A. Zaitsev and A. Galaktionov. ∂2w ∂w =a + beλw . Then the functions λ 2λ w1 = w( C1 x + C2 . ξ = √ . H.52 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1. t) = − ln −β + C exp µx − 2 aλt . V. A. C2 . A. 2 z = µx + σt. Ibragimov (1994). Solution: 2◦ . µ= a aλ . } where C1 . 2 λ w = u(ξ) − References: N. t) is a solution of this equation. x 1 ln t. aλ2 . (1) σ = µ2 . D. 3◦ . Equations of the Form ∂w = a ∂ w + b0 + b1 eλw + b2 e2λw ∂t ∂x2 1.2. Mikhailov (1995). Polyanin (1996). λ 2 1 w(x. are also solutions of the equation. t) = − ln β + C exp µx − 2 aλt . z = kx + βt. Traveling-wave solutions: 2 1 w(x. where the function w(z) is determined by the autonomous ordinary differential equation ak 2 wzz − βwz + beλw = 0. F. ∂t ∂x2 This equation occurs in heat and mass transfer and combustion theory. ∂w = ∂2w β= b − . S. 1◦ . P. Samarskii. Kurdyumov. and C3 are arbitrary constants. 2◦ . λ where C is an arbitrary constant.2. z=2 λ f (τ ) λτ f (τ ) + 2 where the function f (τ ) is deﬁned by √ C2 − 2 ln τ + τ 2 + k √ . k= 4b . w = ln |f (τ )|.1. λ t where the function u(ξ) is determined by the ordinary differential equation 1 1 auξξ + ξuξ + + beλu = 0. and A. For µ= 1 2 aλ. © 2004 by Chapman & Hall/CRC } } ¤~ ¤~ + a + beλw . Traveling-wave solution (k and β are arbitrary constants): w = w(z). C1 t + C3 ) + 2 ln |C1 |. Suppose w(x. Reference: V. P. f (τ ) = λ τ2 + k and C1 and C2 are arbitrary constants. 2. ∂t ∂x2 1◦ . } w = w(z). The solutions of Item 1◦ are special cases of the traveling-wave solutions that satisfy the autonomous equation µ2 wzz − σwz + a + beλw = 0. the general solution of equation (1) can be written out in parametric form as 2 fτ (τ ) dτ + C1 .

2◦ .2. t) is a solution of the equation in question. λ leads to an equation with quadratic nonlinearity: ∂u ∂x 2 (2) ∂ 2u ∂u =u 2 − ∂t ∂x − aλu2 − bλu − cλ. Suppose w(x. ◦ ¤ Reference: V. 3◦ .2. Equations of the Form ∂w = a ∂ eλw ∂w ∂t ∂x ∂x 1. Then the function w1 = w(C1 x + C2 . 1. 4 . λ λ where A. ln √ λ B − 2at 1 1 w(x. . β (1) where C is an arbitrary constant and the parameter β is determined by solving the quadratic equation 2◦ . λ aβ 2 + bβ + c = 0. and µ are arbitrary constants. C4 are arbitrary constants.1. The substitution u = e−λw u √ 1 ln −cλ x − bλt + C . ∂t ∂x ∂x This equation governs unsteady heat transfer in a quiescent medium in the case where the thermal diffusivity is exponentially dependent on temperature. βw + C2 √ y = x/ t. 2 © 2004 by Chapman & Hall/CRC .2. Traveling-wave solutions for a = 0: w(x. The particular solution u = β + C exp(λt + µx) of this equation generates a solution (1). λ C1 1◦ . Self-similar solution: w = w(y). . Zaitsev and A. The ﬁrst solution is self-similar and the second one is an additive separable solution. t) = − 3◦ . eλw dw . . D. is also a solution of the equation. t) = where the function w(y) is determined by the ordinary differential equation a(eλw wy )y + 1 ywy = 0. ∂w =a ∂ eλw ∂w + f (w) . 1◦ . t) = − ln C − 2aλµt + ln λµx2 + Ax + B . . F. t) = − µ= 1√ −cλ. C3 1 ln 2 . Traveling-wave solution in implicit form: x + βt + C1 = a 4◦ . Polyanin (1996). C3 t + C4 ) + where C1 . w(x. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 53 3. Solutions: x+A 2 . C. Traveling-wave solutions for a ≠ 0: w(x. 1 ln β + C exp( µx − aλt) . ∂2w ∂w + a + beλw + ce2λw . Solutions (1) and (2) are special cases of a wider class of traveling-wave solutions w = w(x + σt). = ∂t ∂x2 Equations of this form are encountered in problems of heat and mass transfer and combustion theory. B.

1: w(x. w(x. B. λ where b is an arbitrary constant. S. − λ 8◦ . t) = 2 ln | x + C1 | − ln C2 e−bt − The ﬁrst solution is degenerate. Samarskii. and C3 are arbitrary constants. ax − B. ∂t ∂x ∂x 1◦ . b ∂ ∂u ∂u =a eu . leads to an equation of the form 1. The transformation w = bt + u(x. ∂t ∂x ¤ 6◦ .1. Then the function 2.2. P.9. Additive separable solutions: w(x.2. and the function U = U (ξ) is determined by the ordinary differential equation 2k − kλξUξ = a(eλU Uξ )ξ . Ibragimov (1994). t) = U (ξ) + 2kt. ∂τ ∂x ∂x 3. τ ). where C1 . 2b + 1 ln t. and A. A. w(x. t) = G(θ) − ∂2ϕ ∂ϕ = aϕ 2 .54 5◦ . 3◦ . and the function G = G(θ) is determined by the ordinary differential equation 2b + 1 + bθGθ = (aeλG Gθ )θ . ∂w = ∂ ew ∂w ∂x – a 2 ew . λ where β is an arbitrary constant. H. ew w1 = w(C1 x + C2 . Solution: 7◦ . is also a solution of the equation. ξ = xe−kλt . and C are arbitrary constants. V. Suppose w(x. t) = ln |C1 x + C2 | + bt + C3 . Solution: References: L. ζ = x + β ln t. θ = xtb . A. A. Galaktionov. Ovsiannikov (1959.1: τ= 1 bt e + const b ∂w =a ∂ ∂w 2a . P. Mikhailov (1995). V. 1982). 2◦ . Kurdyumov. and the function F = F (ζ) is determined by the ﬁrst-order ordinary differential equation (C is an arbitrary constant) w(x. Solution: PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE w(x. The substitution ϕ = eλw leads to an equation of the form 1. t) = F (ζ) − −ζ + βλF = aλeλF Fζ + C. where k is an arbitrary constant. + b. C2 . t) = ln © 2004 by Chapman & Hall/CRC ∂t ∂x Solutions: . 1 ln t. N. t + C3 ) − 2 ln |C1 |. a exp[2( ax + B)] + 2 exp( ax + B) + A 2a2 (t + C) where A. t) is a solution of this equation.

a ≠ 0. and C3 are arbitrary constants. ∂t ∂x ∂x 1◦ .9. t) = u(z) − 2aλe−z 2(eu uz )z − eu uz + bλeλu = (1 − λ)uz − 1. λ λ where the function u = u(z) is determined by the ordinary differential equation w(x. ∂ ∂w = ew + aew + b. 3◦ . and the function u = u(x) is determined by the ordinary differential equation b uxx + (ux )2 + C1 e−u − = 0. ∂ ∂w =a ew – bew . Suppose w(x. are also solutions of the equation. Additive separable solution for a = −k 2 < 0: w(x. z = 2 ln x + ln t. C2 . t) = ln C1 cos(kx) + C2 sin(kx) + bt + C3 .4: ∂ ∂u ∂u = eu ∂τ ∂x ∂x 6. 2◦ . where C1 . τ ). Additive separable solution: w(x.2.2. t) is a solution of this equation.2. where C1 .9: ∂2u ∂u = au 2 − bu2 . The substitution u = ew leads to an equation of the form 1. The transformation w = bt + u(x. ∂t ∂x ∂x 1◦ . see 1. Additive separable solution for a = k 2 > 0: ∂w b ≠ 0. are also solutions of the equation. t) = ln C1 cosh(kx) + C2 sinh(kx) + bt + C3 . w(x. C2 .3 for a solution).2. 2◦ .1. ∂t ∂x 5. and C3 are arbitrary constants. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 55 4. 1−λ 1 ln t. a Integrating yields the general solution in implicit form: C3 e−2u − 2C1 e−u + b a −1/2 du = x + C4 . The integral is computable. and C3 are arbitrary constants. so the solution can be rewritten in explicit form (if a = 1 and b > 0. 2◦ . ∂ ∂w ∂w =a ew + beλw . Then the functions λ−1 2λ w1 = w( C1 x + C2 . ∂t ∂x ∂x 1◦ . Then the functions ∂w w1 = w( x + C1 .1. C1 t + C3 ) + 2 ln |C1 |. C2 . where C1 . 3◦ . Solution for λ ≠ 0: © 2004 by Chapman & Hall/CRC . t) is a solution of this equation. Suppose w(x. leads to an equation of the form 1. t) = u(x) − ln(aC1 t + C2 ). τ= 1 bt e + const b + aeu . where C1 and C2 are arbitrary constants.2. C2 t + C3 ) + ln C2 .

+ beλw + c + e–λw . Posashkov (1989). ¤ Reference: A. and A. Functional separable solutions: =a eλw 1 ln eαt C1 cos(x β ) + C2 sin(x β ) + γ λ 1 w = ln eαt C1 cosh(x −β ) + C2 sinh(x −β ) + γ λ w= if abλ > 0. ∂w ∂t ∂x ∂x Traveling-wave solution: =a ∂ weλw ∂w .3. P. where γ = γ1.4 with f (t) = c and g(t) = s. 1. 2a . ∂t ∂x ∂x Functional separable solution: w= bλ 2 1 ln cλt − x + C1 x + C2 . A. Kurdyumov. λ 2a where C1 and C2 are arbitrary constants. Samarskii. 3. P. t) = 2 aC1 2bλt 1 ln C1 ebλt x + e + C2 ebλt . ∂w ∂ ∂w Here.6. Galaktionov and S.2 are roots of the quadratic equation bγ 2 + cγ + s = 0. 8. Reference: V. V. ∂t ∂x ∂x Functional separable solution: w(x. S. t) = ln C1 e2bt − where C1 and C2 are arbitrary constants.2. Mikhailov (1995).56 7.14. A. A. ∂t ∂x ∂x Generalized traveling-wave solution: 2. ∂ ∂w =a weλw + b. ∂ ∂w ∂w =a wew + b(w + 2). C1 and C2 are arbitrary constants and α = λ(bγ + c). ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w =a eλw + b + ce–λw . © 2004 by Chapman & Hall/CRC b (x + C2 )2 . ∂t ∂x ∂x This is a special case of equation 1. A. Equations of the Form ∂w = ∂ f (w) ∂w + g(w) ∂t ∂x ∂x 1. Galaktionov. λ λ w(x. a 2 1 ln C1 x + C1 t + C2 . ¤ β = bλ/a. t) = where C1 and C2 are arbitrary constants. ∂w w(x. if abλ < 0. 2 λ bλ where C1 and C2 are arbitrary constants.

6.2. ∂w ∂t =a ∂ ∂x weλw ∂w ∂x + b + ce–λw . ∂t ∂x ∂x Generalized traveling-wave solutions: w(x. ϕ(t) = C2 e−2cλt − a cλ −1/2 . 2 λ2 bn where C1 and C2 are arbitrary constants. Generalized traveling-wave solution: w(x. Self-similar solutions: w(x. t) = 1 ln C1 x + λ 2 aC1 + bλ t + C2 .2. 5. ∂w ∂t =a ∂ ∂x weλw ∂w ∂x + be–λw . λ bλ2 b where C1 and C2 are arbitrary constants. ∂w ∂t = ∂ ∂x ae2λw + bweλw b x + C1 1 ln √ . where C1 and C2 are arbitrary constants. t) = 2 aC1 2bλt c 1 ln C1 ebλt x + e + C2 ebλt − . − λ C2 − 2at aλ ∂w ∂x + c. 9. 8. λ where C1 and C2 are arbitrary constants.3. Generalized traveling-wave solutions: w(x. ∂ ∂w ∂w =a wn exp λwn + bw1–n . t) = 1 ln z λ 1/n . ∂w ∂t = ∂ ∂x ae2λw + bweλw ∂w ∂x . ∂w ∂t = ∂ ∂x aweλw + beλw ∂w ∂x + b + ce–λw . + b + cebλ/a e−λu . © 2004 by Chapman & Hall/CRC .1. The substitution w = u − b/a leads to an equation of the form 1. z = C1 ebnλt x + 2 aC1 2bnλt e + C2 ebnλt . t) = where C1 and C2 are arbitrary constants. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 57 4.5: ∂ ∂u ∂u = ae−bλ/a ueλu ∂t ∂x ∂x 7. t) = 1 ln λ ϕ(t)x + C1 ϕ(t) + b ϕ(t) λ ϕ(t) dt . Traveling-wave solution: w(x.

∂t ∂x ∂x ∂x 1◦ . ∂w ∂ ∂w + aw =b eλw . x a 1 ln t. Apart from the traveling wave w = w(x + λt). Then the function ∂w w1 = w eC1 x + 1 a C1 eC1 t + C 2 . t) is a solution of the equation in question. t) is a solution of this equation. ∂t ∂x This is a special case of equation 1. is also a solution of the equation. and β are arbitrary constants. aw2 + 2βw + C1 where C1 . The substitution a λw e dw u = exp λ leads to the linear heat equation ∂t u = ∂xx u. This is a special case of equation 1. t . there is also an exact solution of the form w = ϕ(ξ) − 3. Then the function 2 w1 = w(C1 x + C2 . Suppose w(x. one arrives at a simpler equation of the form 1. Other Equations Explicitly Independent of x and t ∂w + a + beλw + ce2λw .1. λ where C1 .58 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1.7 with f (w) = beλw . λ t λ where the function u = u(z) is determined by the ordinary differential equation w(x.3: ∂w ∂ 2 w = + a + beλw + ce2λw . and C3 are arbitrary constants. 2◦ . z = − ln t. λ λ 3◦ . C2 .4. Solution: © 2004 by Chapman & Hall/CRC .2. 2λ x ξ= √ . Suppose w(x. Traveling-wave solution in implicit form: 2b eλw dw = x + βt + C2 . ∂x2 1◦ .2. 2◦ . z = x + βt. λ λ where C1 . . ∂t ∂z 2 1. ∂w = 2.6.3.6. ∂w ∂t = ∂2w ∂x2 + aeλw ∂w ∂x 2 1 ln t. x to the new variables t. +β ∂t ∂x2 ∂x On passing from t. C2 . is also a solution of the equation. 4.8 with f (w) = aeλw . C1 t + C3 ) + ∂2w ∂w =a ∂2w + beλw ∂w 1 ln C1 . t) = u(z) + au − z − 1 a uz + = b(eλ uuz )z . C2 . and C3 are arbitrary constants. e C1 t + C 3 − C1 .6.

6◦ . 2◦ . √ y = x/ t. Then the function w1 = w(C1 x + C2 . t) = w(x. t) = w(x. C4 are arbitrary constants. Solution: 1 ln t. 4◦ . and the function F = F (ζ) is determined by the autonomous ordinary differential equation βλFζ − 1 = aλeλF Fζζ . λ where β is an arbitrary constant. t) = C1 1 z ln C2 exp λ ak 2 + β .2. − λ © 2004 by Chapman & Hall/CRC 7◦ . C2 . k. and the function G = G(θ) is determined by the ordinary differential equation 2b + 1 + bθGθ = aeλG Gθθ . where k is an arbitrary constant. Suppose w(x. and the function U = U (ξ) is determined by the ordinary differential equation 2k − kλξUξ = aeλU Uξξ . ξ = xe−kλt . is also a solution of the equation. C3 t + C4 ) + C3 1 ln 2 . t) = G(θ) − 2b + 1 ln t. t) = U (ξ) + 2kt. . Additive separable solutions: w(x. where C1 . = aeλw ∂t ∂x2 1◦ . . ζ = x + β ln t. ln 2 λ 2C2 (at + C1 ) sinh2 (C2 x + C3 ) 1 . 2 λ 2C2 (C1 − at) where C1 . C1 z = kx + βt. t) = cos2 (C2 x + C3 ) 1 . Solution: . . 2 5◦ . t) = F (ζ) − w(x. Solution: w(x. and C3 are arbitrary constants. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 59 5. λ where b is an arbitrary constant. ∂2w ∂w . Self-similar solution: w = w(y). Traveling-wave solution: w(x. 3◦ . t) is a solution of this equation. note that ln(A/B) = ln |A| − ln |B| for AB > 0. C2 .1. and β are arbitrary constants. θ = xtb . ln 2 λ 2C2 (at + C1 ) cosh2 (C2 x + C3 ) 1 ln . where the function w(y) is determined by the ordinary differential equation aeλw wyy + 1 ywy = 0. . λ C1 where C1 . w(x.

2 C1 e−bt − a 2C2 where C1 . 4◦ . In special cases. where k and β are arbitrary constants. 3◦ . ∂w . t) is a solution of this equation. t) = − ln(kt + C) + ϕ(x). note that ln(A/B) = ln |A| − ln |B| for AB > 0. 2 2C2 a − C1 e−bt b sinh2 (C2 x + C3 ) . t) = ln w(x. ξ = kx + βt. 2 a − C1 e−bt 2C2 b cosh2 (C2 x + C3 ) . © 2004 by Chapman & Hall/CRC . ∂ ∂w ∂w = ae2w + bwew + cew + ∂t ∂x ∂x Generalized traveling-wave solution: 7. Additive separable solutions for λ = 0: w(x. C1 t + C3 ) + 2 ln |C1 |. and the function w(ξ) is determined by the autonomous ordinary differential equation ak 2 ew wξξ − βwξ + beλw = 0. and the function ϕ(t) is determined by the ﬁrst-order separable ordinary differential equation ϕt = aϕ3 + cϕ2 . Solution for λ ≠ 0: 1−λ 1 ln t. ∂x w(x. t) = ln w(x. Traveling-wave solution: w = w(ξ). we have ϕ(t) = (C2 − 2at)−1/2 ϕ(t) = (C2 − ct) −1 if c = 0. λ λ where the function u = u(z) is determined by the ordinary differential equation w(x. and C3 are arbitrary constants. z = 2 ln x + ln t. t) = ln b cos2 (C2 x + C3 ) . Suppose w(x. t) = u(z) − 2aλeu−z 2uzz − uz + bλeλu = (1 − λ)uz − 1. where C1 is an arbitrary constant. Additive separable solution for λ = 1: w(x. where the function ϕ(x) is determined by the autonomous ordinary differential equation aϕxx + b + ke−ϕ = 0. and C3 are arbitrary constants. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w + beλw . Then the functions λ−1 2λ w1 = w( C1 x + C2 . C2 . 2◦ . C2 .60 6. are also solutions of the equation. t) = ln ϕ(t)x + (st + C1 )ϕ(t) + bϕ(t) ϕ(t) dt . = aew ∂t ∂x2 1◦ . whose general solution can be written out in implicit form. where C1 . 5◦ . if a = 0.

1.1 with f (t) = beµt . ∂w ∂w =a ∂ ∂w ∂ ∂w =a eλw + beµt . ∂t ∂x ∂x This is a special case of equation 1.6.6. Equations Explicitly Dependent on x and/or t 1. 10.6.14. ∂t ∂x ∂x This is a special case of equation 1. ∂t ∂x2 This is a special case of equation 1.2 with f (z. 6. ∂w ∂ =a eλw + beλw + ceµt . λ where C is an arbitrary constant and the function ϕ(x) is determined by the second-order linear ordinary differential equation aψxx + λ(bx + c)ψ + λ = 0. =a eλw ∂w ∂t ∂x ∂x Additive separable solution: =a ∂ eλw ∂w + (bx + c)eλw . ∂t ∂x ∂x This is a special case of equation 1.2. ∂t ∂x ∂x This is a special case of equation 1. © 2004 by Chapman & Hall/CRC ψ = eλϕ . 1 ln(λt + C) + ϕ(x). ∂w ∂t ∂x ∂x Additive separable solution: =a ∂ eλw ∂w + beλw+µx . w=− ∂w ∂ ∂w ∂w ∂ ∂w 9. 1 ln(λt + C) + ϕ(x).2.6.14. . 7.6. 5.4 with f (t) = ceµt and g(t) = 0. ∂t ∂x ∂x This is a special case of equation 1.14.14. =a eλw 8.9 with f (x. ∂t ∂x ∂x This is a special case of equation 1.4.1. + ceλw+bx+ct .2 with f (t) = beµt and g(t) = ceνt . ∂w ∂ ∂w ∂w =a eλw + btn e–λw . t) ≡ 0 and g(x.6. + beµt + ce–λw+νt .14. + be–λw+µt . w) = cez+λw .14. t) = be βx+µt .2 with f (t) = 0 and g(t) = bt n . ∂w =a ∂2w + aλ ∂w 2 ∂w =a ∂2w + beβx+µt–λw . 3. 2.6.6. + btn . ∂t ∂x2 ∂x This is a special case of equation 1. w=− ψ = eλϕ . λ where C is an arbitrary constant and the function ϕ(x) is determined by the second-order linear ordinary differential equation aψxx + λbeµx ψ + λ = 0. eλw 4.2 with f (t) = 0 and g(t) = be µt .5.1 with f (t) = btn . EQUATIONS WITH EXPONENTIAL NONLINEARITIES 61 1.

∂ ∂w ∂w =a coshk (βw) .14. 5.3.6. w) = β coshk (z + λw). and h(t) = c coshk (βt).1.2 with f (z. Self-similar solutions: x + C1 1 .1.12 with f (x) = axn .10 with f (w) = b coshk (λw). 2. ∂t ∂x2 This is a special case of equation 1.2. t) = arcsinh √ β C2 − 2at where C1 and C2 are arbitrary constants. ∂t ∂x ∂x This is a special case of equation 1. ∂w ∂t =a ∂ ∂x eλw ∂w ∂x + 2b cosh(λw) + c coshk (βt). w(x.6. 2 .1 with f (w) = b coshk (λw). eλw+µx ∂w =a ∂ ∂w 1. 6.62 11. .6. ∂2w ∂w ∂w = + b coshk (λw) 2 ∂t ∂x ∂x 2 + c coshk (βt) ∂w . This is a special case of equation 1. 7. + b coshk (λw).1. Equations with Hyperbolic Nonlinearities 1.6.6.12 with f (x) = aeµx . ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w =a xn eλw . g(t) = 0. ∂t ∂x ∂x This is a special case of equation 1. ∂x This is a special case of equation 1.1 with f (w) = s coshk (λw). This is a special case of equation 1.3.15.1 with f (w) = a coshk (βw).6.4 with f (t) = c coshk (βt) and g(t) = b.1 with f (w) = a cosh2 (βw).15. ∂w ∂t = ∂2w ∂x2 + b coshk (λw) ∂w ∂x 3. ∂w =a ∂2w ∂w =a ∂2w + β coshk (λw + bx + ct).8 with f (w) = b coshk (λw). ∂t ∂x ∂x This is a special case of equation 1. © 2004 by Chapman & Hall/CRC . 4. Equations Involving Hyperbolic Cosine 1. ∂ ∂w ∂w =a cosh2 (βw) .17. ∂2w ∂w ∂w =a + (bx + c) + coshk (λw).6.6. 12. ∂t ∂x2 This is a special case of equation 1. 2 ∂t ∂x ∂x This is a special case of equation 1. 8.6.17.6.6.6. ∂t ∂x ∂x This is a special case of equation 1.

∂w =a ∂2w ∂x2 + (bx + c) ∂w 2 .1 with f (w) = a sinh2 (βw). 5. ∂t ∂x2 This is a special case of equation 1.6.2.4 with f (t) = c sinhk (βt) and g(t) = −b. w(x. 2.6.3. This is a special case of equation 1.1 with f (w) = a sinhk (βw).15. ∂w =a ∂2w ∂w =a ∂2w + β sinhk (λw + bx + ct).2 with f (z.2. EQUATIONS WITH HYPERBOLIC NONLINEARITIES 63 1. This is a special case of equation 1.6.6. t) = arccosh √ β C2 − 2at where C1 and C2 are arbitrary constants. ∂w =a ∂2w + β tanhk (λw + bx + ct). + b sinhk (λw).3.10 with f (w) = b sinhk (λw).3.6.3.1. 6. © 2004 by Chapman & Hall/CRC . ∂2w ∂w =a + b tanhk (λw).1 with f (w) = b sinhk (λw).14. 4. ∂t ∂x This is a special case of equation 1. ∂ ∂w ∂w =a eλw ∂t ∂x ∂x + 2b sinh(λw) + c sinhk (βt).1. ∂t ∂x2 This is a special case of equation 1. + sinhk (λw).1.6. 2. g(t) = 0. This is a special case of equation 1.6.2 with f (z. ∂t ∂x2 This is a special case of equation 1.1.6. Self-similar solutions: x + C1 1 . ∂w ∂t =a ∂ ∂x sinh2 (βw) ∂w ∂x . and h(t) = c sinhk (βt).1 with f (w) = b tanhk (λw). This is a special case of equation 1. 8.6.8 with f (w) = b sinhk (λw). 7. ∂t ∂x2 This is a special case of equation 1. 1. ∂w ∂t = ∂2w ∂x2 + b sinhk (λw) ∂w ∂x 3. ∂2w ∂w ∂w = + b sinhk (λw) 2 ∂t ∂x ∂x 2 + c sinhk (βt) ∂w . ∂w ∂t =a ∂ ∂x sinhk (βw) ∂w ∂x .15.6.1. Equations Involving Hyperbolic Tangent 1. ∂x This is a special case of equation 1. w) = β tanhk (z + λw).6.1 with f (w) = s sinhk (λw). w) = β sinhk (z + λw). Equations Involving Hyperbolic Sine 1.6.

+ b cothk (λw). 1.1.15.4.2. Equations with Logarithmic Nonlinearities 2 1.1 with f (w) = b ln w. Equations Involving Hyperbolic Cotangent 1.10 with f (w) = b cothk (λw).8 with f (w) = b tanhk (λw).1.2.1 with f (w) = a cothk (βw).1 with f (w) = s tanhk (λw). and h(t) = c tanhk (βt).6. 4.6.64 3.6. This is a special case of equation 1. 2.1. ∂w ∂t = + b tanhk (λw) ∂w ∂x + c tanhk (βt) ∂w ∂x .1 with f (w) = a tanhk (βw).1 with f (w) = b cothk (λw). + b ln w. w) ∂t ∂x2 1. =a cothk (βw) ∂w ∂x . w) = β cothk (z + λw). ∂w ∂t ∂w ∂t = + b cothk (λw) ∂w ∂x 2 + c cothk (βt) ∂w ∂x .6.6. and h(t) = c cothk (βt). This is a special case of equation 1. ∂w =a ∂2w © 2004 by Chapman & Hall/CRC .6. ∂t ∂x2 This is a special case of equation 1.6. ∂w ∂t = ∂2w ∂x2 ∂2w ∂x2 + b tanhk (λw) ∂w ∂x 2 2 . ∂2w ∂w ∂w = + b cothk (λw) 2 ∂t ∂x ∂x ∂2w ∂x2 ∂ ∂x 2 ∂2w .15.2 with f (z. ∂t ∂x2 ∂x This is a special case of equation 1. ∂t ∂x2 This is a special case of equation 1. 1.1.4. 5. 2 ∂t ∂x ∂x This is a special case of equation 1. This is a special case of equation 1.4. ∂ ∂w ∂w =a tanhk (βw) . PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w ∂w + (bx + c) =a + tanhk (λw). 6. 6.6.10 with f (w) = b tanhk (λw).6. This is a special case of equation 1. This is a special case of equation 1.6.1 with f (w) = s cothk (λw). ∂w + (bx + c) + cothk (λw). ∂t ∂x2 This is a special case of equation 1.8 with f (w) = b cothk (λw).3. Equations of the Form ∂w = a ∂ w + f (x.6.6.6. 5. g(t) = 0. ∂w =a 4. t. 3.6. ∂t ∂x ∂x This is a special case of equation 1. ∂w =a ∂2w ∂w =a ∂2w + β cothk (λw + bx + ct).6. g(t) = 0.

3.9.4. References: V.1. EQUATIONS WITH LOGARITHMIC NONLINEARITIES 65 2. 5. 4◦ .7.5 and 1. where the function f (ξ) is determined by the autonomous ordinary differential equation fξξ + (fξ )2 − bfξ + af = 0. w1 = exp(C1 eat )w( x + C2 . t) = exp − ¤ 1 at a(x + A)2 + e ln(1 + Be−at ) + Ceat . t) = exp 1 2 A2 2at e + Beat . ∂t ∂x2 This is a special case of equation 1. 1982). ∂t ∂x2 This is a special case of equations 1. + aw ln w + bw. a w(x. (2) Relations (1) and (2) involve three arbitrary constants.4.6. P. B.6.6. Mikhailov (1995). Then the functions where C1 . There are more complicated solutions of the form w(x.1. Suppose w(x. C2 . 1◦ . P. where the function f (x) is deﬁned implicitly by (1) 1 Be−2f − af + 2 a −1/2 df = x + C. A. ∂t ∂x2 This is a special case of equation 1. S. V. + aw ln w + (bx2 + cx + k)w. 2◦ . Samarskii. t) = exp Aeat x + w(x.6. 6. + aw ln w + (bx + c)w. Functional separable solutions: w(x. Kurdyumov. t) = exp Aeat + f (x) .6. A. ∂t ∂x2 The substitution w = e−b/a u leads to an equation of the form 1. −at ) 4(1 + Be 2B 1 − 4 a(x + A)2 + Beat . ∂t ∂x2 4.7. t + C3 ).1.2: ∂u ∂ 2 u = + au ln u. and A. ∂2w ∂w + aw ln w. t) = exp Aeat + f (x + bt) . = ∂t ∂x2 This is a special case of equation 1. Multiplicative separable solution: w(x. t) is a solution of the equation in question.1. = ∂w ∂2w ∂w = ∂2w ∂w = ∂2w ∂w = ∂2w © 2004 by Chapman & Hall/CRC . and C. Dorodnitsyn (1979.1. where A. 3◦ . A. A.1. B. are also solutions of the equation. and C3 are arbitrary constants.1 with f (w) = aw ln w. + aw ln w + (bx + ct + k)w. A. and C are arbitrary constants. Galaktionov.1.

where C1 and C2 are arbitrary constants. Posashkov (1989). t) = C1 exp −at x −a . are also solutions of the equation. ∂t ∂x On passing from t. t) = 2 C1 − at (C1 − at) where C1 and C2 are arbitrary constants. + (1 + kw) a ln2 (1 + kw) + b ln(1 + kw) + c . a ∂ ∂w = k xk + bw ln w. © 2004 by Chapman & Hall/CRC . Equation (1) has also generalized separable solutions of the following forms: √ √ u(x.6. h(t) = c. A. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w + a(w + b) ln2 (w + b). 4. Galaktionov and S. A. ∂w =a ∂2w ∂x2 + bx ∂w ∂w =a ∂2w + bt ∂w ∂w = ∂2w +a ∂w + cw ln w. ¤ ∂u ∂x 2 + au2 .6 with f (t) = 0. Posashkov (1989).4. t) = ϕ(t) + ψ(t) A exp(x −a ) + A exp(x −a ) if a < 0. 8.2: ∂w ∂ 2 w = + bw ln w.10. 3◦ . A. z = x + at. = ∂t ∂x2 1◦ .1. √ C2 1 + exp −at x −a .66 7. one obtains a simpler solution of the form 1.6. ∂w 1◦ . t + C2 ).1.2.7. ∂t ∂x2 This is a special case of equation 1. g(t) = 0. ∂x2 3.4. + cw ln w. ∂t ∂x This is a special case of equation 1. and p(t) = s(t) = 0.6.4.4: ∂u ∂ 2 u = + ∂t ∂x2 2◦ . ∂t x ∂x ∂x The values k = 1 and k = 2 correspond to problems with axial and central symmetry. t) = ϕ(t) + ψ(t) A sin(x a ) + A cos(x a ) For details. t) is a solution of the equation in question. and p(t) = s(t) = 0.2. A. ∂t ∂x This is a special case of equation 1. The ﬁrst solution is a traveling-wave solution and the second one is a generalized separable solution. see 1. The substitution w = eu − b leads to an equation of the form 1. Other Equations + bw ln w. ∂x2 2.4. 1. Zaitsev (2002). Suppose w(x. u(x. u(x.6 with f (t) = b.1. A. √ √ if a > 0. Polyanin and V. Solutions of equation (1) for a < 0: √ u(x. x to the new variables t. F. Galaktionov and S. Then the functions w1 = exp(C1 ebt ) w( x. D. (1) References: V. ∂t ∂z 2 1. = ¤ ∂w ∂2w Reference: V. respectively. g(t) = bt.2. h(t) = c.1.

Generalized traveling-wave solution: w(x. Suppose w(x. t) is a solution of this equation. Kurdyumov. A. V. P. C1 t + C3 ). t) = exp Aebt θ(x).6.1 with f (w) = a ln w + b.15. Mikhailov (1995).8 with f (w) = a lnk (bw). C2 . −bt ) 4a(1 + Ae 2A where A and B are arbitrary constants. This is a special case of equation 1. Samarskii. Then the function ∂x2 w1 = eC1 w(x + bC1 t + C2 . Traveling-wave solution: c b . ◦ 3 . ¤ Reference: A. ∂t ∂x ∂x This is a special case of equation 1. is also a solution of the equation. b(t + C2 ) b t + C2 b . © 2004 by Chapman & Hall/CRC . ∂w = ∂ (a ln w + b) ∂w where C1 .6.1. t) = exp √ a C1 − 2at where C1 and C2 are arbitrary constants.6. ∂t ∂x 1◦ . w(x. and C3 are arbitrary constants. t) = exp C1 exp − x + b2 C2 t + 1 − aC2 − a b where C1 and C2 are arbitrary constants. t + C3 ). 1◦ . w(x. t) = exp − 1 bx2 + Bebt + (k + 1)ebt ln(1 + Ae−bt ) . S. t) is a solution of the equation in question. 7. where A is an arbitrary constant and the function θ(x) is determined by the second-order ordinary differential equation dθ a d xk + bθ ln θ = 0. is also a solution of the equation. EQUATIONS WITH LOGARITHMIC NONLINEARITIES 67 2◦ . . . ∂w ∂t = ∂2w ∂x2 + a lnk (bw) ∂w ∂x 2 ∂w =a ∂2w + (b ln w + c) ∂w a ln |t + C2 | c C1 − x + 2 − . C2 . w(x. Galaktionov. Multiplicative separable solution: w(x. Functional separable solution: w(x. a a+b C2 x − . 3◦ . and A. A. and C3 are arbitrary constants. The ﬁrst solution represents a traveling wave and the second one is self-similar. Suppose w(x. 2◦ . Then the function 2 w1 = w(C1 x + C2 . P. where C1 . t) = exp 6.4. xk dx dx 5. Solutions: 2 2C1 x + 2aC1 t + C2 − b . t) = exp 2◦ .

2: ∂ ∂u =a ∂t ∂x u2 1 ∂u . + 1 ∂x ∂w = a cos2 (λw + β) © 2004 by Chapman & Hall/CRC w(x. 4.1.68 8.2. 2.6. ∂w + (bx + c) + cosk (λw). ∂w = ∂ (a ln w + b) ∂w + cw ln w + w. ϕ(t) = C2 e−2ct − a c where C1 and C2 are arbitrary constants. ∂t ∂x2 This is a special case of equation 1.5.6. ∂t ∂x ∂x Generalized traveling-wave solution: 9.1.6.6. ϕ(t) . ∂x This is a special case of equation 1.6. ∂w ∂t = ∂2w ∂x2 + b cosk (λw) ∂w ∂x 2 3. ∂2w ∂w ∂w = + b cosk (λw) ∂t ∂x2 ∂x ∂2w + c cosk (βt) ∂w . 2 .10 with f (w) = b cosk (λw). ∂t ∂x2 This is a special case of equation 1. Equations with Trigonometric Nonlinearities 1.1 with f (w) = s cosk (λw).2 with f (z. .6. g(t) = 0. ∂t ∂x2 ∂x This is a special case of equation 1. w(x. 5.8 with f (w) = b cosk (λw). 1. .1. and h(t) = c cosk (βt). Equations Involving Cosine 1. 6. ∂t ∂x ∂x Generalized traveling-wave solution: c a+b C2 x − (C1 − 2at) − . + b cosk (λw).13. ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w = (a ln w + b) + cw. t) = exp ϕ(t)(C1 x) + (a + b)ϕ(t) −1/2 ϕ(t) dt + sϕ(t) dt . ∂t ∂x2 The substitution u = tan(λw + β) leads to an equation of the form 1.6. ∂w =a ∂2w ∂2w ∂w =a + β cosk (λw + bx + ct). w) = β cosk (z + λw).5.1. ∂w =a ∂2w This is a special case of equation 1.1 with f (w) = b cosk (λw). t) = exp √ a C1 − 2at 3a where C1 and C2 are arbitrary constants.

6.1.15.1 with f (w) = s sink (λw). g(t) = 0. ∂t ∂x u2 + 1 ∂x ∂w =a ∂ sin2 (βw) ∂w ∂w = a sin2 (λw) ∂2w + c sink (βt) ∂w 7. 7. ∂w = ∂2w + b sink (λw) ∂w 2 .6. ∂t ∂x2 The substitution u = cot(λw) leads to an equation of the form 1.5.5. w(x. w) = β sink (z + λw).6.6.6. . and h(t) = c sink (βt).1 with f (w) = a cos2 (βw).10 with f (w) = b sink (λw). ∂t ∂x ∂x This is a special case of equation 1. ∂t ∂x ∂x This is a special case of equation 1. © 2004 by Chapman & Hall/CRC .1.15. ∂w =a ∂ cosk (βw) ∂w 1.15. Self-similar solutions: x + C1 1 . ∂t ∂x ∂x This is a special case of equation 1.6.1 with f (w) = a sin2 (βw).13. t) = arcsin √ β 2at + C2 where C1 and C2 are arbitrary constants. ∂t ∂x ∂x This is a special case of equation 1. + b sink (λw). EQUATIONS WITH TRIGONOMETRIC NONLINEARITIES 69 ∂ ∂w =a cos2 (βw) .1. ∂t ∂x2 ∂x This is a special case of equation 1.6. 5. ∂t ∂x2 This is a special case of equation 1. t) = arccos √ β 2at + C2 where C1 and C2 are arbitrary constants. w(x. . 6. Self-similar solutions: x + C1 1 . 2 ∂t ∂x ∂x This is a special case of equation 1.1 with f (w) = a cosk (βw).2.6.1 with f (w) = b sink (λw). ∂w = ∂2w ∂x2 + b sink (λw) ∂w 2 3.6.15.2. ∂2w ∂w ∂w =a + (bx + c) + sink (λw). ∂w =a ∂ sink (βw) ∂w 8. ∂w =a ∂2w ∂w =a ∂2w + β sink (λw + bx + ct). ∂t ∂x2 This is a special case of equation 1.2 with f (z. . Equations Involving Sine 1. .1 with f (w) = a sink (βw). . ∂w 8.8 with f (w) = b sink (λw).1. ∂t ∂x ∂x This is a special case of equation 1. 4. 2.2: ∂ 1 ∂u ∂u =a .6.6.

2. ∂2w ∂w ∂w =a + (bx + c) + tank (λw).6.6.1. .1. . and h(t) = c cotk (βt). ∂w =a ∂2w ∂w =a ∂2w + β cotk (λw + bx + ct). Equations Involving Cotangent 1. ∂w =a ∂2w 2 ∂2w ∂w 2 ∂w ∂w = + b cotk (λw) + c cotk (βt) .6. + β tank (λw + bx + ct).70 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1. + b cotk (λw). .5.1 with f (w) = a tank (βw). ∂t ∂x ∂x This is a special case of equation 1. ∂w =a ∂ cotk (βw) ∂w © 2004 by Chapman & Hall/CRC . ∂t ∂x2 ∂x ∂x This is a special case of equation 1.15.1 with f (w) = b tank (λw). ∂t ∂x2 This is a special case of equation 1.6. ∂t ∂x2 This is a special case of equation 1.8 with f (w) = b cotk (λw). 2 ∂t ∂x ∂x This is a special case of equation 1.2 with f (z.1. 1. 5. 2.6.1 with f (w) = s cotk (λw). 2.1 with f (w) = b cotk (λw).3.5.6.6. ∂t ∂x ∂x This is a special case of equation 1.10 with f (w) = b tank (λw). 6.6. ∂t ∂x2 ∂x This is a special case of equation 1. w) = β cotk (z + λw).6. w) = β tank (z + λw). 6.6. ∂w = ∂2w + b cotk (λw) ∂w 3.1.6.6.4.6.6.8 with f (w) = b tank (λw).2. g(t) = 0. Equations Involving Tangent ∂2w ∂w =a + b tank (λw). g(t) = 0.6. . ∂w = ∂2w + b tank (λw) ∂w 2 ∂w =a ∂2w 3.6. and h(t) = c tank (βt). ∂t ∂x2 This is a special case of equation 1.15. 5. ∂t ∂x2 ∂x This is a special case of equation 1. ∂w + (bx + c) + cotk (λw).1 with f (w) = s tank (λw). 4.10 with f (w) = b cotk (λw).1 with f (w) = a cotk (βw). 2 ∂t ∂x ∂x ∂x This is a special case of equation 1. ∂t ∂x2 ∂x This is a special case of equation 1.2 with f (z. ∂w =a ∂ tank (βw) ∂w + c tank (βt) ∂w 1. ∂t ∂x2 This is a special case of equation 1. 4. ∂w = ∂2w + b tank (λw) ∂w 2 .

For f = f (w) having power-law. K. ∂t Solution: 2.1 to 1. Ibragimov (1989).3.2. respectively. ∂w = ∂w ∂x 2 –1 +1 exp k arctan ∂w ∂x ∂2w ∂x2 . 1. £ f (w) dw dw = C2 x.4. w2 = u(z) − x2 . or logarithmic form. k ≠ 0. The equation has an implicit nonstationary solution independent of the space variable: dw = t + C. w = tan ϕ(z) + arctan 2zϕz − x2 = z cos2 ϕ(z) + 1 ln t . Gazizov. ∂w =a ∂2w 2◦ .3.5. z = x + λt. 1◦ . Traveling-wave solutions: 2 a −1/2 © 2004 by Chapman & Hall/CRC £ w = w(z).1. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 71 1. Akhatov. combustion theory. or equations 1. Stationary solution in implicit form: C1 − 3◦ .1.3. z = t exp −k arctan(x/w) .2.5. Equations Involving Arbitrary Functions 2 1.6.4. w) ∂t ∂x2 1. Equations of the Form ∂w = a ∂ w + f (x. and ecology.7. exponential.1.1. 1. ∂t ∂x2 Kolmogorov–Petrovskii–Piskunov equation.1. ∂w = ∂ eλ arctan w ∂w . Equations of this form are often encountered in various problems of heat and mass transfer (with f being the rate of a volume chemical reaction).1.1.6. 2 −1 1 2 kzu uz where the function u(z) is determined by the ordinary differential equation 2k 2 z 2 uuzz − k 2 z(3zuz − 2u)uz − 4u2 − 1 (k 2 z 2 uz + 4u2 )uz exp k arctan 2 = 0.4. + f (w).2. Sh.6.1 to 1.1. λ 1 ln t . Equations Involving Inverse Trigonometric Functions ∂t ∂x 1 + w2 ∂x Solution in parametric form: 1.8. t. 2z λ 2 λ(C − z) Reference: I. see Subsections 1. and 1. H.4. biology. . λ where z is the parameter and the function ϕ = ϕ(z) is determined by the ﬁrst-order ordinary differential equation (C is an arbitrary constant) ϕz = ¢ ¤¡ 1 1 C −z 1 tan ln −ϕ − .1. equations 1.1. R. f (w) C is an arbitrary constant. 1. and N.

where the function w(ξ) is determined by the ordinary differential equation ab2 wξξ − cwξ + f (ξ. ∂w =a ∂2w w1 = exp C1 ebt w x + C2 . The book by Polyanin and Zaitsev (2003) presents a considerable number of solutions to equation (2) for various f = f (w). w) = 0. 2.2 (e. leads to the equation x ξ= √ t ∂ 2 w 1 ∂w ∂w =a 2 + ξ + f (ξ. + bw ln w + f (t)w. Suppose w(x. t) is a solution of this equation. is also a solution of the equation.. 2◦ . 4◦ . ∂t ∂x2 1◦ . Generalized traveling-wave solution: w(x. where C1 and C2 are arbitrary constants. 1 . C 2 t). t) = exp Aebt x + Bebt + where A and B are arbitrary constants.g. ∂τ ∂ξ 2 ∂ξ which admits exact solutions of the form w = w(ξ).w . ∂2w ∂w + f (bx + ct. w). Suppose w(x. t w1 = w(Cx. (1) The transformation ξ = (λ/a)z. Then the functions ¤ 4. Subsection S. The function w = w(z) is determined by the autonomous ordinary differential equation awzz − λwz + f (w) = 0. ◦ ∂w ∂t =a ∂2w ∂x2 + 1 t f x √ . see Example 1) presents an exact solution of this equation with f (w) deﬁned parametrically. . 3. =a ∂t ∂x2 Solution: w = w(ξ). © 2004 by Chapman & Hall/CRC a 2 2bt A e + ebt b e−bt f (t) dt . brings (1) to the Abel equation U (w) = wξ (2) U Uw − U + aλ−2 f (w) = 0. t) is a solution of this equation.72 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE where λ is an arbitrary constant. t . Then the function where C is an arbitrary constant. w). The transformation τ = ln t. are also solutions of the equation. ξ = bx + ct. 2◦ .5.

A − 4aebt ψ(t) = Bebt + ebt e−bt 2aϕ(t) + f (t) dt. There are also functional separable solutions of the more general form w(x. t) = exp ebt one arrives at the simpler equation ∂2u ∂u = a 2 + bu ln u + f (x)u. and the function ϕ(t) is determined by the ordinary differential equation aϕxx + bϕ ln ϕ + f (x)ϕ = 0. Solution: w(x.1.6. t) where C is an arbitrary constant. 2◦ . t) = exp Aebt + ebt e−bt f (t) dt + Φ(x + λt) . Here. ∂t ∂x 5.4. t) . and ϕ0 (t) are determined by a system of ordinary differential equations (see equation 1. where the functions ϕ2 (t). t) = exp Cebt + ebt e−bt g(t) dt ϕ(x). 5◦ . Functional separable solution: w(x. With the substitution w(x.2: ∂2u ∂u = a 2 + bu ln u. the functions ϕ(t) and ψ(t) are given by ϕ(t) = bebt . where A and λ are arbitrary constants. 4◦ . t) = exp ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) . ∂t ∂x2 1◦ . + bw ln w + f (x) + g(t) w. Multiplicative separable solution: w(x. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 73 3◦ .9) that can be integrated. t) = exp ϕ(t)x2 + ψ(t) . and the function Φ = Φ(z) is determined by the autonomous ordinary differential equation aΦzz + a(Φz )2 − λΦz + bΦ = 0.1. the order of which can be reduced by one. ∂w =a ∂2w e−bt f (t) dt u(x. The substitution w(x. t) = exp ebt leads to a simpler equation of the form 1. ∂t ∂x © 2004 by Chapman & Hall/CRC e−bt g(t) dt u(x.6. where A and B are arbitrary constants. 6◦ .1. ϕ1 (t).

¦ ¤¥ −1 eF dt . Generalized traveling-wave solution: 7. ϕ1 (t). There are also functional separable solutions of the form w(x.9).1. 3◦ . Generalized traveling-wave solution: w(x. which can be integrated. where the functions ϕ(t) and ψ(t) are given by ϕ(t) = eF A − 4a ψ(t) = BeF + eF and A and B are arbitrary constants. Zaitsev and A. F. which can be integrated. and A and B are arbitrary constants. F = f dt. t) = exp Φ(t)x + Ψ(t) . where the functions Φ(t) and Ψ(t) are given by Φ(t) = AeF .6. =a ∂t ∂x2 1◦ . ¦ ¤¥ ∂w =a ∂2w e−F g dt. + f (t)w ln w + g(t)x + h(t) w. ϕ1 (t). t) = exp ϕ(t)x2 + ψ(t) . © 2004 by Chapman & Hall/CRC . F = f dt.74 6. e−F (2aϕ + g) dt.9).1. D. and ϕ0 (t) are determined by a system of ordinary differential equations (see equation 1. F = f dt. D. Functional separable solution: w(x. ∂t ∂x2 1◦ . F. Reference: V. Polyanin (1996). Ψ(t) = BeF + eF e−F (aA2 e2F + g) dt. There are also functional separable solutions of the more general form w(x. t) = exp ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) . 2◦ . Polyanin (1996). and ϕ0 (t) are determined by a system of ordinary differential equations (see equation 1. Reference: V. Zaitsev and A. w(x. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w + f (t)w ln w + g(t)w. where the functions ϕ2 (t). where the functions ϕ(t) and ψ(t) are given by ϕ(t) = AeF + eF ψ(t) = BeF + eF and A and B are arbitrary constants. e−F (aϕ2 + h) dt.6. 2◦ . t) = exp ϕ(t)x + ψ(t) . t) = exp ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) . where the functions ϕ2 (t).

Given a solution of equation (1). 3) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕ2 = 4aϕ2 + f ϕ2 + g. Equations of the Form ∂w = a ∂ w + f (x. 2 1. ∂t ∂x2 Functional separable solution: w(x. ∂2w ∂w + f (x)w ln w + bf (x)t + g(x) w.1. Polyanin and Zaitsev (2003) present a large number of solutions to this equation for various f and g. the solutions of equations (2) and (3) can be constructed successively. 2 ϕ1 = 4aϕ2 ϕ1 + f ϕ1 + h.6. (1) (2) (3) © 9. Polyanin (1996).6. t) = e−bt exp ϕ(x) . F. ¨ ¤§ Reference: V. ∂w ∂t ∂x2 The substitution =a ∂2w + (bw + c) k ln2 (bw + c) + f (t) ln(bw + c) + g(t) . g = const.6. ∂w =a ∂2w ϕ0 = f ϕ0 + aϕ2 + 2aϕ2 + s. bw + c = exp u.6. t) = exp ϕ2 (t)x2 + ϕ1 (t)x + ϕ0 (t) . w) ∂t ∂x2 ∂x 1. t) ∂w + g(x. © 2004 by Chapman & Hall/CRC . ∂2w ∂w ∂w =a + (bx + c) + f (w). where the function ϕ(x) is determined by the ordinary differential equation aϕxx + a(ϕx )2 + f (x)ϕ + g(x) + b = 0. 2.2: + bku2 + bf (t)u + bg(t). 10.2. where the functions ϕn (t) (n = 1. D. ∂2u ∂u ∂u =a 2 +a ∂t ∂x ∂x 2 u = u(x. and s are not speciﬁed. 1 Here. g. the arguments of the functions f . =a ∂t ∂x2 Multiplicative separable solution: w(x. this equation can be reduced by the substitution u(ϕ) = (ϕ x )2 to a ﬁrst-order linear equation. t) leads to an equation of the form 1. and the prime denotes a derivative with respect to t. Equation (1) for ϕ2 = ϕ2 (t) is a Riccati equation. Zaitsev and A. due to the linearity of each of them in the unknown. which has exponential and sinusoidal solutions with respect to x. h. + f (t)w ln w + g(t)x2 + h(t)x + (t) w. 2 ∂t ∂x ∂x This equation governs unsteady mass transfer with a volume chemical reaction in an inhomogeneous ﬂuid ﬂow. t. so it can be reduced to a second-order linear equation. The books by Kamke (1977). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 75 8. For f .

f (t) dt. w). ∂ 2 w an ∂w ∂w =a 2 + + f (t)w ln w. the arguments of the functions f and g are not speciﬁed. ∂t ∂z © 2004 by Chapman & Hall/CRC . Then the function w1 = w(x + C1 e−bt . t + C2 ). is also a solution of the equation. 2◦ . ∂w ∂t =a ∂2w ∂x2 + f (t) ∂w ∂x + g(w). 4. z = x + ∂2w ∂w = a 2 + g(t. Suppose w(x. 3. one obtains a simpler equation ∂2w ∂w = a 2 + g(w). ∂t x ∂x ∂x This equation can be rewritten as 2. Generalized traveling-wave solution: w = w(z). ∂t ∂z which has a traveling-wave solution w = w(kz + λt). ∂t ∂x x ∂x Functional separable solution: w(x. where C1 and C2 are arbitrary constants. F = f dt. ϕe−F dt. where the function w(z) is determined by the ordinary differential equation awzz + (bz + c)wz + f (w) = 0. ψ(t) = BeF + 2a(n + 1)eF where A and B are arbitrary constants.76 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1◦ . On passing from t. t) is a solution of the equation in question. t) = exp ϕ(t)x2 + ψ(t) . where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4aϕ2 + f ϕ. we obtain ϕ(t) = eF A − 4a −1 eF dt . one obtains a simpler equation On passing from t. a ∂ ∂w ∂w = n xn + f (t)w ln w. z = x + C1 e−bt . x to the new variables t. ∂w ∂t =a ∂2w ∂x2 + f (t) ∂w ∂x + g(t. w). z = x + f (t) dt. Integrating the ﬁrst equation and then the second. ψt = 2a(n + 1)ϕ + f ψ. x to the new variables t.

ª 1◦ . ψt = h(t)ψ + aϕ + g(t)ϕ + s(t). 6.6. With equation (1) solved. Generalized traveling-wave solution: w(x. t) = exp x2 ϕ(t) + xψ(t) + χ(t) . 2 (1) (2) aϕ2 (t) + g(t)ϕ(t) + s(t) dt. where C is an arbitrary constant. h(t) dt . where the functions ϕ = ϕ(t). χt = hχ + 2aϕ + aψ 2 + gψ + s. where the functions ϕ = ϕ(t) and ψ = ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = f (t) + h(t) ϕ + p(t). ∂2w ∂w ∂w =a + xf (t) + g(t) + h(t)w ln w + x2 r(t) + xp(t) + (t) w. the solutions of equations (2) and (3) can be obtained with ease. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 77 5.1. H(t) H(t) = exp 2◦ . 2 ∂t ∂x ∂x Multiplicative separable solution: w(x. See equation 1. h. Integrating (1) and then (2).6. and r. equation (1) for ϕ = ϕ(t) is a Riccati equation. For r ≡ 0. equation (1) is a Bernoulli equation. E(t) E(t) = exp f (t) dt + h(t) dt . ∂w ∂t =a ∂2w ∂x2 + xf (t) + g(t) ∂w ∂x + h(t)w ln w + xp(t) + (t) w. we obtain (C1 and C2 are arbitrary constants) ϕ(t) = C1 E(t) + E(t) ψ(t) = C2 H(t) + H(t) p(t) dt. ψt = (4aϕ + f + h)ψ + 2gϕ + p. The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a large number of solutions to the Riccati equation for various f . © 2004 by Chapman & Hall/CRC . t) = exp Cebt + ebt e−bt h(t) dt ϕ(x). and χ = χ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4aϕ2 + (2f + h)ϕ + r. In the general case.2. so it can be reduced to a second-order linear equation. 2 ∂t ∂x ∂x Functional separable solution: w(x. since these are linear in their unknowns ψ = ψ(t) and χ = χ(t). ψ = ψ(t). and the function ϕ(t) is determined by the ordinary differential equation aϕxx + f (x)ϕx + bϕ ln ϕ + g(x)ϕ = 0. (1) (2) (3) ª 7. t) = exp xϕ(t) + ψ(t) . so it is easy to integrate.7 with r(t) = 0. ∂w ∂2w ∂w + f (x) =a + bw ln w + g(x) + h(t) w.

2 ∂t ∂x x ∂x Functional separable solution: w(x. t) dx + g(t) with g(t) being an arbitrary function. t) + t f (τ ) dτ . D.3: ∂ 2u ∂u ∂u = a 2 + bu . ∂w ∂2w ∂ ∂w + bw =a + f (x. Reference: A. t) . t). ¤¬ Reference: A. t) = − b θ ∂x where the function θ = θ(x. equation (1) for ϕ = ϕ(t) is a Riccati equation. ∂2w ∂w ∂w =a + bw + f (x. ∂t ∂x ∂x © 2004 by Chapman & Hall/CRC . t. the solution of the linear equation (2) can be obtained with ease. (1) (2) « For p ≡ 0.1. t0 z =x+b t t0 (t − τ )f (τ ) dτ . t). With equation (1) solved. Polyanin (2002). h.5.6.4. w) ∂t ∂x2 ∂x 1. 2 ∂t ∂x ∂x ∂x Solution: 2a 1 ∂θ . Equations of the Form ∂w = a ∂ w + f (x. t)w . ∂2θ ∂θ ∂θ = a 2 + f (x. leads to the Burgers equation 1. 2 ∂t ∂x ∂x The transformation w = u(z.3: w= ∂2u b ∂u =a 2 + ∂t ∂x 2 where F (x.6. t. ∂t ∂x ∂x 2. t) satisﬁes the linear equation 3. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE g(t) ∂w ∂2w ∂w + xf (t) + =a + h(t)w ln w + x2 p(t) + (t) w. Fortsyth (1906). In the general case.3. t) = ¤¬ ∂u ∂x 2 + F (x.78 8. 2 ∂t ∂x ∂x Let us substitute ∂u ∂x and then integrate the resulting equation with respect to x to arrive at an equation of the form 1. 2 1. so it is easy to integrate. so it can be reduced to a second-order linear equation. w) ∂w + g(x. f (x. w(x. ∂w ∂2w ∂w + bw =a + f (t). where t0 is any number. where the functions ϕ = ϕ(t) and ψ = ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4aϕ2 + (2f + h)ϕ + p. equation (1) is a Bernoulli equation. ψt = hψ + 2(a + g)ϕ + s. The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a large number of solutions to the Riccati equation for various f . R. and r. t) = exp ϕ(t)x2 + ψ(t) .

Then the function w1 = w x + ψ(t). ∂t ∂x ∂x2 Suppose w(x.3: ∂ 2u ∂u ∂u = a 2 + bu . where t0 is any number. ψ(t) = − f (t)ϕ(t) dt. For exact solutions of the above ordinary differential equation with various f (w) and g(w). t) is a solution of this equation. 7. is also a solution of the equation. t0 z =x+ t t0 f (τ ) dτ + b t t0 (t − τ )g(τ ) dτ . ∂w =a ∂2w ∂x2 + f (w) ∂w ∂x . where A and B are arbitrary constants. t) = exp ϕ(t)(x + C1 ) + ϕ(t) where C1 and C2 are arbitrary constants. ∂w =a ∂2w z = x + λt. ∂w + f (t)w ∂w =a ∂2w where C is an arbitrary constant. where the function w = w(z) is determined by the second-order autonomous ordinary differential equation awzz + f (w) − λ wz + g(w) = 0. 8. Remark. a = a(t). ∂t Traveling-wave solution: are deﬁned implicitly by a dw = z + B. © 2004 by Chapman & Hall/CRC .1. + g(t)w + h(t). w(x. leads to the Burgers equation 1. ∂t ∂x2 ∂x Traveling-wave solution: w = w(z). This remains true if the equation coefﬁcient a is an arbitrary function of time. ϕ(t) = C exp g(t) dt . ∂t ∂x2 ∂x Generalized traveling-wave solution: 6.1. . t) + t g(τ ) dτ . w = w(z). t + ϕ(t). see Polyanin and Zaitsev (2003). w = u(z. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 79 ∂w ∂2w ∂w + bw + f (t) =a + g(t). ∂t ∂x ∂x 5. ∂w ∂2w ∂w + f (t) ln w + g(t) =a .5. ϕ(t) = − f (t) dt + C2 −1 .6. z = x + λt aϕ(t) + g(t) dt . which can be reduced with the change of variable wz = u(w) to a ﬁrst-order equation. 2 ∂t ∂x ∂x The transformation 4. λw − F (w) + A F (w) = f (w) dw. ∂w + f (w) + g(w).

we obtain an equation of the form On passing from t. where C is an arbitrary constant.3. Then the function w1 = w(x + C1 e−bt .80 9. z = x + 1.3. ∂w ∂t =a ∂2w ∂x2 + f (w) + g(t) ∂w ∂x . Suppose w(x.6. where the function w(z) is determined by the ordinary differential equation awzz + f (w) + bz wz + h(w) = 0. 2 ∂t ∂x ∂x 1◦ . t) is a solution of this equation. where the function w(z) is determined by the ordinary differential equation awzz + f (w) + bz wz + g(w) = 0. w).6. 13. g(t) dt. © 2004 by Chapman & Hall/CRC . w) − c wξ + g(ξ. 2 ∂t ∂x ∂x 1◦ . z = x + form 1. ∂t Solution: w = w(ξ). PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w ∂w + f (w) + bx =a + g(w). 2◦ . is also a solution of the equation. ∂t ∂z ∂z 12.7: ∂2w ∂w ∂w = a 2 + f (w) . ∂t ∂z ∂z =a ∂2w ∂x2 + f (w) + g(t) ∂w ∂x + h(w). x to the new variables t. 2◦ . t) is a solution of this equation. 10. ∂w =a ∂2w ∂x2 + f (bx + ct. x to the new variables t. z = x + C1 e−bt . where C1 and C2 are arbitrary constants. t + C2 ). Suppose w(x. Generalized traveling-wave solution: w = w(z). ∂w ∂t On passing from t. ξ = bx + ct. ∂2w ∂w ∂w =a + f (w) + g(t) + bx + h(w). where the function w(ξ) is determined by the ordinary differential equation ab2 wξξ + bf (ξ. z = x + Ce−bt + e−bt ebt g(t) dt. t). is also a solution of the equation. g(t) dt. w) ∂w ∂x + g(bx + ct. w) = 0. Then the function w1 = w(x + Ce−bt . one obtains a simpler equation of the 11. Generalized traveling-wave solution: w = w(z).8: ∂2w ∂w ∂w = a 2 + f (w) + h(w).

© 2004 by Chapman & Hall/CRC . where the functions ϕ(t). ∂t ∂x a The substitution u = exp 4. χt = 2aϕ + bψ 2 + h. ∂w ∂t =a ∂2w ∂x2 +b ∂w ∂x b w a 2 + f (x. ∂w =a ∂2w +b ∂w 2 Here. t). respectively. t)u. which can be reduced. Generalized separable solution quadratic in x: =a +b w(x. t) = At + B + g(t) dt + ϕ(x). t) is a solution of this equation. Having determined ϕ. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 81 2 2 1. Then the functions w1 = w( x + C1 . + f (t)x2 + g(t)x + h(t). ψ(t).3.1.6. Equations of the Form ∂w = a ∂ w + b ∂w ∂t ∂x2 ∂x + f (x. which are linear in the unknowns ψ and χ. 2. and χ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4bϕ2 + f .6. t): ∂2u b ∂u = a 2 + f (x.3.4. A and B are arbitrary constants. are also solutions of the equation. to a second-order linear equation: b ψ ψxx + ba−2 f (x) − A ψ = 0.4. ® ∂w =a ∂2w +b ∂w 2 where C1 and C2 are arbitrary constants. w) 1. (1) (2) (3) ∂w ∂2w ∂w 2 Equation (1) for ϕ is a Riccati equation. Additive separable solution: w(x. t) + C2 ect . + f (x) + g(t). Suppose w(x. + cw + f (t). 3. ∂t ∂x2 ∂x This is a special case of equation 1. with the change of variable ϕx = a ψx . ∂t ∂x2 ∂x 1◦ . leads to a linear equation for u = u(x. one ﬁnds the solutions of equation (2) and then (3). t) = ϕ(t)x2 + ψ(t)x + χ(t). t.4. it can be easily integrated by separation of variables. ∂t ∂x2 ∂x This is a special case of equation 1. In the special case f = const.6. and the function ϕ(x) is determined by the ordinary differential equation aϕxx + b ϕx )2 + f (x) − A = 0. ψt = 4bϕψ + g.6.

w1 = w(x. ξ = x + λt. ∂t ∂x2 ∂x This is a special case of equation 1. t) = Aect + ect e−ct f (t) dt + Θ(ξ).6. where A is an arbitrary constant and the function ϕ(x) is determined by the ordinary differential equation aϕxx + b(ϕx )2 + cϕ + f (x) = 0. Generalized separable solution quadratic in x: w(x. ∂w ∂t ∂x2 ∂x Additive separable solution: =a ∂2w +b ∂w 2 + cw + f (x) + g(t). t) = Aect + ϕ(x). ∂w =a ∂2w +b ∂w 2 where A is an arbitrary constant and the function ϕ(x) is determined by the ordinary differential equation aϕxx + b(ϕx )2 + cϕ + f (x) = 0. ∂t ∂x2 ∂x 1◦ . e−ct f (t) dt 2 + cU . 7. + g(t)w + h(t). t) is a solution of this equation. where A and λ are arbitrary constants. w(x. + cw + f (x). Then the function 5.6. t) = ϕ(x) + Aect + ect e−ct g(t) dt.82 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . where C1 and C2 are arbitrary constants. 8.1 with f (t) = b. Suppose w(x. Additive separable solution: w(x. is also a solution of the equation. t + C1 ) + C2 ect . ∂t ∂x2 ∂x This is a special case of equation 1. and the function Θ(ξ) is determined by the autonomous ordinary differential equation aΘξξ + b Θξ 4◦ . t) = 3◦ . ∂w =a ∂2w +b ∂w 2 ∂w =a ∂2w +b ∂w 2 + cw2 + f (t)w + g(t). The substitution w = U (x. F. D. t) + ect leads to the simpler equation ∂2U ∂U ∂U =a 2 +b ∂t ∂x ∂x ° ¤¯ 2a ct c(x + C2 )2 e ln |C1 e−ct − 4b| + C3 ect + ect − −ct − 4b C1 e C1 e−ct f (t) dt.6. Polyanin and V.2. 2◦ . Solution: w(x. Zaitsev (2002). 2 − λΘξ + cΘ = 0. 6. © 2004 by Chapman & Hall/CRC .6. Reference: A.

where C is an arbitrary constant. t. ∂t ∂x2 ∂x ∂x There are generalized separable solutions of the form + cw w(x. + kw2 + f (t)w + g(t). t) + Cekt . 2◦ . and the function ϕ(ξ) is determined by the ordinary differential equation aϕξξ + b(ϕξ )2 + 1 ξϕξ + f (ξ) − C = 0. t) is a solution of this equation. ∂w =a ∂2w +b ∂w 2 + f (t) ∂w + cw2 + g(t)w + h(t).1. w). ∂t ∂x2 ∂x The substitution u = exp(λw) leads to the linear equation 9. ∂t ∂x ∂x 1◦ . t. ∂t ∂x2 ∂x t t ◦ 1 .2 are roots of the quadratic equation bλ2 + cλ + k = 0. Additive separable solution: w(x.5. t) = ϕ(x) + Cekt + ekt e−kt h(t) dt. w) ∂w + g(x. Equations of the Form ∂w = a ∂ 2 w + b ∂w 2 + f (x. Suppose w(x. 3. + kw + g(x) + h(t). =a + aλ ∂2u ∂u = a 2 + λf (x. Then the function 10.5. ∂w =a ∂2w ∂x2 +b ∂w 2 ∂t ∂x ∂x This is a special case of equation 1. t)e–λw . ∂w =a +b ∂w w1 = w(Cx. t) = ϕ(t) + ψ(t) exp(λx).6. t)u + λg(x. 2. is also a solution of the equation. there is also a exact solution of the form w = Cτ + ϕ(ξ). ∂t ∂x 1 x + f √ . Suppose w(x. C 2 t). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 83 + f (x. is also a solution of the equation. 2 ∂ξ which admits an exact solution of the form w = w(ξ). 2 1. In the special case f = f (ξ). where λ = λ1. 2◦ . ∂w =a ∂2w +b ∂w 2 + f (x) ∂w where the function ϕ(x) is determined by the ordinary differential equation aϕxx + b(ϕx )2 + f (x)ϕx + kϕ + g(x) = 0. t) is a solution of this equation. 3◦ . where C is an arbitrary constant. ξ = xt−1/2 leads to the equation ∂2w ∂w ∂w =a 2 +b ∂τ ∂ξ ∂ξ 2 ∂w ∂2w ∂w 2 ∂2w 2 + 1 ∂w ξ + f (ξ.6. Then the function ∂x2 w1 = w(x. The transformation τ = ln t. w) ∂t ∂x2 ∂x ∂x 1. © 2004 by Chapman & Hall/CRC ∂w .6. t). where C is an arbitrary constant.6.w . t) + g(x.

³ ¤² Reference: V. t) + h(x. so it can be readily integrated. t)e–λw . t. t). t. and χ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4f ϕ2 + gϕ. ψt = (4f ϕ + g)ψ. (4) (4f ϕ + g) dt . t)u + λh(x. χt = gχ + 2aϕ + f ψ 2 + h. t)u. ∂t ∂x ∂x The substitution u = exp(λw) leads to the linear equation =a ∂x2 + aλ + f (x. we can obtain the solutions of equation (2) and then (3). t. ∂w =a ∂2w + f (t) ∂w 2 2 + g(x. F. A2 . Zaitsev and A. + g(x. where the functions ϕ(t). ψ(t). w) ∂x + g(t)w + h(t). Then the functions w1 = w( x + C1 . ∂t ∂x2 ∂x 1◦ . t) ∂ 2u ∂u ∂u = a 2 + f (x. t) + g(x. t) + C2 exp ± g(t) dt . Generalized separable solution quadratic in x: w(x. which are linear in the unknowns ψ and χ. 2◦ . Equations of the Form ∂w = a ∂ 2 w + f (x. w) ∂w + h(x. D. G= g dt. ∂t ∂x ∂x a 5. t) + λg(x. and A3 are arbitrary constants. t) is a solution of this equation.84 4.6. Having determined ϕ. Finally. respectively. ∂t ∂x ∂x ∂w ∂2w ∂w 2 ∂w 1. © 2004 by Chapman & Hall/CRC . ∂w ∂t =a ∂2w ∂x2 +b PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w 2 ∂x b w The substitution u = exp a + f (x. where C1 and C2 are arbitrary constants. t) ∂w ∂x + g(x. e−G (2aϕ + f ψ 2 + h) dt. Polyanin (1996). are also solutions of the equation. χ = A 3 eG + eG where A1 . we have ϕ = e G A1 − 4 ψ = A2 exp −1 eG f dt . t). Suppose w(x. (1) (2) (3) Equation (1) is a Bernoulli equation for ϕ. t) = ϕ(t)x2 + ψ(t)x + χ(t). w) ∂w ∂t ∂x2 ∂x 1. leads to the linear equation ∂2u ∂u b ∂u = a 2 + f (x. To the limit passage A1 → ∞ in (4) there corresponds a degenerate solution with ϕ ≡ 0.6.

t) = ϕ(t) + ψ(t) cosh x −b . Note two special cases of solution (6) where the exponentials combine to form hyperbolic functions: √ 1 1 w(x. If the functions f .6. and h. hence. b < 0. © 2004 by Chapman & Hall/CRC . (6) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = bf ϕ2 + 4ABψ 2 + gϕ + h. and h are proportional. β = const). Generalized separable solution quadratic in : √ w(x. g. B = −2. ∂t ∂x2 ∂x 1◦ . where C1 is an arbitrary constant. The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a large number of solutions to this equation for various f . (2) (3) Equation (2) is a Riccati equation for ϕ = ϕ(t). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 85 + bf (t)w2 + g(t)w + h(t). ψt = (2bf ϕ + g − ab)ψ. where C2 is an arbitrary constant. t) = ϕ(t) + ψ(t) A exp x −b + B exp −x −b . then the solution of (2) is given by dϕ = bϕ2 + αϕ + β f dt + C2 . (7) (8) Solving (8) for ϕ to express it in terms of ψ and then substituting the resulting expression into (7). one arrives at a second-order nonlinear equation for ψ (if f . 2◦ . =a + f (t) ´ ∂w ∂2w ∂w 2 b < 0. g. A = 2 . g. 2. Note two special integrable cases of equation (2). t) = ϕ(t) + ψ(t) sinh x −b . Given a solution of (2). its order can be reduced). h = const. ψt = 2bf ϕψ + gψ − abψ. (1) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients (the arguments of f . On integrating the left-hand side of equation (5) and solving for ϕ. t) = ϕ(t) + ψ(t) exp x −b . and h are not indicated) ϕt = bf ϕ2 + gϕ + h. this equation is autonomous and.1. g = αf . Generalized separable solution of a more general form: √ √ w(x. Solution of equation (2) with h ≡ 0: ϕ(t) = eG C1 − b −1 (2bf ϕ + g) dt . one can ﬁnd ϕ = ϕ(t) in explicit form. B = 2 . g. w(x. one can solve the linear equation (3) for ψ = ψ(t) to obtain ψ(t) = C exp −abt + where C is an arbitrary constant. so it can be reduced to a second-order linear equation. G= g dt. (4) f eG dt . √ 1 1 A = 2. (5) h = βf (α.

ψt = 2bf ϕψ + gψ − abψ. A. + bw + h(x) + p(t). ∂t ∂x ∂x 1◦ . 2◦ .6. t) = ϕ(t)x2 + ψ(t)x + χ(t).86 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 3◦ . Posashkov (1989. t) = ϕ(t) + ψ(t) exp(λx). where C is an arbitrary constant. =a ∂x2 + f (t) + bf (t)w w(x. Suppose w(x. ¶ ¤µ References: V. ∂w =a ∂2w + f (x) ∂w 2 ∂w ∂2w ∂w 2 ∂w + g(x) ∂w where the function ϕ(x) is determined by the ordinary differential equation aϕxx + f (x)(ϕx )2 + g(x)ϕx + bϕ + h(x) = 0. Galaktionov and S. t) + Cebt . one arrives at a second-order nonlinear equation for ψ (if f . A. one arrives at an equation of the form + bf (t)w2 + h(t)w + p(t). g. is also a solution of the equation. g.2 are roots of the quadratic equation λ2 + bλ + c = 0. (10) (11) Solving (11) for ϕ to express it in terms of ψ and then substituting the resulting expression into (10). © 2004 by Chapman & Hall/CRC · 6. Additive separable solution: w(x. where λ = λ1.2: ∂2w ∂w ∂w = a 2 + f (t) ∂t ∂z ∂z =a ∂2w + f (t) ∂w 2 2 ∂w + g1 (t)x + g0 (t) + h(t)w + p(t)x2 + q(t)x + (t). h = const was considered). On passing from t.6. h = const. this equation is autonomous and. 4. Polyanin (1996). hence. Zaitsev and A. x to the new variables t. the case f . ∂w ∂t =a ∂2w ∂x2 + f (t) ∂w ∂x 2 + g(t) ∂w ∂x + bf (t)w2 + h(t)w + p(t). Then the function ∂x2 w1 = w(x. D. 5. ∂w . Generalized separable solution (c is an arbitrary constant): √ w(x. t) = ϕ(x) + Cebt + ebt e−bt p(t) dt. b > 0. t) is a solution of this equation. V. ∂t ∂x ∂x There are generalized separable solutions of the form 3. z = x + 1. + cf (t)w2 + g(t)w + h(t). F. g(t) dt. ∂t ∂x2 ∂x ∂x Generalized separable solution quadratic in x: w(x. its order can be reduced). t) = ϕ(t) + ψ(t) cos x b + c . (9) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = bf ϕ2 + ψ 2 + gϕ + h.

10. can be reduced to a second-order linear equation. u= F (w) dw. where F (w) = exp f (w) dw . t) ∂w ∂x + g(x. ∂t ∂x2 9. © 2004 by Chapman & Hall/CRC . see Kamke (1977) and Polyanin and Zaitsev (2003). t): ∂u ∂u ∂ 2 u = + g(x) . ∂t ∂x2 ∂x Some exact solutions of this equation. ∂t ∂x ∂x 8. where F (w) = exp f (w) dw . Given a solution of (1). t) + (1 − k)g(x. and χ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4f ϕ2 + (2g1 + h)ϕ + p. t): ∂u ∂u ∂ 2 u = + xg(t) + h(t) . ψ(t). 2 ∂t ∂x ∂x This equation can be reduced to the linear heat equation (see Polyanin. ∂w = ∂2w ∂x2 + f (w) ∂w ∂x 2 ∂t The substitution + xg(t) + h(t) ∂w ∂x . ∂2w ∂w ∂w = + f (w) ∂t ∂x2 ∂x The substitution u= 2 + g(x) ∂w . can be found in Polyanin (2002).6. (1) (2) (3) Equation (1) is a Riccati equation for ϕ = ϕ(t) and. t)w k . (1) is a Bernoulli equation. 2002). ∂w ∂t =a ∂2w ∂x2 – ak 1 w ∂w ∂x 2 + f (x. The substitution u = w 1−k leads to the linear equation ∂2u ∂u ∂u = a 2 + f (x. t)w + h(x. leads to a linear equation for u = u(x. ψt = (4f ϕ + g1 + h)ψ + 2g0 ϕ + q. F (w) dw. In the special case p ≡ 0. ∂w ∂t ∂x2 The substitution = ∂2w + f (w) ∂w ∂x 2 . For solution of such equations. hence.1. since these are linear in their respective unknowns ψ and χ. t)u + (1 − k)h(x. so it can be readily integrated. for arbitrary g. u= F (w) dw. t). equations (2) and (3) can be easily solved. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 87 where the functions ϕ(t). t): ∂u ∂ 2 u = . leads to the linear heat equation for u = u(x. ∂x where F (w) = exp f (w) dw . 7. χt = hχ + 2aϕ + f ψ 2 + g0 ψ + s. leads to a linear equation for u = u(x.

and h(w). ∂t ∂x2 ∂x f (w) dw . hence. Then the function ∂x2 w1 = w(x. ϕx ) + bϕ = 0. 2◦ . + bw + g(t). ∂w =a ∂2w + f x. ∂w = ∂2w ∂x2 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE + f (w) ∂w ∂x 2 ∂t The substitution + g(x. Additive separable solution: w(x. w. Note that in the special case h ≡ 0. z = x + λt. t) + Cebt . (1) 1. g(w). is also a solution of the equation. is also a solution of the equation. Additive separable solution: w(x. (2) For exact solutions of the ordinary differential equations (1) and (2) for various f (w).6. Equations of the Form ∂w = a ∂ w + f x. t) ∂w ∂x . Suppose w(x. where C is an arbitrary constant. ∂t ∂x 1◦ .88 11. see the book by Polyanin and Zaitsev (2003). equation (2) becomes linear and. t) = ϕ(x) + Cebt + ebt e−bt g(t) dt. t) = ϕ(x) + Cebt + ebt e−bt p(t) dt. 12. where F (w) = exp ∂u ∂u ∂ 2 u = + g(x. where C is an arbitrary constant. t) is a solution of this equation. The substitution wz = u(w) leads to the ﬁrst-order equation auuw + f (w)u2 + [g(w) − λ]u + h(w) = 0. t) . where the function ϕ(x) is determined by the second-order ordinary differential equation aϕxx + f (x)(ϕx )k + g(x)ϕx + bϕ + h(x) = 0. 2. 2 ∂t ∂x ∂x ∂x 1◦ . where the function w = w(z) is determined by the autonomous ordinary differential equation awzz + f (w)(wz )2 + [g(w) − λ]wz + h(w) = 0. © 2004 by Chapman & Hall/CRC . ∂w ∂t ∂x2 ∂x ∂t ∂x2 Traveling-wave solution: =a ∂2w + f (w) ∂w ∂x 2 + g(w) ∂w ∂x + h(w). Then the function w1 = w(x. can be readily integrated. ∂w w = w(z). ∂w where the function ϕ(x) is determined by the second-order ordinary differential equation aϕxx + f (x. u= leads to the linear equation F (w) dw. 2◦ .7. t) + Cebt . Suppose w(x. 2 1. t. ∂2w ∂w k ∂w ∂w =a + f (x) + g(x) + bw + h(x) + p(t). t) is a solution of this equation.

∂w ∂t ∂x ∂x2 f (t. .6. λ) dt . ∂t ∂x x ∂x Functional separable solution: w(x. where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4f ϕ2 + gϕ. Equations of the Form ∂w = f (x. G= g dt. The substitution z = dx √ leads to an equation of the form 1. This equation can be rewritten in the form ∂ 2 w nf (t) ∂w ∂w = f (t) 2 + + g(t)w ln w. Multiplicative separable solution: w(x. where C1 and C2 are arbitrary constants.6. t. one obtains ϕ(t) = eG A − 4 −1 f eG dt .8. © 2004 by Chapman & Hall/CRC f ϕe−G dt.6. 2◦ . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 89 3. the arguments of f and g are omitted. t) = A exp λx + aλ2 t + where A and λ are arbitrary constants. ψ(t) = BeG + 2(n + 1)eG where A and B are arbitrary constants. 1. t) = exp ϕ(t)x2 + ψ(t) . t) is a solution of this equation. Then the function w1 = C1 w(x + C2 . ψt = 2(n + 1)f ϕ + gψ. w. 2 1. ∂t ∂z 2 2. is also a solution of the equation. . f (t) ∂ ∂w ∂w = n xn ∂t x ∂x ∂x + g(t)w ln w. t) ∂ w + g x. 1 ∂w ∂2w ∂w + wf t.1: ax2 + b ∂w ∂ 2 w = + f (w). Successively integrating. Suppose w(x. =a 2 ∂t ∂x w ∂x 1◦ . ∂w ∂t = (ax2 + b) ∂2w ∂x2 + ax ∂w ∂x + f (w).1.1. t).

ψt = g(t)ψ + h(t). w(x. + aw ln w. Suppose w(x. 2 ∂t ∂x x ∂x Functional separable solution: w(x. ∂t ∂x ∂x 1◦ . The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a considerable number of solutions to this equation for various f . t) = exp ϕ(t)x2 + ψ(t) . and p. Having solved equation (1). can be easily integrated. where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients ϕt = 4f ϕ2 + (2g + s)ϕ + p. t) = exp ϕ(t)e−λx + ψ(t) . and the function ϕ(t) is determined by the ordinary differential equation (f ϕx )x + aϕ ln ϕ = 0. (1) (2) ¸ For p ≡ 0. s. g. equation (1) is a Bernoulli equation and. ∂2w ∂w ∂w = f (x) 2 + fx (x) + aw ln w. (1) is a Riccati equation for ϕ = ϕ(t). 4. where C is an arbitrary constant. G(t) = exp g(t) dt . G(t) . Multiplicative separable solution: w(x. 5. 2◦ . PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE h(t) ∂w ∂2w ∂w + xg(t) + = f (t) + (t)w ln w + x2 p(t) + q(t) w. ∂t ∂x ∂x This equation can be rewritten in the form ∂w = ∂ f (x) ∂w −1 f (t)G(t) dt h(t) dt. Then the function w1 = exp(C1 eat ) w(x. ∂w ∂ ∂w eλx = f (t) ∂t ∂x ∂x Functional separable solution: + g(t)w ln w + h(t)w. is also a solution of the equation. hence. © 2004 by Chapman & Hall/CRC .90 3. where C1 and C2 are arbitrary constants. t + C2 ). t) = exp(Ceat ) ϕ(x). one can ﬁnd ψ = ψ(t) from the linear equation (2). so it can be reduced to a second-order linear equation. where the functions ϕ(t) and ψ(t) are determined by the ordinary differential equations ϕt = λ2 f (t)ϕ2 + g(t)ϕ. In the general case. Integrating yields ϕ(t) = G(t) A − λ2 ψ(t) = BG(t) + G(t) where A and B are arbitrary constants. t) is a solution of the equation in question. ψt = sψ + 2(f + h)ϕ + q.

Then the function w1 = exp(Ceat )w(x. t) + Ceat . 2◦ . is also a solution of the equation. t). 8. Then the function ∂x2 w1 = w(x. ∂w = f (x) ∂2w + g(x) ∂w k + h(x) ∂w where the function ϕ(x) is determined by the ordinary differential equation f (x)ϕxx + g(x)(ϕx )k + h(x)ϕx + aϕ + p(x) = 0. t) = exp Ceat + eat e−at h(t) dt ϕ(x). ∂t ∂x ∂x 1◦ . t) = ϕ(x) + Ceat + eat e−at q(t) dt. where the function ϕ(x) is determined by the ordinary differential equation (f ϕx )x + aϕ ln ϕ + g(x)ϕ = 0. t). is also a solution of the equation. ∂t ∂x ∂x 1◦ . + aw + p(x) + q(t). t) is a solution of this equation. © 2004 by Chapman & Hall/CRC . is also a solution of the equation. ∂w + g(x) + aw ln w + h(x) + (t) w. 2◦ . where C is an arbitrary constant. Additive separable solution: w(x. Additive separable solution: w(x. + aw + p(x) + q(t). where C is an arbitrary constant. t) + Ceat . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 91 ∂ ∂w = f (x) + aw ln w + g(x) + h(t) w. ∂t ∂x2 ∂x 1◦ . ∂t ∂x ∂x 1◦ . Then the function ∂x2 w1 = w(x. Suppose w(x. 2◦ . where C is an arbitrary constant. where C is an arbitrary constant. ¹ 7. 2◦ . ∂w = f (x) ∂2w + g(x) ∂w 2 + h(x) ∂w where the function ϕ(x) is determined by the ordinary differential equation f (x)ϕxx + g(x)(ϕx )2 + h(x)ϕx + aϕ + p(x) = 0. t) is a solution of this equation. Multiplicative separable solution: w(x. t) = ϕ(x) + Ceat + eat e−at q(t) dt. Suppose w(x. Then the function 6. t) is a solution of this equation. 9.1. t) = exp Ceat + eat e−at s(t) dt ϕ(x).6. t) is a solution of this equation. ∂w = f (x) ∂2w where the function ϕ(x) is determined by the ordinary differential equation f (x)ϕxx + g(x)ϕx + aϕ ln ϕ + h(x)ϕ = 0. ∂w w1 = exp(Ceat )w(x. is also a solution of the equation. Suppose w(x. Multiplicative separable solution: w(x. Suppose w(x.

t. Additive separable solution: w(x. is also a solution of the equation. + aw + h(t). t. B. where A. 2◦ . F (t) = exp where A. t) = ϕ(t)(x2 + Ax + B) + ϕ(t) ϕ(t) = F (t) C − 2a ∂2w −1 g(t) dt. Suppose w(x. 3. ϕ(t) f (t) dt . t) = F (t)(Ax + B) + F (t) g(t) dt. F (t) dt . ∂t ∂x2 Generalized separable solution: w(x.9. w) ∂t ∂x2 ∂x 1. + f (x)w + bx + c. 2 1. B. and C are arbitrary constants. 2◦ . Generalized separable solution quadratic in x: w(x.92 10. where A and B are arbitrary constants. PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w ∂w + g x.6. Then the function w1 = w(x.6. where the function ϕ(x) is determined by the ordinary differential equation f (x)ϕxx + g(x. B. = aw 4. + f (t)w + g(t). and x0 are arbitrary constants. t) = t b1 sinh(kx) + b2 cosh(kx) + ϕ(x). Degenerate solution linear in x: w(x. ∂t ∂x2 Generalized separable solution: w(x. where C is an arbitrary constant.1 with b = 0. Here. t) is a solution of this equation. ϕx ) + aϕ = 0. © 2004 by Chapman & Hall/CRC . ∂t ∂x2 1◦ . t) + Ceat . whose general solution is given by x 1 f (ξ) sinh k(x − ξ) dξ. ϕ(x) = C1 sinh(kx) + C2 cosh(kx) − ak x0 where A. 2. Equations of the Form ∂w = aw ∂ w + f (x. t) = (bx + c)t + Ax + B − ∂w = aw ∂2w 1 a x x0 (x − ξ)f (ξ) dξ. w) ∂w + g(x. = f (x) 2 ∂t ∂x ∂x 1◦ . t) = ϕ(x) + Ceat + eat e−at h(t) dt.10. the function ϕ(x) is determined by the linear nonhomogeneous ordinary differential equation with constant coefﬁcients aϕxx − ak 2 ϕ + f (x) = 0. and x0 are arbitrary constants. ∂w ∂2w ∂w = aw – ak2 w2 + f (x)w + b1 sinh(kx) + b2 cosh(kx). F (t) ∂w = aw ∂2w F (t) = exp f (t) dt . ∂t ∂x2 This is a special case of equation 1. + cw2 + f (t)w + g(t).

H(t) = exp h(t) dt . g(t)F (t) dt. the function ϕ(x) is determined by the linear nonhomogeneous ordinary differential equation with constant coefﬁcients aϕxx + ak 2 ϕ + f (x) = 0. ∂t ∂x2 The transformation w(x. where the functions F (t) and G(t) are given by F (t) = exp f (t) dt . leads to a simpler equation of the form 1. τ ). leads to a simpler equation of the form 1.1: ∂2u ∂u = au 2 . ϕ(x) = C1 sin(kx) + C2 cos(kx) − ak x0 where A. and x0 are arbitrary constants. ∂w ∂t ∂x2 The transformation = aw ∂2w + xf (t) + g(t) ∂w ∂x + h(t)w. t) = G(t)u(z.6. τ ). B. z = xF (t) + where the functions F (t) and H(t) are given by F (t) = exp f (t) dt . t) = t b1 sin(kx) + b2 cos(kx) + ϕ(x). ∂2w ∂w ∂w = aw + f (t) + g(t)w + h(t). ∂2w ∂w + ak2 w2 + f (x)w + b1 sin(kx) + b2 cos(kx).1. 8.5. ∂τ ∂z 9. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 93 5. 2 ∂t ∂x ∂x This is a special case of equation 1. z =x+ f (t) dt. w(x. = aw ∂t ∂x2 Generalized separable solution: w(x. ∂τ ∂z © 2004 by Chapman & Hall/CRC . G(t) = exp g(t) dt . τ= F 2 (t)H(t) dt.1.1: ∂2u ∂u = au 2 . 2 ∂t ∂x ∂x The transformation w(x. Here.9. whose general solution is given by x 1 f (ξ) sin k(x − ξ) dξ.1: ∂2u ∂u = au 2 . ∂w = aw ∂2w + xf (t) ∂w ∂x + g(t)w. ∂τ ∂z 7. G(t) = exp g(t) dt leads to a simpler equation of the form 1. t) = G(t)u(z. 6. t) = H(t)u(z. ∂2w ∂w ∂w = aw + f (t) + g(t)w.6. τ ). τ= F 2 (t)G(t) dt.9. z = xF (t). τ= G(t) dt.1.10.1.9.

Here. w) ∂w ∂t ∂x2 ∂x 1. t) = ϕ(t)Θ(x) + ψ(t). w) ∂x + cw2 + f (t)w + g(t). Generalized separable solution quadratic in involving an exponential of x: w(x. ∂w = aw ∂2w ∂x2 + f (x)w ∂w + g(x)w2 + h(t)w. F (x) = exp F (x) a f (x) dx . and B2 are arbitrary constants. t. where the functions ϕ(t). aΘxx + f (x)Θx = C. B1 . © 2004 by Chapman & Hall/CRC 2 (2) (3) . one obtains ϕ(t) = G(t) A1 − C ψ(t) = A2 ϕ(t) + ϕ(t) Θ(x) = B1 −1 G(t) dt . see Kamke (1977) and Polyanin and Zaitsev (2003). ψ(t). For exact solutions of this equation with various f (x) and g(x). º λ= −c a+b 1/2 . 2 ∂t ∂x ∂x Generalized separable solution: w(x. H(t) = exp h(t) dt . PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w ∂w + f (x)w = aw + g(t)w + h(t). 11. and Θ(x) are determined by the system of ordinary differential equations ϕt = Cϕ2 + g(t)ϕ. h(t) dt. (1) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients (the arguments of f and g are not indicated) ϕt = cϕ2 + f ϕ + g. ψt = (aλ ϕ + 2cϕ + f )ψ. ∂w = aw ∂2w +b ∂w 2 2 + g(x.6. ψt = Cϕ + g(t) ψ + h(t). 1. t. and the function ϕ(x) is determined by the second-order linear ordinary differential equation aϕxx + f (x)ϕx + g(x)ϕ = B. t) = ϕ(x)H(t) A − B H(t) dt .10. Equations of the Form ∂w = (aw + b) ∂ 2 w + f (x. −1 ∂t ∂x Multiplicative separable solution: w(x. ϕ(t) C dx + B2 + F (x) a F (x) dx dx 1 .94 10. t. w) ∂w + h(x. where A1 . Integrating successively. where C is an arbitrary constant. A2 . ∂t ∂x2 ∂x 1◦ . t) = ϕ(t) + ψ(t) exp( λx). G(t) = exp g(t) dt . A and B are arbitrary constants.

its order can be reduced. Generalized separable solution involving a trigonometric function (A is an arbitrary constant): w(x. t) = ϕ(t) + ψ(t) cos(λx + A). the solution of equation (3) can be obtained in the form ψ(t) = C exp where C is an arbitrary constant. λ= −c a+b 1/2 (aλ2 ϕ + 2cϕ + f ) dt . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 95 Equation (2) is a Riccati equation for ϕ = ϕ(t). 0 Given a solution of (2). g = const and. (4) .6. t) = ϕ(t) + ψ(t) cosh(λx + A). 3◦ . F. V. g = const and. The substitution u = ϕ − ϕ0 leads to a Bernoulli equation. Polyanin (1996). (5) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients (the arguments of f and g are not speciﬁed) ϕt = cϕ2 − bλ2 ψ 2 + f ϕ + g. hence. for g ≡ 0. (9) (10) Solving equation (10) for ϕ to express it in terms of ψ and substituting the resulting expression into (9). The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a large number of solutions to this equation for various f and g. equation (2) is a Bernoulli equation. f . © 2004 by Chapman & Hall/CRC . ψt = (aλ2 ϕ + 2cϕ + f )ψ. In another special case. t) = ϕ(t) + ψ(t) sinh(λx + A). so it can be reduced to a second-order linear equation. where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = cϕ2 + bλ2 ψ 2 + f ϕ + g. ψt = (aλ2 ϕ + 2cϕ + f )ψ. 4◦ . one arrives at a second-order nonlinear equation for ψ. Generalized separable solution involving hyperbolic cosine (A is an arbitrary constant): w(x. ¼ ¤» References: V. (8) where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = cϕ2 + bλ2 ψ 2 + f ϕ + g. Zaitsev and A. which is a root of the quadratic equation cϕ2 + f ϕ0 + g = 0. which is autonomous if f . A. its order can be reduced. λ= −c a+b 1/2 . Galaktionov (1995). which is easy to integrate. ϕ = ϕ 0 . which is autonomous if f . one arrives at a second-order nonlinear equation for ψ. D. g = const.1. In particular. 2◦ . a particular solution of (2) is a number. ψt = (−aλ2 ϕ + 2cϕ + f )ψ. λ= c a+b 1/2 . (6) (7) Solving equation (7) for ϕ to express it in terms of ψ and then substituting the resulting expression into (6). Generalized separable solution involving hyperbolic sine (A is an arbitrary constant): w(x. hence.

V.10. so it is easy to integrate. D. 4.2 for u = u(z. P (t) = exp where A is an arbitrary constant. ∂w = aw ∂2w ∂x2 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE + f (t) ∂w 2 ∂t ∂x ∂x Generalized separable solution quadratic in x: w(x. w(x. equation (1) can be reduced. t) = w(x. and χ(t) are determined by a system of ﬁrst-order ordinary differential equations with variable coefﬁcients. Equation (1) is a Bernoulli equation for ϕ = ϕ(t). since both are linear in their respective unknowns ψ and χ. and χ(t) are determined by the system of ﬁrst-order ordinary differential equations with variable coefﬁcients (the arguments of f . where the functions ϕ(t). ∂w = (aw + b) ∂2w ∂x2 +c ∂w ∂x 2 b . to a second-order linear equation. ∂t ∂x2 ∂x ∂x There is a generalized separable solution quadratic in x: w(x. equation (2) and then (3) can be solved with ease. ∂w = aw ∂2w + f (x) ∂w 2 + g(x)w ∂w ∂x + h(x)w2 + p(t)w. The general solution of equation (2) is given by ψ(t) = P (t) A − C −1 2 P (t) dt . where the functions ϕ(t). u(z. © 2004 by Chapman & Hall/CRC ½ + g(t) ∂w + h(t)w + (t). t) = ϕ(t)x2 + ψ(t)x + χ(t). ∂t ∂x2 ∂x Multiplicative separable solution: 3. F. ¿ ¤¾ References: V. After that. on dividing it by ϕ.6. In the special case f ≡ 0.96 2. h. see Kamke (1997) and Polyanin and Zaitsev (2003). t) + 5. ψ(t). ψt = (4f ϕ + 2aϕ + h)ψ + 2gϕ. Zaitsev and A. (1) (2) p(t) dt . t). A. Polyanin (1996). Galaktionov (1995). . ψt = Cψ + p(t)ψ. g. ∂t The transformation + f (t) ∂w ∂x + kw2 + g(t)w + h(t). (1) (2) (3) χt = (2aϕ + h)χ + f ψ 2 + gψ + s. z =x+ a leads to an equation of the form 1. where ϕ(x) and ψ(t) are determined by the following system of ordinary differential equations (C is an arbitrary constant): aϕϕxx + f (x)(ϕx )2 + g(x)ϕϕx + h(x)ϕ2 = Cϕ. t) = ϕ(t)x2 + ψ(t)x + χ(t). t) = ϕ(x)ψ(t). f (t) dt ∂2w ∂w 2 ∂w ∂w = (aw +b) +f (t) + g1 (t)x+g0 (x) +h(t)w +p2 (t)x2 +p1 (t)x+p0 (t). which is not written out here. ψ(t). for exact solutions of this equation with various g(x) and h(x). and s are not speciﬁed) ϕt = 2(2f + a)ϕ2 + hϕ.

(2) Given a particular solution. t) = F (t)u(x.10. t) = (4λt + C)−1/4 g(x). τ ). Á ¤À (1) Then the transformation dx u2 2 . ∂t ∂ξ Using the change of variable v = z −3 . D. C1 t + C2 ). ∂t ∂x2 1◦ .11. Zaitsev and A. Suppose w(x.4: ∂v ∂ ∂v v −4/3 . where C1 and C2 are arbitrary constants. The transformation w(x. ∂t ∂x2 1◦ . z= u2 u simpliﬁes the original equation bringing it to the form ξ= ∂2z ∂z = az 4 2 . Reference: V. we obtain an equation of the form 1. Polyanin (1996). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 97 2 1. t) ∂w + g(x. t) is a solution of this equation. are also solutions of the equation.1. ∂w = awm ∂2w where C1 and C2 are arbitrary constants. where C and λ are arbitrary constants. t. F. τ= F m (t) dt. Suppose u = u(x) is a nontrivial solution of the second-order linear ordinary differential equation auxx + f (x)u = 0. is also a solution of the equation.g. see Polyanin and Zaitsev. Then the function 4 w1 = C1 w(x.. t). 2003) λ Ag 2 = − u2 + u2 B + A a where A and B are arbitrary constants (A ≠ 0). Multiplicative separable solution: w(x. Equations of the Form ∂w = aw m ∂ w + f (x. 2◦ . u = u(x). w dx . w) ∂t ∂x2 ∂x 1. ∂τ ∂x © 2004 by Chapman & Hall/CRC Â . Suppose w(x. m −2 w1 = C1 w( C1 x + C2 . leads to a simpler equation of the form 1. t) is a solution of this equation. and the function g = g(x) is determined by Yermakov’s equation agxx + f (x)g + λg −3 = 0. F (t) = exp f (t) dt .6. the general solution of the nonlinear equation (2) can be expressed as (e. ∂2w ∂w = aw4 + f (x)w5 .18: ∂2u ∂u = aum 2 .6.1. =a ∂t ∂ξ ∂ξ 3◦ .1. of the linear equation (1).9. Then the functions 2. 2◦ . + f (t)w.

6a(t + C1 ) where C1 .98 3.9.9. Polyanin (1996). τ )H(t). is also a solution of the equation. D. Equations of the Form ∂w =a ∂ w ∂w +f (x. 2◦ . ∂w = awm ∂2w + f (t)x + g(t) ∂w ∂x + h(t)w. τ= F 2 (t)H m (t) dt. (x + C2 )2 + C3 (t + C1 )−1/3 + (t + C1 )−1/3 (t + C1 )1/3 f (t) dt. G(t) = exp g(t) dt . t) is a solution of this equation. Zaitsev and A. ∂t ∂x2 The transformation 4. τ= F 2 (t)Gm (t) dt.6. τ )G(t).18: ∂2u ∂u = aum 2 . ∂t ∂x ∂x Generalized separable solutions linear and quadratic in x: w 2 w(x. t. where C1 and C2 are arbitrary constants. t) = − . Suppose w(x. ∂τ ∂z 1. leads to a simpler equation of the form 1. where the functions F (t) and H(t) are given by F (t) = exp f (t) dt . and C3 are arbitrary constants. F. t) = u(z. w(x. ∂w = awm ∂2w + xf (t) ∂w ∂x + g(t)w. © 2004 by Chapman & Hall/CRC Ä ¤Ã ∂t ∂x2 The transformation 5. where the function ϕ = ϕ(x) is determined by the ordinary differential equation aϕm ϕxx + f (x)ϕm+1 + λϕ = 0. C1 t + C2 ).1. z = xF (t) + g(t)F (t) dt. C2 .18: ∂2u ∂u = aum 2 . Multiplicative separable solution (C and λ are arbitrary constants): w(x. t) = C1 x + aC1 t + C2 + ∂w =a ∂ ∂w f (t) dt. t) = u(z. leads to a simpler equation of the form 1. H(t) = exp h(t) dt . w(x.1. w) ∂t ∂x ∂x ∂x 1. ∂τ ∂z Reference: V. The ﬁrst solution is degenerate. z = xF (t). = awm ∂t ∂x2 1◦ . + f (t). PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w + f (x)wm+1 .12. t) ∂w +g(x. where the functions F (t) and G(t) are given by F (t) = exp f (t) dt . w(x. t) = (mλt + C)−1/m ϕ(x). Then the function m w1 = C1 w(x.

∂w 4. ∂t ∂x ∂x This is a special case of equation 1.6.1. ∂t ∂z ∂z 6. ∂w ∂ ∂w leads to a simpler equation of the form 1.13. ∂ ∂w ∂w ∂w =a w + f (t) + g(t)w. x) = u(z.13. t) = G(t)u(z. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 99 ∂ ∂w =a w + f (t)w + g(t). ∂x z = xF (t). Multiplicative separable solution: w = (λt + C)−1 ϕ(x). G(t) = exp g(t) dt . ∂t ∂x ∂x ∂x This is a special case of equation 1.1: ∂ ∂u ∂u =a u . 2◦ . 2. is also a solution of the equation.13. © 2004 by Chapman & Hall/CRC . τ= F 2 (t)G(t) dt.8 with m = 1.10. ∂ ∂w ∂w ∂w =a w + xf (t) + g(t) + h(t)w. C1 t + C2 ). w(x. Suppose w(x.13. + f (x)w2 .6. where the functions F (t) and G(t) are given by F (t) = exp f (t) dt .1.6. τ= G(t) dt. where C1 and C2 are arbitrary constants. and the function ϕ(x) is determined by the ordinary differential equation a(ϕϕx )x + f (x)ϕ2 + λϕ = 0.10 with m = 1. ∂w ∂ ∂w =a w + bw2 + f (t)w + g(t). ∂t ∂x ∂x 1◦ .4 with m = 1. t) is a solution of this equation. z =x+ f (t) dt. ∂τ ∂z ∂z 7.5 with m = 1. ∂t ∂x ∂x This is a special case of equation 1.1. Then the function =a w w1 = C1 w(x. w(t. G(t) = exp g(t) dt . τ ).10. τ )G(t). The transformation 5. ∂t ∂x ∂x ∂x This is a special case of equation 1.6.1: ∂ ∂u ∂u =a u . where λ and C are arbitrary constants. ∂ ∂w ∂w =a w ∂t ∂x ∂x The transformation + xf (t) ∂w + g(t)w.6. leads to a simpler equation of the form 1. 3.

where C1 and C2 are arbitrary constants.10. A. Suppose u = u(x) is any nontrivial particular solution of the second-order linear ordinary differential equation auxx − 1 f (x)u = 0. t) = u(x.1.7: ∂ ∂u ∂u =a um . D. F. 3 ∂t ∂x 3◦ .10. t. ∂t ∂x ∂x The substitution u = w m leads to an equation of the form 1. see 1. Zaitsev and A.1. w) ∂t ∂x ∂x ∂x 1.2: =a wm ∂2u a ∂u = au 2 + ∂t ∂x m ∂u ∂x 2 ∂w ∂ ∂w + mf (t).6.10.6. C1 −4/3 t + C2 ).1. ∂t ∂x ∂x The substitution u = w m leads to an equation of the form 1. ∂t ∂x ∂x 1◦ . 3 The transformation dx . which admits a generalized separable solution of the form u = ϕ(t)x2 + ψ(t)x + χ(t).4: ∂z ∂ ∂z z −4/3 . + f (t)w1–m .11.6. bringing it to equation 1. Galaktionov (1995).2: ∂w ∂2u a ∂u = au 2 + ∂t ∂x m Ç ¤Æ ∂u ∂x 2 + mf (t)u + mg(t).2 or 1.10. D. References: V. Suppose w(x. Zaitsev and A.1: ∂2v ∂v = av 4 2 − 1 f (x)v 5 . 4. F.100 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1. Equations of the Form ∂w = a ∂ w m ∂w +f (x.13. ∂ ∂w =a wm + f (t)w + g(t)w1–m . is also a solution of the equation. which admits a generalized separable solution of the form u = ϕ(t)x2 + ψ(t)x + χ(t). © 2004 by Chapman & Hall/CRC . ∂ ∂w =a wm ∂t ∂x ∂x The transformation 2. t) ∂w +g(x. leads to a simpler equation of the form 1. 2◦ . Polyanin (1996). ∂τ ∂x ∂x If m = −1 or m = −2. τ )F (t). ∂w + f (t)w. ∂ ∂w ∂w =a w–4/3 + f (x)w–1/3 .3 for solutions of this equation. Polyanin (1996). z = wu3 ξ= u2 simpliﬁes the original equation.10. V. F (t) = exp f (t) dt . t) is a solution of this equation.10. w(x.6. 3. τ= F m (t) dt. =a ∂t ∂ξ ∂ξ Ç ¤Æ Å Reference: V. Then the function w1 = C1 w(x. The substitution w = v −3 leads to an equation of the form 1.1.

10. where C1 and C2 are arbitrary constants.6. 2◦ . ψt − f (t)ψ + C1 ψ m+1 = 0. u(x. u(x. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + bw1+m + f (t)w + g(t)w1–m .6. Suppose w(x. u(x. For b = 0. ∂ ∂w =a wm + g(x)wm+1 + f (t)w.1: ∂ 2u a ∂u = au 2 + ∂t ∂x m ∂u ∂x 2 + bmu2 + mf (t)u + mg(t). t) = ϕ(t) + ψ(t) sinh(λx + C). D. ∂w ∂t =a ∂ ∂x wm ∂w ∂x + f (x)w1+m . © 2004 by Chapman & Hall/CRC . where the functions ϕ(t) and ψ(t) are determined by systems of appropriate ﬁrst-order ordinary differential equations. ∂w where the functions ϕ = ϕ(x) and ψ = ψ(t) are determined by the ordinary differential equations (C1 is an arbitrary constant) a(ϕm ϕx )x + g(x)ϕm+1 + C1 ϕ = 0. which admits generalized separable solutions of the following forms: u(x. where λ and C are arbitrary constants. Polyanin (1996).13. is also a solution of the equation. Galaktionov (1995). C1 t + C2 ). t) = ϕ(t) + ψ(t) cosh(λx + C). t) = ϕ(t) + ψ(t) cos(λx + C).1. 1 ψ = ϕm+1 . The general solution of the second equation is given by (C2 is an arbitrary constant) ψ(t) = eF C2 + mC1 −1/m emF dt . A. see equation 1. Multiplicative separable solution: w = (λmt + C)−1/m ϕ(x). ∂t ∂x ∂x Multiplicative separable solution: w = ϕ(x)ψ(t). 6. and C is an arbitrary constant.6. Then the function m w1 = C1 w(x. Zaitsev and A. t) = ϕ(t) + ψ(t) exp( λx). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 101 5. Ê ¤É È References: V. F = f (t) dt. the parameter λ is a root of a quadratic equation. and the function ϕ(x) is determined by the equation aψxx + (m + 1)f (x)ψ + λ(m + 1)ψ m+1 = 0. 7.4: The substitution u = w m leads to an equation of the form 1. t) is a solution of this equation. V. 1◦ . The book by Polyanin and Zaitsev (2003) presents exact solutions of this equation for various f (x). F.

Zaitsev and A.3). τ )G(t). is also a solution of the equation. ∂τ ∂z ∂z In the special case m = −2.102 8.3 for solutions of this equation. τ= F 2 (t)H m (t) dt. ∂ ∂w =a wm ∂t ∂x ∂x The transformation w(x.7: ∂ ∂u ∂u =a um . Equations of the Form ∂w = a ∂ eλw ∂w ∂t ∂x ∂x 1. λ where C1 and C2 are arbitrary constants. F. eλw ∂w =a ∂ ∂w + f (x. t) = u(z. t) is a solution of this equation. G(t) = exp g(t)dt . w(t. τ )G(t). Polyanin (1996). w(x.1.10. F.10. x) = u(z.7: ∂ ∂u ∂u =a um . Ì ¤Ë Reference: V. Then the function 2 w1 = w(C1 x + C2 . z =x+ f (t) dt. τ= F 2 (t)Gm (t) dt. ∂t ∂x ∂x 1◦ . D. ∂ ∂w ∂w =a wm ∂t ∂x ∂x The transformation + xf (t) ∂w + g(t)w. this equation can be transformed to the linear heat equation (see 1.10. Polyanin (1996). leads to a simpler equation of the form 1. ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE + f (t) ∂w ∂x + g(t)w. ∂τ ∂z ∂z 9. + xf (t) + g(t) ∂w + h(t)w. w) © 2004 by Chapman & Hall/CRC .6. G(t) = exp g(t) dt . Suppose w(x. ∂x z = xF (t). H(t) = exp h(t) dt . D.10.10. leads to a simpler equation of the form 1.7: ∂ ∂u ∂u =a um . ∂x g(t)F (t) dt. t) = u(z. τ )H(t). z = xF (t) + where the functions F (t) and H(t) are given by F (t) = exp f (t) dt . ∂ ∂w ∂w =a wm ∂t ∂x ∂x The transformation 10. ∂τ ∂z ∂z If m = −1 or m = −2.14. τ= Gm (t) dt. t) − ln |C1 |.10.2 or 1. see 1.1. Zaitsev and A.1.1.1. leads to a simpler equation of the form 1. + f (t). 1. where the functions F (t) and G(t) are given by F (t) = exp f (t) dt .1. t. Ì ¤Ë Reference: V.

µ is a root of a quadratic equation. Polyanin (1996). leads to a simpler equation of the form 1. The transformation w(x. τ= exp λF (t) dt. ψ = eλϕ .1: ∂ ∂u ∂u =a eλu . F (t) = f (t) dt.2: ∂2u ∂u = au 2 + λf (t)u + λg(t).1: ∂2u ∂u = au 2 + bu2 + λf (t)u + λg(t). The substitution u = eλw leads to an equation of the form 1.6. t) is a solution of this equation.2.6. © 2004 by Chapman & Hall/CRC Ï ¤Î . Zaitsev and A. and C is an arbitrary constant. ∂ ∂w ∂w =a eλw + f (x)eλw . λ where C1 and C2 are arbitrary constants.6. Then the function 1 w1 = w(x. is also a solution of the equation.2. 4.2. ∂τ ∂x ∂x 2. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 103 2◦ . 2◦ . 3. Polyanin (1996). see equation 1. 5. λ where λ and C are arbitrary constants.1: ∂2u ∂u = au 2 + λf (x)u + λ(bx + c). ∂t ∂x ∂x For b = 0. F. ∂t ∂x This equation admits generalized separable solutions of the following forms: u(x. D. The substitution u = eλw leads to an equation of the form 1. ∂t ∂x This equation admits a generalized separable solution of the form u = ϕ(t)x 2 + ψ(t)x + χ(t). u(x. where the functions ϕ(t) and ψ(t) are determined by systems of appropriate ﬁrst-order ordinary differential equations.1. F. ∂w ∂t =a ∂ ∂x eλw ∂w ∂x + f (t) + g(t)e–λw . τ ) + F (t).6.10. Ï ¤Î Í ∂w ∂ ∂w Reference: V. t) = u(x. t) = ϕ(t) + ψ(t) cos(µx + C). Additive separable solution: 1 ln(λt + C) + ϕ(x). ∂t ∂x ∂x 1◦ . t) = ϕ(t) + ψ(t) cosh(µx + C). Zaitsev and A. t) = ϕ(t) + ψ(t) exp( µx).9. D. The substitution u = eλw leads to an equation of the form 1. + beλw + f (t) + g(t)e–λw . t) = ϕ(t) + ψ(t) sinh(µx + C). w=− Reference: V. u(x. =a eλw u(x. ∂w ∂t =a ∂ ∂x eλw ∂w ∂x + f (x) + (bx + c)e–λw . Suppose w(x.9. and the function ϕ(x) is determined by the second-order linear ordinary differential equation aψxx + λf (x)ψ + λ = 0. ∂t ∂x This equation admits a generalized separable solution of the form u = λ(bx + c)t + ϕ(x). C1 t + C2 ) + ln C1 .6.14.

1. see equation 1. t) is a solution of the equation in question. t. w = ϕ(x) + ψ(t).15.2. f (w)wz z + 1 zwz = 0.10. (3) For f (w) = aw . at z = 0. and C3 are arbitrary constants. is also a solution of the equation. k2 3◦ . For f (w) = awm . Suppose w(x. (1) λw + C1 where C1 . see also the book by Lykov (1967).7. see equation 1. w1 = const): w = w0 w = w1 w → w0 w = w1 −1 at at t=0 x=0 (initial condition). C2 . and f (w) = (αw + βw + γ) . 1. ∂t ∂x ∂x This equation is frequently encountered in nonlinear problems of heat and mass transfer (with f being the thermal diffusivity or diffusion coefﬁcient) and the theory of ﬂows through porous media. (boundary condition). © 2004 by Chapman & Hall/CRC −2 . see Subsection 1. 2◦ .104 6. F = f (t) dt. To λ = 0 there corresponds a stationary solution. . w → w0 2 Then the boundary conditions for the ordinary differential equation (2) are as follows: as −1 z → ∞. to the linear equation aUxx + λg(x)U + λC1 = 0. with the change of variable U = e λϕ . 2 (2) Solutions of this form usually correspond to constant w in the initial and boundary conditions for the original partial differential equation (w0 . w. and for f (w) = a ln w + b. Then the function 2 w1 = w(C1 x + C2 .2. as x → ∞ (boundary condition).1. z= √ t where the function w(z) is determined by the ordinary differential equation w = w(z). for f (w) = eλw . 1◦ . k. ψt − f (t) + C1 eλψ = 0. ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE + g(x)eλw + f (t).6. ∂w = ∂ f (w) ∂w where C1 . and λ are arbitrary constants. The general solution of equation (2) is given by (C2 is an arbitrary constant) ψ(t) = F − 1 ln C2 + λC1 λ eλF dt . (1) (2) Equation (1) can be reduced. ∂ ∂w =a eλw ∂t ∂x ∂x Additive separable solution: where the functions ϕ = ϕ(x) and ψ = ψ(t) are determined by the ordinary differential equations (C1 is an arbitrary constant) a(eλϕ ϕx )x + g(x)eλϕ + C1 = 0. the general solutions of (2) were obtained by Fujita (1952). Self-similar solution: x (0 ≤ x < ∞). ∂w ∂t ∂x ∂x ∂x 1. Equations of the Form ∂w = ∂ f (w) ∂w + g x. C1 t + C3 ). f (w) = aw .4. Traveling-wave solution in implicit form: f (w) dw = kx + λt + C2 .2. C2 .

6. those corresponding to a problem with the initial and boundary conditions of (3) with w0 = 0 and w1 = 1. For example. Expressions (5) deﬁne a parametric representation of f = f (w). assuming in (5) that φ(z) = w0 z + 1 (w0 − w1 )e−λz λ (λ > 0. one can verify that equation (2) is satisﬁed by w = φz . 2φzz (5) where φ = φ(z) is an arbitrary function. © 2004 by Chapman & Hall/CRC . f (w) = s + φ − zφz .1 for various f = f (w). we integrate equation (2) with respect to z and then apply the hodograph transformation (where w is regarded as the independent variable and z as the dependent one) to obtain 1 z dw + A .6. No. he obtained a large number of exact solutions to the original equation for various f = f (w). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 105 TABLE 1 Solutions equation 1. the explicit representation is obtained by eliminating z. 5◦ . By direct substitution. All solutions are written out in implicit form. where z = xt −1/2 . A is an arbitrary constant. The explicit form of w = w(z) is obtained by the inversion of z = z(w). 1 2 3 4 5 6 7 8 9 10 11 1 2 1 8 1 16 Function f = f (w) n n n w − w2n 2 2(n + 1) n (1 − w)n−1 − (1 − w)2n 2(n + 1) n n w−2n − w−n 2(1 − n) 2 1 2 Solution z = z(w) 1 − wn (1 − w)n w−n − 1 cos cos2 cos3 1 2 πw 1 2 πw 1 2 πw Conditions n>0 n>0 0<n<1 sin2 1 2 πw sin(πw) πw + sin(πw) sin2 (πw) 5 + cos(πw) cos 1 2 πw 1 + 2 πw − 1 1 2 πw cos 1 1 − sin( 2 πw) w arccos w + 1 1 √ − 2 2 1 − w2 π − 2(1 − w) arcsin(1 − w) 1 √ − 2 4 2w − w2 w arcsin w 1 2 √ + w 4 4 1 − w2 1 2 (1 arccos w arcsin(1 − w) √ 1 − w2 − ln w − ln w) 4◦ . are listed below in Table 1. (4) f (w) = − zw 2 Substituting a speciﬁc expression z = z(w) for z on the right-hand side of relation (4). There is another way to ﬁnd an f (w) for which equation (2) admits exact solutions. We now describe a simple method for ﬁnding an f (w) such that equation (2) admits an exact solution. w1 > w0 ). and are valid within the range of their spatial localization 0 ≤ w ≤ 1.15. and s is an arbitrary constant. To this end.1. The method just outlined was devised by Philip (1960). Some of his results. one obtains a one-parameter family of functions f (w) for which z = z(w) solves equation (2). z = z(w).

t) = ¯ ¯ ¯ 1 w(x. A. W. Doyle and P. see equation 1. Another conservation law: ∂ ∂w ∂ (xw) + F (w) − xf (w) ∂t ∂x ∂x where F (w) = f (w) dw.13. and A. τ ) t0 ∂w (x. R. R. A. 9◦ . N. 1962. It follows from (6) and (7) that w is also ¯ a solution of equation (2) with f (w) = aw m + Aw 2 . f (w) = aw m . Then w = w(z) is also a solution of ¯ ¯ ¯ ¯ 1 the more complicated equation [F (w)wz z + 2 zwz = 0 with F (w) = f (w) + Ag(w) (A is an arbitrary constant). Fijalkow (1984). A. is a particular solution of ¯ equation (2) if f (w) is a power-law function. w − w0 w = w0 + (w1 − w0 )e−λz . i. Ibragimov (1994). in the form of a conservation law. Svirshchevskii (1983). V.106 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE and eliminating z. For f (w) = a(w 2 + b)−1 . Ovsiannikov (1959.3 (Example 10). τ ) ∂x dτ .6. the method outlined gives the following one-parameter family of functions: n n w2n + Awn−1 . 8◦ .5. = 0. V. H.. © 2004 by Chapman & Hall/CRC . P. one can construct other f (w). The transformation ¯ t = t − t0 . Zaitsev and A. W. A. (6) where the function g = g(w) is deﬁned parametrically by g(w) = 1 . Here is one more method for constructing an f (w) for which equation (2) admits exact solutions. w (9) In the special case of power-law dependence. where b is some constant.15. S. Strampp (1982). t) of a similar equation. P. Kurdyumov. and C = − 2 λ−2 . Munier. P. 1982). 6◦ . J. one obtains f (w) = A + B + C ln(w − w0 ). Note that this solution satisﬁes the 2 boundary conditions of (3). t) (8) takes a nonzero solution w(x.1. t) of the original equation to a solution w(x. ¯ (7) For example. V. x= ¯ x x0 t m−2 w(y. For the ﬁrst solution presented in Table 1. Polyanin (1996).e. f (w) = aw m . wz ¯ w = w(z). 7◦ . the function w = bz 2/m . Galaktionov. x=x0 w(x. ¯ ¯ ¯ ∂ ¯ ∂w ¯ ∂w ¯ = f(w) ¯ .2 and Subsection S. D. 1 1 where A = − 1 sλ−1 . B = 2 λ−2 1 + ln(w1 − w0 ) . ¯ ∂t ∂x ¯ ∂x ¯ 1 f¯(w) = 2 f w 1 . R. M. J. V. Mikhailov (1995). The equation in question is represented in conservative form. transformation (8) leads to equation (9) where f¯(w) = aw−m−2 . and E. A. Likewise. Dorodnitsyn and S. F. Burgan.1: L. Samarskii. Suppose w = w(z) is a solution of equation (2) with an f (w). Feix. Ñ ¤Ð References for equation 1. t) dy + f w(x0 . Vassiliou (1998). f (w) = wn − 2 2(n + 1) for which z = 1 − w n is a solution of equation.

© 2004 by Chapman & Hall/CRC Ò Ô ¤Ó Ô ¤Ó 3 . A2 . A4 and C1 and C2 are arbitrary constants. 2 where C1 and C2 are arbitrary constants. Galaktionov (1994). 2. (6) Z= A4 g(w) A4 where A1 . ◦ Ò 1◦ . (2) The book by Polyanin and Zaitsev (2003) present a considerable number of solutions to equation (2) for various ϕ = ϕ(w). A2 . and C 1 and C2 are arbitrary constants. In this case. Z = ϕ(t)x + ψ(t). ∂t ∂x ∂x This equation governs unsteady heat conduction in a quiescent medium in the case where the thermal diffusivity and the rate of reaction are arbitrary functions of temperature. Zaitsev (2002).6. where the function w(Z) is determined by the inversion of (6). − w = w(Z). ψ(t) = −ϕ(t) A1 ϕ(t) dt + A2 dt + C2 . Reference: V. Let g = g(w) be arbitrary and let f = f (w) be deﬁned by dw 1 A1 w + A3 Z dw exp −A4 . (3) f (w) = g(w) g(w) dw . A3 . and A3 are some numbers. and A4 are some numbers (A4 ≠ 0). Z = √ 2A3 t + C1 A3 3A3 where the function w(Z) is determined by the inversion of (4). g(w) = f (w) where A and B are some numbers. A. z = x + λt. there are generalized travelingwave solutions of the form w = w(Z).1. Let the function f = f (w) be arbitrary and let g = g(w) be deﬁned by A + B. ∂w w = w(z). (5) f (w) = g(w) g(w) dw A2 1 exp −A4 − . ϕ(t) . which is deﬁned implicitly by 1 f (w) dw = At − Bx2 + C1 x + C2 . F. Polyanin and V. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 107 ∂ ∂w = f (w) + g(w). A3 . D. Then there are generalized traveling-wave solutions of the form A2 A1 x + C2 − (2A3 t + C1 ). there is a functional separable solution. (1) The substitution 1 y(w) = f (w)wz λ brings (1) to an Abel equation of the second kind: yyw − y = ϕ(w). where ϕ(w) = −λ−2 f (w)g(w). Let now g = g(w) be arbitrary and let f = f (w) be deﬁned by A1 A2 w + B A2 A3 + Z dw. Ò ϕ(t) = C1 e2A4 t − −1/2 Reference: A. Traveling-wave solutions: 4◦ . where the function w(z) is determined by the autonomous ordinary differential equation [f (w)wz ]z − λwz + g(w) = 0. 2◦ . (4) Z = −A2 g(w) where A1 . In this case.

1. S. + h(x). and c are any numbers (the prime denotes a derivative with respect to w). ◦ 7 .1. Elenin. and Ibragimov (1994). 1. √ √ p(x) = C1 e −ax + C2 e− −ax if a < 0. f (w) dw = g(t) dt − x x0 ∂w = ∂ f (w) ∂w + g(t) (x − ξ)h(ξ) dξ + C1 x + C2 . Ö ¤Õ if ac > 0. Elenin. Ö ¤Õ Reference: V. Galaktionov (1994). b. Dorodnitsyn. C1 and C2 are arbitrary constants. g(w) = aϕ(w) + b + c[aϕ(w) + b]. and c are arbitrary constants.1 to 1.3. A. If f = dF (w)/dw and g = aF (w) + bw + c. 2 where C1 and C2 are arbitrary constants. For speciﬁc equations of this form.1 to 1. Let f (w) and g(w) be as follows: f (w) = wϕw (w). and x0 is any number. and a.1. Kurdyumov. ∂t ∂x ∂x f (w) Functional separable solution in implicit form: 3. © 2004 by Chapman & Hall/CRC . Dorodnitsyn. G.1.11 to 1. Ö ¤Õ √ √ C1 sin( ax) + C2 cos( ax) if a > 0. Galaktionov. only a limited number of equations were extracted that possess symmetries other than translations.3. 6◦ . see Subsections 1. Then there are functional separable solutions deﬁned implicitly by √ √ b ϕ(w) = eat C1 cos(x ac ) + C2 sin(x ac ) − a √ √ b at ϕ(w) = e C1 cosh(x −ac ) + C2 sinh(x −ac ) − a and C1 and C2 are arbitrary constants. then there is a conservation law e−bt p(x)w Here. where C1 and C2 are arbitrary constants. Galaktionov (1994). 9◦ . where ϕx = cp(x). where F (w) is an arbitrary function. Galaktionov. Dorodnitsyn (1979). g(w) = a w + 2 ϕ(w) . ϕ (w) where ϕ(w) is an arbitrary function and a. ϕw (w) where ϕ(w) is an arbitrary function and a is any number. G.2.1. and 1. b.13. t + e−bt p(x)x F (w) − p(x)(F (w))x + ϕ(x) x = 0. and Samarskii (1986). if ac < 0. A. As a result. 1982). 1. Reference: V. Kurdyumov. C1 x + C2 if a = 0. see also Dorodnitsyn and Svirshchevskii (1983). and A.2.4. Let the functions f (w) and g(w) be as follows: f (w) = ϕ (w). Then there are functional separable solutions deﬁned implicitly by ϕ(w) = C1 e2at − 1 a(x + C2 )2 . Group classiﬁcation of solutions to the equation in question was carried out by Dorodnitsyn (1979. References: V. A. V. 8◦ . V. Samarskii (1986). A. A. A. P.108 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 5◦ .

dτ = dt.6. and λ are arbitrary constants. 3◦ . Then the function w1 = w(x + C1 eat . where C1 and C2 are arbitrary constants. λw − kG(w) + C1 G(w) = g(w) dw. ∂τ ∂z ∂z ∂z where Φ(u) = 1 f u2 1 .3. where the function w(z) is determined by the ordinary differential equation [f (w)wz ]z − (az + b)wz = 0. C2 . © 2004 by Chapman & Hall/CRC . Fokas and Y. ∂w 1◦ . is also a solution of the equation. t) is a solution of the equation in question. ∂τ ∂ζ ∂ζ 5.1. t + C2 ). the original equation is an unnormalized Burgers equation 1.6. The transformation dz = w dx + f (w)wx + G(w) dt. ∂t ∂x ∂x ∂x 1◦ . 4. where C1 . Yortsos (1982). For f (w) = a and g(w) = bw. S. 2◦ . a one obtains a simpler equation of the form 1. ∂ ∂w ∂w ∂w = f (w) + g(w) . C. τ ): ∂ ∂w ∂w = f (w) . Functional separable solution in implicit form: f (w) dw = C1 e−at (ax + b) + C2 . The above transformation brings it to the solvable equation ∂ ∂u = ∂τ ∂z Ø ¤× a ∂u u2 ∂z + b ∂u . 2◦ .1 for w(ζ. 4◦ . z = x + C1 eat . Generalized traveling-wave solution: w = w(z). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 109 ∂ ∂w ∂w = f (w) – (ax + b) . Suppose w(x. u = 1/w dz = zx dx + zt dt Example. leads to an equation of the similar form ∂ ∂u ∂u ∂u = Φ(u) + Ψ(u) .15. x to the new variables τ= 1 1 − e−2at . On passing from t.1. k. ∂t ∂x ∂x ∂x This equation governs unsteady heat and mass transfer in an inhomogeneous ﬂuid ﬂow in the cases where the thermal diffusivity is arbitrarily dependent on temperature. Traveling-wave solution in implicit form: k2 f (w) dw = kx + λt + C2 . u G(w) = g(w) dw. 2a ζ = e−at x + b .5. u Ψ(u) = 1 1 g u u −G 1 . 2u2 ∂z Reference: A.

15. Polyanin (1996). where G(t) = B exp g(t) dt .1: ∂ ∂U ∂U = f (U ) . F. = ∂ f (w) ∂w ∂x + xg(t) + h(t) ∂w ∂x . leads to a simpler equation of the form 1. x to the new variables t. ∂ ∂w ∂w ∂w = f (w) + g(t) .110 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 3◦ . ∂w = ∂ f (w) one obtains a simpler equation of the form 1. ∂w ∂t ∂x The transformation w = U (z.6. Then the original equation has the following generalized traveling-wave solution: w = w(Z). On passing from t. © 2004 by Chapman & Hall/CRC . g(w) = A2 + A4 Z. Let f (w) and g(w) be deﬁned as f (w) = Zw A1 w + A3 where Z dw . where C1 is an arbitrary constant. ∂τ ∂z ∂z 8. if A3 = 0. one obtains a simpler equation of ∂ ∂w ∂w = f (w) . z = xG(t). and the function ϕ(t) is determined by the ﬁrst-order separable ordinary differential equation ϕt = A 3 ϕ3 + A 4 ϕ2 .15.6. z = xG(t) + h(t)G(t) dt.6. ∂t ∂x ∂x ∂x This equation governs unsteady heat conduction in a moving medium in the case where the thermal diffusivity is arbitrarily dependent on temperature. Conservation law: ◦ −1 (2) if A4 = 0. ∂t ∂x ∂x ∂x On passing from t. ∂t ∂z ∂z ∂w + xg(t) ∂w 7.15. whose general solution can be written out in implicit form. Z = ϕ(t)x + (A2 t + C1 )ϕ(t) + A1 ϕ(t) ϕ(t) dt. the function w(Z) is determined by the inversion of (1). Z = Z(w) (1) is a prescribed function (chosen arbitrarily). x to the new variables (A and B are arbitrary constants) τ= G2 (t) dt + A. 6. z): ∂ ∂w ∂w = f (w) . Zaitsev and A. D. G2 (t) dt. τ= G(t) = exp g(t) dt . τ ). G(w) = g(w) dw.1: g(t) dt. z = x + the form 1.1 for w(τ . ∂τ ∂z ∂z Ú ¤Ù Reference: V. . In special cases. Dt (w) + Dx −f (w)wx − G(w) = 0. solutions of equation (2) are given by ϕ(t) = (C2 − 2A3 t)−1/2 ϕ(t) = (C2 − A4 t) 4 .

The transformation u = w/x. Traveling-wave solution: w = w(z). ∂t ∂ξ © 2004 by Chapman & Hall/CRC . t + C2 ). ξ = 1/x leads to an equation of the similar form ∂2u ∂u = F (ξ)um 2 . and the function ϕ = ϕ(x) is determined by the generalized Emden–Fowler equation −1 ϕxx + λ f (x) ϕ1−m = 0. Suppose w(x. z = x + λt. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 111 ∂ ∂w ∂w = f (w) + g(w) + h(w). g(w). w) ∂ w ∂t ∂x2 ∂w = ∂ f (w) ∂w + ax + g(w) ∂w 1. where the function w(z) is determined by the ordinary differential equation [f (w)wz ]z + [az + g(w)]wz + h(w) = 0. a solution of equation (1) is given by x (x − ξ) dξ + Ax + B. 2◦ . Multiplicative separable solution: w(x. 2003) present a large number of solutions to equation (1) for various f (x).16. (1) For m = 1. ∂w + h(w). where the function w(z) is determined by the autonomous ordinary differential equation [f (w)wz ]z + [g(w) − λ]wz + h(w) = 0. is also a solution of the equation. The books by Polyanin and Zaitsev (1995. Equations of the Form ∂w = f (x. is also a solution of the equation. 9. Then the function w1 = w(x + C1 e−at . t) is a solution of this equation. ∂t ∂x ∂x ∂x 1◦ . 2003) present a large number of exact solutions to equation (2) for various f (w). 10. 2 1. 2◦ . (2) The books by Polyanin and Zaitsev (1995. C1 t + C2 ). and x0 are arbitrary constants.1. this equation governs unsteady heat conduction in a medium moving at a constant velocity in the case where the thermal diffusivity and the reaction rate are arbitrary functions of temperature. (1) The substitution y(w) = f (w)wz brings (1) to the Abel equation yyw + [g(w) − λ]y + f (w)g(w) = 0. where C and λ are arbitrary constants. where C1 and C2 are arbitrary constants. t) = (mλt + C)−1/m ϕ(x).6. where C1 and C2 are arbitrary constants. Generalized traveling-wave solution: w = w(z). t) is a solution of this equation. Suppose w(x. ∂t ∂x ∂x ∂x For g ≡ const. ∂2w ∂w = f (x)wm . ∂t ∂x2 1◦ . ϕ(x) = −λ f (ξ) x0 where A. F (ξ) = ξ 4−m f (1/ξ). z = x + C1 e−at .6. B. 3◦ . Then the function m w1 = C1 w(x. and h(w).

C2 . Then the function 2 w1 = w(C1 x + C2 . . Conservation laws: Dt (u) + Dx (−wx ) = 0. f (w) To obtain F = F (u) in explicit form. and u is deﬁned in © 2004 by Chapman & Hall/CRC . and λ are arbitrary constants.1: f (w) ∂ ∂u ∂u = F (u) . t) = f (x)ϕxx − ϕ2 = 0. dx . where Dt = Item 4◦ .6. and C3 are arbitrary constants. 2◦ . PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE f (x) ∂ 2 w ∂w . Dt (xu) + Dx (w − xwx ) = 0. C2 . Suppose w(x. is also a solution of the equation. so its general solution is given by ψ(x) = C1 ϕ(x) + C2 ϕ(x) where C1 and C2 are arbitrary constants. ϕ2 (x) ∂w = f (w) ∂2w where C1 .112 2. Dx = wx is the partial derivative of w with respect to x. f (x)ψxx − ϕψ = 0. The second equation has a particular solution ψ(x) = ϕ(x). The substitution u = dw leads to an equation of the form 1. u= dw . 3. a where the functions ϕ(x) and ψ(x) are determined by the system of ordinary differential equations w(x. λF (w) + C1 F (w) = dw . ∂t ∂x ∂x where the function F is deﬁned parametrically as F (u) = f (w). and the function U (z) is determined by the ordinary differential equation f (U )Uzz + 1 C2 zUz = 0. and C3 are arbitrary constants. ∂t ∂x2 1◦ . one should eliminate w from the two relations. The ﬁrst equation can be treated independently from the second. k. 3◦ . ∂ ∂x . 5◦ . z= √ C2 t + C 3 where C1 .15. Self-similar solution: x + C1 . t) is a solution of this equation. 4◦ . C1 t + C3 ). f (w) where C1 . C2 . = ∂t aw + b ∂x2 Generalized separable solution linear in t: 1 ϕ(x)t + ψ(x) − b . 2 w = U (z). ∂ ∂t . Traveling-wave solution in implicit form: k2 dw = kx + λt + C2 .

3◦ . Zaitsev and A. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 113 w ∂2w ∂w .3: 4. ∂t ∂x2 ∂x 1◦ . © 2004 by Chapman & Hall/CRC . ∂t ∂ξ 5. and C3 are arbitrary constants. where C1 . C1 . F. t. and C3 are arbitrary constants. = x4 f ∂t x ∂x2 The transformation u = w/x.6. one arrives at the simpler equation z= ax2 ∂2u ∂u 1 = u4 f (u) 2 + (ac − 4 b2 )u5 f (u). – aw3 .6. t) = C1 sin x where F = a 2.1. C2 . ∂ w ∂w + g(t)w + h2 (t)x2 + h1 (t)x + h0 (t).16. w. t) = ϕ(t)x2 + ψ(t)x + χ(t). Ü ¤Û Reference: V. Equations of the Form ∂w = f (x. D. ψt = 6f (t)ϕψ + g(t)ψ + h1 (t).6. 2◦ . ∂2u ∂u = f (u) 2 . Multiplicative separable solution for a < 0: w(x.17. t) = C exp Ý ∂w ∂2w ∂w 2 √ x a+a f (t) dt . ∂t ∂z which has a traveling-wave solution u = u(z + λt). ∂t ∂x ∂x Generalized separable solution quadratic in x: = f (t) w(x. Multiplicative separable solutions for a > 0: = f (t) +w w(x. where the functions ϕ = ϕ(t). and χ = χ(t) are determined by the system of ordinary differential equations ϕt = 6f (t)ϕ2 + g(t)ϕ + h2 (t). 2 1. where C is an arbitrary constant. t. ∂w ∂t ∂x2 ∂x 1. Polyanin (1996). dx + bx + c √ w(x. f (t) dt. t) = u(z. Multiplicative separable solution for a > 0: w(x. C2 . F =a f (t) dt. ξ = 1/x leads to a simpler equation of the form 1. ψ = ψ(t). w √ ∂t ax2 + bx + c With the transformation ∂w = w4 f ∂2w ∂x2 . ∂w |a| + C2 cos x |a| 2 2 eF C3 + 2a(C1 + C2 ) −1/2 e2F dt . t) = C1 ex √ a √ a −1/2 + C2 e−x eF C3 + 8aC1 C2 e2F dt . t) ax2 + bx + c. χt = 2f (t)ϕχ + f (t)ψ 2 + g(t)χ + h0 (t). w) ∂ w + g x.

t) = (x + C1 )ϕ(t). 4. where C and λ are arbitrary constants. and the function ϕ = ϕ(x) is determined by the ordinary differential equation [f (x)ϕm ϕx ]x + λϕ = 0. Φ = ϕm+1 f (x) brings (1) to the generalized Emden–Fowler equation z= Φzz + F (z)Φ m+1 = 0. Suppose w(x. (1) dx . f (x) 1 The transformation (2) The book by Polyanin and Zaitsev (2003. ∂t ∂x ∂x ∂x Degenerate solution linear in x: w(x.114 3.7) presents a large number of solutions to equation (2) for various F = F (z). t) = (mλt + C)−1/m ϕ(x). Then the function ∂w m w1 = C1 w(x. z= dx . ∂t ∂ξ ∂ξ where the function F = F (ξ) is deﬁned parametrically by F = f (x)[ψ(x) 3m+4 m+1 . t) = 1 x+ F (t) g(t) dt + C1 . Multiplicative separable solution: w(x. C1 t + C2 ). f (x) leads to an equation of the similar form ∂ ∂u ∂u = F (ξ)um . f (x) © 2004 by Chapman & Hall/CRC . F (t) F (t) = f (t) dt + C2 . 3◦ . 5. t). 2◦ . Sections 2. is also a solution of the equation. ξ=− ψ(x) m+2 m+1 dx. ψ(x) = dx . ∂w ∂ ∂w ∂w + f (t)w = g(t) w2 . where C1 and C2 are arbitrary constants. ∂ ∂w = f (x)wm . ∂w PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂ ∂w + f (t)w = g(t) w . where C1 and C2 are arbitrary constants. ψ(x) = dx . ∂t ∂x ∂x 1◦ . t) = ψ(x) 1 m+1 u(ξ. where the function F = F (z) is deﬁned parametrically by F = λ(m + 1)f (x). t) is a solution of this equation. where the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt = 2g(t)ϕ3 − f (t)ϕ2 . ∂t ∂x ∂x ∂x Degenerate solution linear in x: w(x.3 and 2. The transformation w(x. m+2 m+1 ξ=− ψ(x) dx.

w(x. = f (x)wm ∂t ∂x2 1◦ . ß ¤Þ 2(m + 2) m Reference: V. t) = (mλt + C)−1/m ϕ(x). ψt = 2f ϕψ + mg. Then the function 8. is also a solution of the equation. Zaitsev and A. t) = (mλt + C)−1/m ϕ(x). t) is a solution of this equation. Polyanin (1996). ∂w + g(t)w1–m .6. + g(x)wm+1 . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 115 6. equation (1) becomes ϕxx + (b/a)xk−n ϕ + (λ/a)x−n ϕ1−m = 0. Suppose w(x. Multiplicative separable solution (C and λ are arbitrary constants): w(x. ∂t ∂x ∂x 1◦ . F. In the special case of f (x) = axn and g(x) = bxk . F ψ = F − m+2 A + m m gF f dt. D. 2◦ . C1 t + C2 ). C1 t + C2 ). Suppose w(x. 2003) present a large number of solutions to equation (2) for various values of n. t) = ϕ(t)x2 + ψ(t) 1/m . m m+2 dt . ∂w = ∂ f (x)wm ∂w where C1 and C2 are arbitrary constants. where the functions ϕ = ϕ(x) and ψ = ψ(x) are determined by the system of ﬁrst-order ordinary differential equations 2(m + 2) 2 fϕ . ∂2w ∂w + g(x)wm+1 . © 2004 by Chapman & Hall/CRC (1) . where the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕm ϕxx + g(x)ϕm+1 + λϕ = 0. m. m w1 = C1 w(x. and k. 2◦ . (2) (1) The books by Polyanin and Zaitsev (1995. ϕt = m Integrating yields ϕ= 1 . is also a solution of the equation. F =B− where A and B are arbitrary constants.1. ∂w ∂ = f (t) wm ∂t ∂x ∂x Functional separable solution: 7. where the function ϕ = ϕ(x) is determined by the ordinary differential equation [f (x)ϕm ϕx ]x + g(x)ϕm+1 + λϕ = 0. Then the function m w1 = C1 w(x. where C1 and C2 are arbitrary constants. t) is a solution of this equation. Multiplicative separable solution (C and λ are arbitrary constants): w(x.

2003) present a large number of solutions to equation (3) for various values of n. G(t) = exp g(t) dt . leads to a simpler equation of the form 1. f (x) 1 z= brings (1) to the equation Φ = ϕm+1 Φzz + F (z)Φ m+1 + G(z)Φ = 0. The books by Polyanin and Zaitsev (1995.116 The transformation PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE dx . z =x+ g(t) dt. where the functions F = F (z) and G = G(z) are deﬁned parametrically by F = λ(m + 1)f (x). Zaitsev and A. dx z= .1. F. f (x) (2) In the special case of f (x) = axn and g(x) = bxk . â The transformation w(x. á ¤à Reference: V. and k. Polyanin (1996).7: ∂ ∂u ∂u = um . ∂τ ∂z ∂z 10. D.7: ∂ ∂u ∂u = um . m. ∂w ∂t = f (t) ∂ ∂x wm ∂w ∂x + xg(t) + h(t) ∂w ∂x + (t)w. τ ) exp h(t) dt . where the functions S(t) and G(t) are given by S(t) = exp s(t) dt . Polyanin (1996).10.10. n+k (3) where A = λa(m + 1) a(1 − n) 1−n and B = ab(m + 1) a(1 − n) 1−n . z = xG(t) + h(t)G(t) dt. equation (2) becomes Φzz + Az 1−n Φ m+1 + Bz 1−n Φ = 0. 9. © 2004 by Chapman & Hall/CRC . τ= f (t) exp m h(t)dt dt leads to a simpler equation of the form 1. dx . t) = u(z. τ= f (t)G2 (t)S m (t) dt. F.1. z= f (x) G = (m + 1)f (x)g(x). t) = u(z. The transformation w(x. D. n n 1 n+k n ≠ 1. Zaitsev and A. ∂w ∂t = f (t) ∂ ∂x wm ∂w ∂x + g(t) ∂w ∂x + h(t)w. ∂τ ∂z ∂z á ¤à Reference: V. τ )S(t).

t) is a solution of this equation. is also a solution of the equation. C1 t + C2 ) + 1 ln C1 . Suppose w(x. D. 2◦ . is also a solution of the equation. Additive separable solution: w(x. Reference: V.1. H(t) = exp h(t) dt . τ )H(t). t) = − 1 1 ln(βt + C) + ln β β A − βx dx + B . where the functions G(t) and H(t) are given by G(t) = exp g(t) dt . Then the function w1 = w(x. B. t) is a solution of this equation. Then the function w1 = w(x. f (x) where A.15. Polyanin (1996). 1◦ .6: ∂u ∂ ∂u um . F. = zk ∂τ ∂z ∂z ä ¤ã Reference: V. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 117 11. ∂w ∂t = ∂ ∂x f (x)eβw ∂w ∂x + g(x)eβw .6. β where β and C are arbitrary constants. ∂w ∂t = xk f (t) ∂ ∂x wm ∂w ∂x + xg(t) ∂w ∂x + h(t)w. z = xG(t). Zaitsev and A.1. 12. © 2004 by Chapman & Hall/CRC . Zaitsev and A. ä ¤ã ψ = eβϕ . 2◦ . C1 t + C2 ) + 1 ln C1 . The transformation w(t. Polyanin (1996). 13. β where C1 and C2 are arbitrary constants. and C are arbitrary constants. 1◦ . Additive separable solution: w(x. ∂w ∂t = ∂ ∂x f (x)eβw ∂w ∂x . Suppose w(x. β where C1 and C2 are arbitrary constants. D. leads to a simpler equation of the form 1. and the function ϕ(x) is determined by the second-order linear ordinary differential equation [f (x)ψx ]x + βg(x)ψ + β = 0. x) = u(z. t) = − 1 ln(βt + C) + ϕ(x). F. τ= f (t)G2−k (t)H m (t) dt.

see equation 1.19. is also a solution of the equation. see equation 1.1. Then the function w1 = w(x + C1 e−at . ∂w ∂t = ∂ ∂x PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE f2 (t)w2n + f1 (t)wn ∂w ∂x + g1 (t)w + g2 (t)w1–n .1. ∂w ∂t = x1–n ∂ ∂x f (w) ∂w ∂x . where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕt = (n + 1) f2 (t)ϕ3 + ng1 (t)ϕ. Then the function n+1 w1 = w(C1 x. 15.6.6. © 2004 by Chapman & Hall/CRC . t) = ϕ(t)x + ψ(t) 1/n . deﬁned by equation 1. where the function w(z) is determined by the ordinary differential equation f (w)wzz + [az + g(w)]wz + h(w) = 0.1. where the function w(z) is determined by the ordinary differential equation (n + 1)[f (w)wz ]z + z n wz = 0.6. t) is a solution of this equation. are reducible to equations of this form. z = xt− n+1 1 (0 ≤ x < ∞).2. t + C2 ).118 14. 2◦ . z = x + C1 e−at . where C1 and C2 are arbitrary constants. t) is a solution of this equation. For n = 1. (1) which is often accompanied by the boundary conditions of (3) in 1. The general solution of equation (1) with f (w) = a(w + b)−1 and arbitrary n can be found in Zaitsev and Polyanin (1993). 1◦ . C1 t + C2 ). n 1 (n + 1) f2 (t)ϕ2 ψ + ng1 (t)ψ + f1 (t)ϕ2 + ng2 (t). Generalized traveling-wave solution: w = w(z). and for f (w) = aw m .15. where C1 and C2 are arbitrary constants. 2◦ . Self-similar solution for n ≠ −1: w = w(z).6. Suppose w(x.15. ψt = n n which is easy to integrate (the ﬁrst equation is a Bernoulli equation and the second one is linear in ψ). 16. Generalized traveling-wave solution: w(x. 2 ∂t ∂x ∂x 1◦ . ∂2w ∂w ∂w = f (w) + ax + g(w) + h(w). Nonlinear problems of the diffusion boundary layer. is also a solution of the equation. Suppose w(x.15.

Let us integrate (1) with respect to z and then apply the hodograph transformation (with w regarded as the independent variable and z as the dependent one) to obtain f (w) = − 1 z n+1 w z n dw + A . It follows from (3) and (4) that w is also a solution of ¯ equation (1) with f (w) = aw m + Aw m−n−1 n+1 . © 2004 by Chapman & Hall/CRC . ¯ n+1 (A is any). + g(w). æ ¤å Reference: V. An explicit form of the solution. Reference: V. To λ = 0 there corresponds a stationary solution. F. 17. if f (w) is a power-law function of w. ϕw (w) ∂w = 1 ∂ xn f (w) ∂w where ϕ(w) is an arbitrary function. 2 where C is an arbitrary constant. one obtains a one-parameter family of functions f (w) for which z = z(w) solves equation (1). Let w = w(z) be a solution of equation (1) with some function f (w). A is any. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 119 3◦ . A. setting z = (1 − w)k . In this case. g(w) = a(n + 1)w + 2a ϕ(w) . λw + F (w) + C1 æ ¤å ξ = ln |x| + λt.6. It involves the following. there is an exact solution of the form w = w(ξ). is determined by the inversion of z = z(w). For example. wz ¯ w = w(z). C1 . 5◦ . Let f (w) and g(w) be deﬁned by f (w) = wϕw (w). 4◦ . then w = bz m is a solution ¯ of equation (1). with b being a constant. D. f (w) = aw m .1. F (w) = f (w) dw. where λ. w = w(z). there is a functional separable solution deﬁned implicitly by ϕ(w) = Ce2at − 1 ax2 . Then w = w(z) ¯ ¯ ¯ ¯ is also a solution of the more complicated equation (n + 1)[F (w)wz ]z + z n wz = 0 with F (w) = f (w) + Ag(w) where the function g = g(w) is deﬁned parametrically by g(w) = 1 . (3) (4) For example. For n = −1. where the function w(ξ) is deﬁned implicitly by f (w) dw = ξ + C2 . Galaktionov (1994). We now describe a simple way to ﬁnd functions f (w) for which equation (1) admits exact solutions. xn 1◦ . one obtains from (2) the corresponding f (w): f (w) = A(1 − w)k−1 − k (1 − w)k(n+1) . (n + 1)(nk + 1) A is any. and C2 are arbitrary constants. ∂t ∂x ∂x This is a nonlinear equation of heat and mass transfer in the radial symmetric case (n = 1 corresponds to a plane problem and n = 2 to a spatial one). There is another way to construct f (w) for which equation (1) admits exact solutions. Zaitsev and A. Polyanin (1996). (2) Substituting a speciﬁc z = z(w) for z on the right-hand side of (2).

6. F. τ )H(t). ∂w ∂ w + g x. w. ∂w z = xG(t) + + xg(t) + h(t) ∂w ∂x .15.3: ∂2w ∂w = ϕ(w) 2 . leads to a simpler equation of the form 1. τ= f (t)G2 (t) dt. fw is the partial derivative of f with respect to w. g(w) = b ϕ .1: ∂ ∂w ∂w = ϕ(w) . 2 ∂t ∂x ∂x The transformation 18. 2 ∂t ∂x ∂x ∂x With the transformation w(x. there is a functional separable solution deﬁned implicitly by bx2 . t. ∂τ ∂z ∂ ∂w = f (t) ϕ(w) ∂t ∂x ∂x The transformation 19. G(t) = exp g(t) dt . w) = g(t) dt − x x0 ∂w = ∂2 f (x. + h(x). t) = u(z. where C1 and C2 are arbitrary constants.18. Functional separable solution in implicit form: 20. Equations of the Form ∂w =f x. In this case. w. leads to a simpler equation of the form 1. ϕ where ϕ = ϕ(w) is an arbitrary function. τ= F k+2 (t)H k (t) dt.6. 2 1. D.6. ∂2w ∂w ∂w = f (t)ϕ(w) + xg(t) + h(t) .16. Polyanin (1996). w) + g(t) (x − ξ)h(ξ) dξ + C1 x + C2 . τ= f (t)G2 (t) dt. w) Here. ∂x2 f (x. ∂w k ∂ 2 w ∂w ∂w =a + f (t)x + g(t) + h(t)w. ϕ(w) = −bt − 4a Ce where C is an arbitrary constant. ∂τ ∂z ∂z è ¤ç Reference: V. ∂w ∂t ∂x ∂x2 ∂x 1. z = xG(t) + h(t)G(t) dt.120 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . z = xF (t) + g(t)F (t) dt. Let f (w) and g(w) be deﬁned as follows: f (w) = aϕ− n+1 2 ϕ ϕ n+1 2 dw. © 2004 by Chapman & Hall/CRC . and x0 is any number. h(t)G(t) dt. Zaitsev and A. G(t) = exp g(t) dt . ∂t fw (x.

3. © 2004 by Chapman & Hall/CRC 2◦ . C1 t + C2 ) + C3 . t) = (Akt + B)−1/k Θ(x) + C. 3. and the function Θ(x) is determined by the second-order ordinary differential equation k f (x) Θx Θxx + AΘ = 0. B.6. Solution: A(k + 1) exp A k+1 dx + C1 dx + C2 f (x) dx dx + C2 f (x) if k ≠ −1. ∂t ∂x ∂x2 This equation occurs in the nonlinear theory of ﬂows in porous media. t) is a solution of this equation. 2◦ .1. it governs also the motion of a nonlinear viscoplastic medium. k. C4 are arbitrary constants. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 121 where the functions F (t) and H(t) are given by F (t) = exp one arrives at the simpler equation ∂u ∂u =a ∂τ ∂x See equation 1. A. the special case 1. Solution: . t) = At + B + ϕ(z). . −1 2 w1 = C1 w(C1 x + C2 . and the function ϕ(z) is determined by the ordinary differential equation (2) k 2 f kϕz )ϕzz = λϕz + A. z = kx + λt. where A. .18. w(x. . (1) where A. The general solution of equation (2) can be rewritten in parametric form as f (u) du uf (u) du + C1 . w(x. Suppose w(x.6. B. C1 . ∂z 2 h(t) dt . and C are arbitrary constants. is also a solution of the equation. Relations (1) and (3) deﬁne a traveling-wave solution for A = 0 and an additive separable solution for λ = 0. if k = −1. Additive separable solution: w(x. 1◦ . ∂t ∂x ∂x2 1◦ . and C2 are arbitrary constants. t) = At + B + ϕ(x). . and λ are arbitrary constants. Then the function k w1 = C1 w(x. Suppose w(x. t) is a solution of the equation in question. ∂w = f (x) ∂w k k f (t) dt . B. C1 t + C3 ) + C4 . H(t) = exp ∂2u . where the function ϕ(x) is given by 1 ϕ(x) = ϕ(x) = C1 3◦ . ∂w ∂ 2 w ∂w =f . . C2 . z = k 2 + C2 . 2. and C3 are arbitrary constants. (3) ϕ=k λu + Ak λu + Ak where C1 and C2 are arbitrary constants. is also a solution of the equation. Then the function where C1 . ∂2w where C1 .

Equation ∂w ∂w =a ∂t ∂x wx = β1 w x − β 3 ¯¯ . 2 ∂x ¯ 1 f¯(z) = 2 f z 1 . and u(x) is determined by the ordinary differential equation ak(u x )k uxx + u = 0. x ξ= √ . b2 . C1 C2 t + C4 ) + C5 . . We have 1 f¯(wx ) = (β1 + β2 wx )2 f (wx ). ∂x2 k ≠ 0. Equation k+2 2 u 2ak 1 − k+2 du = x + C4 . a ∂2w ∂w = 2 . Suppose w(x. ¯¯ α Special case 1. β4 − β 2 w x ¯¯ where the subscripts x and x denote the corresponding partial derivatives.122 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE w(x. t) = ∂w leads to an equation of the form 1. Self-similar solution: where α. Solution: w(x.6. 1◦ . ¯ leads to an equation of the similar form ∂w ¯ ∂w ¯ = f¯ ∂t ∂x ¯ 6◦ . t) = √ t Θ(ξ). 2. ∂t ∂x ∂x 5◦ . where C1 . ¯ w = β 3 x + β4 w + γ 3 . and the γi are arbitrary constants such that α ≠ 0 and β1 β4 − β2 β3 ≠ 0. where C1 . Solution: é © 2004 by Chapman & Hall/CRC é é z = x2 cos−2 1 z − arctan ψ(z) 2 é w = bx tan 1 z − arctan ψ(z) 2 a t+C . takes the original equation to an equation with the same form. The substitution u(x. The transformation ¯ t = αt + γ1 . t). . . ¯ ∂2w ¯ . ∂t wx + b2 ∂x2 é wx = ∂w . C5 are arbitrary constants. Then the function k k+2 w1 = C1 w(C2 x + C3 . 4◦ . t) = C1 − b2 (x + C2 )2 − 2at + C3 . the βi . b2 a t+C . The hodograph transformation x = w(x. t) = (t + C1 )−1/k u(x) + C2 . ∂x 1.1: ∂x ∂ ∂u ∂u = f (u) . t) = x ¯ ¯ 3◦ . where C1 and C2 are arbitrary constants. . C2 . z w(x.15. the general solution of which can be written out in the implicit form as C3 − Special case 2. Solution: w(x. ¯ k ∂2w . t where the function Θ(ξ) is determined by the ordinary differential equation 2f Θξ Θξξ + ξΘξ − Θ = 0. 2◦ . x = β 1 x + β2 w + γ 2 . t) is a solution of this equation. is also a solution of the equation. and C3 are arbitrary constants.

I. . see Subsection S. ∂t ∂y 5.18. .2. 2◦ . . H. The function z = z(x. is also a solution of the equation. ϕz = 2z 2 2 2 z The function z = z(x. and C3 are arbitrary constants. wx = . Gazizov. ∂x2 1◦ . ψz = (1 + ψ 2 ) − − . Equation ∂w = k exp ∂t ∂w ∂x ∂2w . B. t) in the solution is deﬁned implicitly. t) in the solution is deﬁned implicitly. t) + C3 . at 2 b where C is an arbitrary constant.6. 2 ∂t ∂x ∂x The hodograph transformation. t) is a solution of this equation. N. according to which x is taken to be the independent variable and w the dependent one. where C1 . (x + B)2 + A2 © 2004 by Chapman & Hall/CRC ì ¤ë References for equation 1.3: E. Then the function −1 2 w1 = C1 w(C1 x + C2 . η) = xξ. Suppose w(x.6. leads to a similar equation for u = u(y. t) + u(ξ. ∂w where h(z) = z −2 h(1/z). 4. C at cos−2 ϕ(z) + ln 2 . The Euler transformation w(x. w = y. EQUATIONS INVOLVING ARBITRARY FUNCTIONS where C is an arbitrary constant. ∂η ∂ξ for details. Solution: w = bx tan ϕ(z) + z= b 2 x2 at C ln 2 2 b . Lenskii (1966). where C1 . ∂w ∂ 2 w ∂x2 = f t. t): ∂2u ∂u = g(y)f (u)h(uy ) 2 . . C. ∂t ∂x 1◦ . Sh. Ibragimov (1994). and D are arbitrary constants. x = u. Ibragimov (1989). ξ) 2 . and the functions ϕ(z) and ψ(z) are determined by the system of ordinary differential equations 1 C ψ ψ ψ . C1 eC3 t + C4 ) + C3 x + C5 . C5 are arbitrary constants. Solution: x+B A where A. and the function ψ = ψ(z) is determined by the ordinary differential equation ψz = ê 123 1 (1 + ψ 2 ) 2 1− ψ z . . t) is a solution of this equation. Akhatov. leads to the linear equation x= ∂u . and N.1. Special case 3. C2 . is also a solution of the equation. K. w(x. Then the function −1 w1 = C1 w(C1 x + C2 . ∂2w ∂w ∂w = f (x)g(w)h(wx ) . . H. 2◦ . ∂ξ t=η ∂2u ∂u = −f (η. Suppose w(x. t) = 2A arctan − (x + B) ln kt + C − (2 + ln 2)x + D. 3.3 (Example 7). V. R.

∂t ∂z ∂z References: Yu. Ryazantsev (1985).19. V. [ϕ(w)wξ ]ξ + 1 2 ξf (ξ) − (1) g(ξ) wξ = 0. F.6. 3. P. respectively. ∂w 1 y ∂w ∂ ∂w y + √ g √ = ϕ(w) . and Yu. A. convective diffusion to a ﬂat plate and that in liquid ﬁlms). Zaitsev and A. ∂x ∂y ∂y ∂y This equation is encountered in nonlinear problems of the steady diffusion boundary layer (mass exchange between solid particles. ∂w ∂ ∂w ∂w + g(x)y n = ϕ(w) .6. where h(x) = B exp − g(x) dx . w → a. respectively.15. ∂t ∂z ∂z î ¤í î ¤í References: A. to a drop or a bubble.6. x where the function w(ξ) is determined by the ordinary differential equation w = w(ξ). f (x) z = yh(x). Solving the original partial differential equation with simple boundary conditions of the ﬁrst kind. is reduced to solving equation (2) with the boundary conditions ξ = 0. where h(x) = B exp − g(x) dx . w = b. D. f (x) z = yh(x). f 1◦ . V. Polyanin. y = 0. f (x) leads to a simpler equation of the form 1. S. D. 1982). The transformation (A and B are arbitrary constants) 1. Polyanin (1980. drops. f (x)y n–1 t= hn+1 (x) dx + A. Nonlinear Equations of the Thermal (Diffusion) Boundary Layer ∂w ∂ ∂w + g(x)y = ϕ(w) . w = a.1: ∂ ∂w ∂w = ϕ(w) . Polyanin (1996). where a are b are some constants. . © 2004 by Chapman & Hall/CRC ξ → ∞. D. x = 0. √ x ∂x x x ∂y ∂y ∂y This is a generalization of the linear equation of the thermal boundary layer on a ﬂat plate. the coordinates x and y are reckoned along and normal to the body surface.16: ∂ ∂w ∂w = z 1−n ϕ(w) . y → ∞. Gupalo. w = b.124 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1. f (x) ∂w t= h2 (x) dx + A. Polyanin (1996). or bubbles and the ambient medium. w → a. The transformation (A and B are arbitrary constants) 2. Zaitsev and A. the coordinates x and y are reckoned along and normal to the interphase surface. (2) 2◦ . The value n = 2 corresponds to a solid particle and n = 1. f (x) leads to a simpler equation of the form 1. D. ∂x ∂y ∂y ∂y This equation is encountered in nonlinear problems of the steady diffusion boundary layer (mass exchange between drops or bubbles and a ﬂow). F. Self-similar solution: y ξ= √ .17.

where the functions a = a(t). Suppose w(x. k is a real number. + ¨ 1◦ . t) = C1 exp i [C2 x + (kC1 − C2 )t + C3 ] . and C3 are arbitrary real constants. 1. D. t) = A ñ ñ 2 2 exp[i(C1 t + C2 )] . Zaitsev (2002). β = β(t). NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 125 Remark.7. Reference: A. i + k|w|2 w = 0. B. b = b(t). D. © 2004 by Chapman & Hall/CRC ð ¤ï . Here. Equations of the Form i ∂w + ∂ w + f (|w |)w = 0 Involving ∂t ∂x2 Arbitrary Parameters Throughout this subsection. i 2 = −1.7. bt = −2aβ − 2bα. including nonlinear optics.1. superconductivity. Polyanin and V. Zaitsev (2002). A. g(ξ) = 2 Pr ξFξ (ξ) − F (ξ) . t) = (ax + b) exp i(αx2 + βx + γ) . A3 . are also solutions of the equation. Solutions: 2 2 w(x. Solution: w(x. B.7. t) = √ exp i 4t t where A. Reference: L. Faddeev and L. The second and third solutions are valid for k > 0. ð ¤ï ð ¤ï References: H. Schlichting (1981). ñ ñ 2 ∂w ∂2w w(x. ◦ 3 . D. 2◦ . This equation occurs in o various chapters of theoretical physics. F. Nonlinear Schrodinger Equations and Related Equations 2 1. Then the functions w1 = A1 w( A1 x + A2 . t). 1 w2 = e−i(λx+λ t+B) w(x + 2λt. C2 . The plus or minus signs in the expression of w1 are chosen arbitrarily. and Pr is the Prandtl number (x the coordinate along the plate and y the coordinate normal to the plate surface). A. The third solution describes the motion of a soliton in a rapidly decaying case. A2 . where A1 . and plasma physics. where F (ξ) is the Blasius solution in the hydrodynamic problem on the longitudinal homogeneous translational ﬂow of a viscid incompressible ﬂuid past a ﬂat plane. ¨ 1. w(x. k cosh(C1 x + C3 ) 2 exp[iBx + i(A2 − B 2 )t + iC1 ] . αt = ka2 − 4α2 . t) = C1 w(x. α = α(t). A2 t + A3 ). The classical thermal boundary layer equation is deﬁned by 1 f (ξ) = Pr Fξ (ξ). and γ = γ(t) are determined by the autonomous system of ordinary differential equations at = −6aα.¨ 1. βt = 2kab − 4αβ. Polyanin and V. w is a complex functions of real variables x and t. k cosh(Ax − 2ABt + C2 ) (x + C2 )2 C1 2 + i(kC1 ln t + C3 ) . and λ are arbitrary real constants. C1 . Takhtadjan (1987). ∂t ∂x2 Schr¨ dinger equation with a cubic nonlinearity. γt = kb2 − β 2 . F. t) is a solution of the Schrodinger equation in question.

Zakharov (1984). k det M(x. w(x. C2 = − ∞ −∞ w ∂w ∂w −w ∂x ∂x References: V. . t) . Novikov. . t) = Mn. 2 2 det R(x. . Takhtadjan (1987). t) = γn ei(λn x−λn t) . ô ¤ó 7◦ . The bar over a symbol denotes the complex conjugate. N . . S. C3 = ∞ −∞ ∂w ∂x 2 + |w|4 − 1 dx. C1 t + C 2 where C1 . . Self-similar solution: 1 u(z). N. t) n. B. References: G. C3 = ∞ −∞ 2 ∂w ∂x 2 − |w|4 dx. k = 1. C2 .126 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 4◦ . N RN +1. ◦ 8 . (rightmost column). V. . for t → ∞. ∂x ∂x 2 ∂w ∂ w ∂w ∂ w 1 − = g1 + ∂t ∂t 2 ∂x ∂x Here. . D.k (x. L.1 with f (u) = ku 2 . it is obtained by augmenting M(x. . f1 = w − w. B. Auto-B¨ cklund transformations preserving the form of the equation (with k = 1): a f2 = w + w. k = 1. 1/2 dx. . t) with a column on the right and a row at the bottom. Ibragimov (1985). (bottom row). and V. Pitaevskii. and C3 are arbitrary constants. t) for n = 1. the bar over a symbol denotes the complex conjugate. (lower right diagonal entry). The Schr¨ dinger equation with a cubic nonlinearity admits inﬁnitely many integrals of motion.7. N Rn.5. Manakov. E. Reference: L. . 2 6◦ .n (x. λ n − λk gn (x. S. t) = Here. N The above solution can be represented. t) = 1 RN +1. dx. . . The ﬁrst three integrals for k = −2: C1 = ô ¤ó ∞ −∞ 1 − |w|2 dx. where the λn and γn are arbitrary complex numbers that satisfy the constraints Im λn > 0 (λn ≠ λk if n ≠ k) and γn ≠ 0. Zakharov and A.N +1 (x.k (x. and the function u = u(z) is determined by the ordinary differential equation uzz + k|u|2u − 1 iC1 (zuz + u) = 0. for n = 1. ◦ 5 . t) = gn (x. Lamb (1974). . t) is of order N + 1. The square matrix R(x. © 2004 by Chapman & Hall/CRC ò − ag2 + i f1 |f1 |2 + |f2 |2 .k (x. . P. Faddeev and L. A. 4 . t) = √ C1 t + C 2 x + C3 z= √ . t) = 0 ô ¤ó (bulk of the matrix). E. L. For other exact solutions.N +1 (x. N -soliton solutions for k > 0: w(x. o The ﬁrst three integrals for k = 2: C1 = ∞ −∞ |w|2 dx. H. Shabat (1972). t) . . t) = 1 + g n (x. . t) for n. t)gn (x. see equation 1. The entries of R are deﬁned as Rn. C2 = ∞ −∞ w ∂w ∂w −w ∂x ∂x It is assumed here that the initial distribution w(x. M(x. i ∂w ∂ w − = iaf1 − f2 g1 . as the sum of N single-soliton solutions. . 1 g2 = i af1 − 2 f2 g1 . 0) decays quite rapidly as |x| → ∞. t) is an N × N matrix with entries Mn. g1 = iε b − 2|f1 |2 ò where a and b are arbitrary real constants and ε = 1.

αt = Aa2 − 4α2 . t) = C1 exp i [C2 x + (AC1 + B − C2 )t + C3 ] . M. L.7. N. Dodd. b = b(t). where the functions a = a(t). ∂w ∂ 2 w + + (A|w|2 + B)w = 0. and C3 are arbitrary real constants. S. where a = a(t). β = β(t). bt = −2aβ − 2bα. t) = (ax + b) exp i(αx2 + βx + γ) . B.-L. Zakharov (1984). 3◦ . i + (A|w|2 + B|w| + C)w = 0. βt = 2Aab − 4αβ. C2 . γt = Ab2 − β 2 + B. Novikov. α = α(t). Then the functions n 2n w1 = B1 w( B1 x + B2 .7. B. i ¨ 1◦ . b = b(t). Solution: w(x. Ankiewicz (1997). B1 t + B3 ). There is an exact solution of the form w(x. The Schr¨ dinger equation with a cubic nonlinearity is integrable by the inverse scattering o method. 2◦ . Izergin (1993). The numbers A and B are assumed real. C. J. and λ are arbitrary real constants. Sulem and P. N. S. C. Korepin. K.1 with f (u) = Au 2 + Bu + C. Ablowitz and H.5. α = α(t). t) = √ exp i 4t t where C1 . E. o + 1◦ . D. see equation 1. © 2004 by Chapman & Hall/CRC ÷ ÷ 2 t+C) w(x + 2λt. are also solutions of the equation. N. B. 2◦ . i 1◦ . ∂t ∂x2 Schr¨ dinger equation with a power-law nonlinearity. R. see the literature cited below. see equation 1. P. The plus or minus signs in the expression of w1 are chosen arbitrarily. β = β(t).¨ 1.7.5. ∂t ∂x2 Schr¨ dinger equation with a cubic nonlinearity. and C are assumed real. w(x. D. and H. Gibbon. ∂w ∂ 2 w + + A|w|2n w = 0. Eilbeck. and V. and A. Shabat (1972). ∂w ∂2w w2 = e−i(λx+λ where B1 . G. Morris (1982). t) is a solution of the Schrodinger equation in question. Segur (1981). ∂t ∂x2 Schr¨ dinger equation with a cubic nonlinearity. Zakharov and A. Sulem (1999). Manakov. E. J.1 with f (u) = Au 2 + B. B2 . Akhmediev and A. Takhtadjan (1987). C. o 2. C. N. t) = (ax + b) exp i(αx2 + βx + γ) . The numbers A. . (x + C2 )2 C1 2 + i(AC1 ln t + Bt + C3 ) . and γ = γ(t) are real functions of a real variable. and γ = γ(t) are determined by the autonomous system of ordinary differential equations at = −6aα. 3. Faddeev and L. Solutions: 2 2 w(x. For other exact solutions. E. o 4. Pitaevskii. A. L. B3 . t). NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 127 9◦ . V. V. Suppose w(x. ö ¤õ References: V. Bogoliubov. For other exact solutions. J. The numbers A and n are assumed real.

Then the functions 2 w1 = C1 eiC2 w( C1 x. Solution: w(x. are also solutions of the equation.2. o 1.7. C2 . C2 . z= √ C1 t + C 2 + f (|w |)w = 0 . 1. Suppose w(x. see equation 1. Self-similar solution: w(x. i ∂w ∂t + 1 xn ∂ ∂x xn ∂w ∂x + A|w|2 w = 0. Solution: w(x. w(x.5. © 2004 by Chapman & Hall/CRC ø t+C) U (x + 2λt). To ¨ n = 1 there corresponds a two-dimensional Schrodinger equation with axial symmetry and to n = 2. i 2 = −1. t) = 2 A cosh (C1 nx + C2 ) 2n (x + C2 )2 AC1 1−n C1 +i t + C3 . ϕ(x. 2◦ . and the function u = u(x) is determined by the ordinary differential equation x−n (xn ux )x − C1 u + Au3 = 0. Equations of the Form i ∂w + 1 ∂ xn ∂w ∂t xn ∂x ∂x Involving Arbitrary Parameters Throughout this subsection. o ¨ 1◦ . 1 where C1 . C1 t + C3 ). 2 n 5◦ . 4◦ .128 2◦ . Schr¨ dinger equation with a cubic nonlinearity. t) . and C3 are arbitrary constants. where C1 and C2 are arbitrary real constants.1 with f (u) = Au 2n . a three-dimensional Schr¨ dinger equation with central symmetry. Its solution can be represented in implicit form. t) = u(x) exp iϕ(x. 2 2n (n + 1)C1 2 exp[i(C1 t + C3 )]. w is a complex function of real variables x and t. and the function U = U (y) is determined by the autonomous ordinary differential equation Uyy + AU 2n+1 = 0. C2 . w(x. t) = e−i(λx+λ 2 where C and λ are arbitrary constants. x + C3 . and C3 are arbitrary real constants. 3◦ . Solutions: PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 2 w(x. ø where C1 . t) = √ exp i 4t 1−n t 1 where C1 . t) = C1 t + C2 dx xn u2 (x) + C3 . 3◦ .7. Multiplicative separable solution: w(x. t) = u(x)ei(C1 t+C2 ) . and the function u = u(z) is determined by the ordinary differential equation 1 1 uzz + k|u|2n u − iC1 zuz + u = 0. For other exact solutions. t) = C1 exp i [C2 x + (A|C1 |2n − C2 )t + C3 ] . t) is a solution of the Schrodinger equation in question. t) = (C1 t + C2 )− 2n u(z). and C3 are arbitrary real constants.

∂t x ∂x ∂x Schr¨ dinger equation with a cubic nonlinearity. t) . ϕ(x. t) . and the function u = u(x) is determined by the ordinary differential equation 2 x−n (xn ux )x − C1 u − C2 x−2n u−3 + Au3 = 0. and C3 are arbitrary real constants. Solution: w(x. i 1 ∂ ∂w ∂w + n xn + (A|w|2 + B)w = 0. C2 . are also solutions of the equation. 4(t + C2 ) n(t + C2 )n where C1 . Suppose w(x. C2 . 2◦ .¨ 1. 4(t + C2 ) n(t + C2 )n where C1 . t) = C1 t + C2 dx + C3 . Solution: w(x. ϕ(x. t) = C1 (t + C2 )− n+1 2 exp iϕ(x. and C3 are arbitrary real constants. Solution: w(x. NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 129 where C1 . and C3 are arbitrary real constants. i ∂w ∂t + 1 ∂ xn ∂x 1 xn ∂w ∂x + (A|w|2 + B|w| + C)w = 0.7. where C1 and C2 are arbitrary real constants. C2 . Then the functions 2 k 2k w1 = C1 eiC2 w( C1 x. C2 . t) = u(x) exp iϕ(x. This is a special case of equation 1. Multiplicative separable solution: where C1 . ∂t ∂x ∂x Schr¨ dinger equation with power-law nonlinearity. 4◦ . C2 . t) . o w(x. t) = 2 AC1 x2 − + Bt + C3 . and C3 are arbitrary real constants. 2◦ . Multiplicative separable solution: w(x. i xn xn ¨ 1◦ .5. t) = C1 (t + C2 )− n+1 2 exp iϕ(x. 3◦ .2 with f (u) = Au2 + Bu + C. ∂w + ∂ ∂w where C1 . o 4. t) = u(x)ei(C1 t+C2 ) . C1 t + C3 ). t) = 2 AC1 x2 − + C3 . 2. t) is a solution of the Schrodinger equation in question. where C1 and C2 are arbitrary real constants. The plus or minus signs are chosen arbitrarily. t) = u(x)ei(C1 t+C2 ) . and C3 are arbitrary real constants. xn u2 (x) 1◦ . 3. + A|w|k w = 0. and the function u = u(x) is determined by the ordinary differential equation x−n (xn ux )x − C1 u + (Au2 + B)u = 0. and the function u = u(x) is determined by the ordinary differential equation 2 x−n (xn ux )x − C1 u − C2 x−2n u−3 + (Au2 + B)u = 0. ϕ(x.7. and the function u = u(x) is determined by the ordinary differential equation x−n (xn ux )x − C1 u + A|u|k u = 0. © 2004 by Chapman & Hall/CRC ù ù .

Calogero and A. v(x. Then the function 2 4 w1 = C1 w(C1 x + C2 . t) = 2 n+1 2 exp iϕ(x. Garnon and P. 3◦ . ∂t ∂x This equation is encountered in plasma physics (propagation of Alfven and radio waves). 2−k−nk 2A|C1 |k x + (t + C2 ) 2 + C3 . and C3 are arbitrary real constants.5 with f (z) = a(1 + z 2 )−1/2 . t) = u(x) exp iϕ(x.5.3. See also Calogero and Degasperis (1982). Other Equations Involving Arbitrary Parameters 1. 2. t) are listed in Table 2. t) .130 3◦ . a is a real number. Winternitz (1993). t) = u(t) exp[iv(x.4. ϕ= . t) = C1 t + C2 2 x−n (xn ux )x − C1 u − C2 x−2n u−3 + A|u|k u = 0. 4(t + C2 ) 2 − k − nk where C1 .5. i + ∂x2 + ia ¨ 1◦ . 4.1 with f (u) = a 1 − e−ku . 4◦ . χ=− ψ 2 + aψu2 ) dt + C4 . ∂w ∂2w ∂ where C1 . ∂w ∂2w © 2004 by Chapman & Hall/CRC û ¤ú û ¤ú û ¤ú |w|2 w = 0. 1. ϕ(x.7. . w(x. 3. Degasperis (1982). t) . Solution: PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE dx + C3 . i ∂w ∂t +a ∂2 ∂x2 w 1 + |w|2 = 0. t)]. and C3 are arbitrary real constants. .5 f1 (x) = 1 and f2 (x) = a.7. This is a special case of equation 1. t)w = 0. and C3 are arbitrary real constants. t) = ϕ(t)x2 + ψ(t)x + χ(t). u= √ 4(t + C1 ) 4(t + C1 ) t + C1 where C1 . is also a solution of the equation. . This is a special case of equation 1.7. C2 . C4 are arbitrary real constants. .7. Suppose w(x. C1 t + C3 ). 2◦ . ∂w ∂2w .7. C2 . ∂t ∂x2 Exact solutions of this equation for some speciﬁc f (x. ∂t ∂x2 This equation is encountered in plasma theory and laser physics. 2 1 C3 − 2aC2 ln |t + C1 | C2 .4 with f (z) = az 2 . + f (x.5. ψ= . + Reference: R. and the function u = u(x) is determined by the ordinary differential equation w(x. 1978). Bullough (1977. Solution: w(x. C2 . Reference: F. and equation 1. t)w|w|2 + g(x.4 with f1 (t) = 1 and f2 (t) = a.4. see Item 2◦ of equation 1. See also equation 1. Solution: Here. i + a 1 – e–k|w| w = 0. t) is a solution of the Schrodinger equation in question. i + (1 + ia) Reference: L. t) and g(x. xn u2 (x) where C1 .7. t) = C1 (t + C2 )− ϕ(x. For another solution. K.

t) = C1 exp iϕ(x. t) 0 ic/t (c1 +ic2)/x2 1 −3/2 +βt−1 4 αxt Solution structure w(x. t) = (ax + b) exp i(αx2 + βx + γ) . z = x2/t √ w = ψ(z)/ x. t) 1+ib 1+ib (1+ib)/x (1+ib) exp αxt−1/2 f1(t)+if2(t) exp 2h(t)x 1+ib (1+ib)e−x (1+ib)e−kx (1+ib)x−k 1+ib g(x. C2 . i + 1◦ . 2 ϕ(x. Here. z = x2/t Solution 1: w = ϕ(t) exp ix . ∂w ∂2w where C1 . z = x2/t w = ψ(z)/x. t where C1 . t)w = 0 a 0 0 0 0 0 arbitrary arbitrary arbitrary arbitrary arbitrary f (x. w is a complex function of real variables x and t. Equations with Cubic Nonlinearities Involving Arbitrary Functions Throughout this subsection. t) = C2 x − C2 t + 2 C1 f (t) + g(t) dt + C3 . ∂t ∂x2 Schr¨ dinger equation with a cubic nonlinearity. i 2 = −1. t)w|w| + g(x. t 3◦ . C2 . and C3 are arbitrary real constants. t) Solution 1: w = ϕ(t) exp ix . z = x2/t w = ϕ(t) exp −h(t)x w = ψ(z)/x. 2◦ . w(x.7. t Solution 2: w = ψ(z)/x. 4t Solution 2: w = ψ(z)/x. Solution: w(x. t) = √ exp iϕ(x. t) . t) . f (t) and g(t) are real functions of a real o variable. 4t Solution 2: w = ψ(z)/x. 1.4. and C3 are arbitrary real constants. z = x2/t w = ϕ(t) exp w = ϕ(t) exp 1 2x 2 2 iht(t)x 0 ic x+ic (c1 +ic2)x−2 c1xt−3/2 −ic2t−1 1 2 kx+ixt w = x(k−2)/2 ψ(z). + f (t)|w|2 + g(t) w = 0. Solution: ϕ(x.¨ 1. z = x2/t 1.7. © 2004 by Chapman & Hall/CRC . Solution: C1 w(x. z = x2/t √ Solution 1: w = ϕ(t) exp − 2ic1x . NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 131 TABLE 2 ¨ Structure of exact solutions for the Schrodinger type ∂ 2w ∂w 2 equation i ∂t + (1 + ia) ∂x2 + f (x. t) = (x + C2 )2 + 4t 2 C1 f (t) + tg(t) dt + C3 .

t) = u(t) exp[iϕ(x. ϕ(x. whose general solution is given by u(t) = C3 eG(t) + 2eG(t) 2◦ . 4. i ∂w + ∂2w ∂x2 + [f1 (x) + if2 (x)]w|w|2 + [g1 (x) + ig2 (x)]w = 0. ∂t ∂x2 Equations of this form occur in nonlinear optics. + [f1 (t) + if2 (t)]w|w|2 + [g1 (t) + ig2 (t)]w = 0. w(x. t) = u(x) exp[iC1 t + iθ(x)]. ü where the functions u = u(x) and θ = θ(x) are determined by the system of ordinary differential equations 2ux θx + uθxx + f2 (x)u3 + g2 (x)u = 0. 2. F. bt = −2aβ − 2bα. Polyanin and V. ∂t ∂x2 Solutions: w(x. G(t) = ln t + 2 g2 (t) dt. 2t e−G(t) f2 (t) dt −1/2 . t) = u(t) exp[iϕ(x. Solutions: w(x. where the function u = u(t) is determined by the Bernoulli equation ut + f2 (t)u3 + g2 (t) + Integrating yields u(t) = C3 eG(t) + 2eG(t) þ ¤ý 1 u = 0. t) = C1 x − C1 t + [f1 (t)u2 (t) + g1 (t)] dt + C2 . ü ϕ(x. and γ = γ(t) are determined by the system of ordinary differential equations at = −6aα. γt = f (t)b2 − β 2 + g(t). t)]. t)]. ü e−G(t) f2 (t) dt −1/2 .132 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE where the functions a = a(t). Zaitsev (2002). t) = u(t) exp[iϕ(x. uxx − C1 u − u(θx )2 + f1 (x)u3 + g1 (x)u = 0. αt = f (t)a2 − 4α2 . βt = 2f (t)ab − 4αβ. Reference: A. t) = C1 x + 2 [−C1 f1 (t) + g1 (t)u2 (t) + h1 (t)] dt + C2 . ü ∂w ∂2w 2 ϕ(x. i + 1◦ . t) = (x + C1 )2 + 4t [f1 (t)u2 (t) + g1 (t)] dt + C2 . D. G(t) = 2 g2 (t) dt. Here. Solutions: w(x. i ∂2w ∂w + [f1 (t) + if2 (t)] + [g1 (t) + ig2 (t)]w|w|2 + [h1 (t) + ih2 (t)]w = 0. © 2004 by Chapman & Hall/CRC . the function u = u(t) is determined by the Bernoulli equation ut + f2 (t)u3 + g2 (t)u = 0. t)]. ∂t Solutions: 3. α = α(t). β = β(t). b = b(t).

Solutions: w(x. F. i where the functions u = u(x) and θ = θ(x) are determined by the system of ordinary differential equations 2f1 ux θx + f1 uθxx + f2 uxx − f2 u(θx )2 + g2 u3 + h2 u = 0. D. D. w(x. see Kuramoto and Tsuzuki (1975). where the functions u = u(x) and θ = θ(x) are determined by the system of ordinary differential equations f1 uxx − f1 u(θx)2 − f2 uθxx − 2f2 ux θx + g1 u3 + h1 u = 0. t) = u(x) exp[iC1 t + iθ(x)]. whose general solution is given by u(t) = C3 eF (t) + 2eF (t) ¡ÿ e−F (t) g2 (t) dt −1/2 . whose general solution is given by u(t) = C3 eF (t) + 2eF (t) F (t) = 2 7. ∂t Solutions: + [g1 (x) + ig2 (x)]w|w|2 + [h1 (x) + ih2 (x)]w = 0. Reference: A. NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 133 Here. ∂w + [f1 (x) + if2 (x)] ∂2w ∂x2 e−F (t) g1 (t) dt −1/2 . f2 uxx + C1 u − f2 u(θx)2 + f1 uθxx + 2f1 ux θx + g2 u3 + h2 u = 0. + [g1 (t) + ig2 (t)]w|w|2 + [h1 (t) + ih2 (t)]w = 0. hn = const. f1 uxx − C1 u − f1 u(θx)2 − 2f2 ux θx − f2 uθxx + g1 u3 + h1 u = 0. the function u = u(t) is determined by the Bernoulli equation 2 ut + g2 (t)u3 + [h2 (t) − C1 f2 (t)]u = 0. F. 5. F (t) = 2 2 h2 (t) − C1 f2 (t) dt. ∂t ∂x2 With fn . the function u = u(t) is determined by the Bernoulli equation 2 ut + g1 (t)u3 + [h1 (t) − C1 f1 (t)]u = 0. 6. gn . Zaitsev (2002). ∂2w ∂w + [f1 (x) + if2 (x)] + [g1 (x) + ig2 (x)]w|w|2 + [h1 (x) + ih2 (x)]w = 0. ¡ÿ ¢ ¢ Reference: A. t)]. 2 h1 (t) − C1 f1 (t) dt. © 2004 by Chapman & Hall/CRC . Here. ∂t ∂x2 Solutions: w(x. Zaitsev (2002). + [f1 (t) + if2 (t)] ϕ(x. t) = u(t) exp[iϕ(x.7. this equation is used for describing two-component reaction-diffusion systems near a bifurcation point. t) = C1 x + ¢ ∂w ∂2w 2 [C1 f2 (t) − g2 (t)u2 (t) − h2 (t)] dt + C2 . t) = u(x) exp[iC1 t + iθ(x)]. Polyanin and V. Polyanin and V.¨ 1.

t) . Multiplicative separable solution: w(x. C 1 . Integrating yields the general solution in implicit form: dU (A2 + B)U 2 − 2F (U ) + C1 £ = C2 ξ. and C are arbitrary constants): w(x. (1) £ 2 ∂w + ∂2w t+C1 ) w(x + 2λt + C2 . t) = C1 exp iϕ(x. C4 are arbitrary real constants. Solution: w(x. Then the function w1 = e−i(λx+λ 2◦ . B. where the function ψ = ψ(z) is determined by the ordinary differential equation w(x. B. © 2004 by Chapman & Hall/CRC . C2 . 3◦ . u2 (x) where C1 . 2 ϕ(x. t) = C2 x − C2 t + f (|C1 |)t + C3 . where the function U = U (ξ) is determined by the autonomous ordinary differential equation Uξξ + f (|U |)U − (A2 + B)U = 0. t) . ϕ(x. where C1 . t) = U (ξ)ei(Ax+Bt+C) . and C2 . C3 . F (u) = uf (|u|) du. and C3 are arbitrary real constants. F (U ) = U f (|U |) dU . (2) Relations (1) and (2) involve arbitrary real constants A. Traveling-wave solution: w(x. . C1 t ϕ(x. i ¨ 1◦ . 4◦ . 3 ψzz + f (|ψ|)ψ − (Az + B)ψ = 0. i 2 = −1. . where the function u = u(x) is deﬁned implicitly by du C1 u2 − 2F (u) + C3 Here. and λ are arbitrary real constants. is also a solution of the equation. C1 .7. o 1. C. C2 . t) is a solution of the Schrodinger equation in question. Equations of General Form Involving Arbitrary Functions of a Single Argument Throughout this subsection. t) = u(x) exp iϕ(x. t) = u(x)ei(C1 t+C2 ) . 6 . t) = 1 √ exp iϕ(x. . + f (|w|)w = 0. 5◦ . = C4 x. C2 . f (u) is a real function of a real variable. ξ = x − 2At. t) = (x + C2 )2 + 4t f |C1 t|−1/2 dt + C3 . Solution (A. t) . t) = C1 t + C2 dx + C3 . t + C3 ). Solution: w(x. where C1 . ∂t ∂x2 Schr¨ dinger equation of general form. 7◦ . Solutions: ◦ £ z = x − At2 . t) = ψ(z) exp i(Axt − 2 A2 t3 + Bt + C) . and C3 are arbitrary real constants.134 PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE 1.5. and the function u = u(x) is determined by the autonomous ordinary differential equation 2 uxx − C1 u − C2 u−3 + f (|u|)u = 0. Suppose w(x. . w is a complex function of real variables x and t.

3. t + C3 ). ϕ(x. Zaitsev (2002). k 2 uzz − k 2 u(ϕz )2 − λuϕz − Au + f (|u|)u = 0. ∂t x ∂x ∂x Schr¨ dinger equation of general form. ϕ(x. Polyanin and V. 3◦ . Suppose w(x. Traveling-wave solution: ¦ w(x. and the function u = u(x) is determined by the ordinary differential equation 2 x−n (xn ux )x − C1 u − C2 x−2n u−3 + f (|u|)u = 0. Solution: w(x. a threedimensional Schr¨ dinger equation with central symmetry. a and b are real numbers.¨ 1. Zaitsev (2002). 2◦ . where C1 and C2 are arbitrary real constants. t) = u(x) exp iϕ(x. k. © 2004 by Chapman & Hall/CRC . t) . f (u) and g(u) are real functions of a real variable. t) = x2 + 4t f |C1 |t− n+1 2 dt + C2 . z = kx + λt. Polyanin and V. and C3 are arbitrary real constants. t) = u(z) exp iAt + iϕ(z) . and λ are arbitrary real constants. Then the function w1 = eiC1 w(x + C2 . t) = x b f (|C1 |) + t g(|C1 |) − f (|C1 |) + C2 . t) = C1 t− n+1 2 exp iϕ(x. i 1◦ . t) is a solution of the generalized Landau–Ginzburg equation. is also a solution of the equation. where C1 . t) . F. F. t) . t) = C1 exp iϕ(x. 1950) and to describe two-component reactiondiffusion systems near a point of bifurcation (Kuramoto and Tsuzuki. and the function u = u(x) is determined by the ordinary differential equation x−n (xn ux )x − C1 u + f (|u|)u = 0.7. where C1 and C2 are arbitrary real constants. ∂t ∂x2 Generalized Landau–Ginzburg equation. and C3 are arbitrary real constants. To n = 1 there o ¨ corresponds a two-dimensional Schrodinger equation with axial symmetry and to n = 2. 1975). and the functions u = u(z) and ϕ = ϕ(z) are determined by the system of ordinary differential equations k 2 uϕzz + 2k 2 uz ϕz + λuz = 0. Reference: A. where A. o 2. C2 . NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 135 8◦ . D. ϕ(x. ¥ ¡¤ ¥ ¡¤ Reference: A. t) = C1 t + C2 dx xn u2 (x) + C3 . Equations of this form are used for studying second-order phase transitions in superconductivity theory (see Landau and Ginzburg. a a where C1 and C2 are arbitrary real constants. C2 . 1◦ . Multiplicative separable solution: w(x. D. 1 ∂ ∂w ∂w + n xn + f (|w|)w = 0. ∂2w ∂w = (a + ib) + f (|w|) + ig(|w|) w. f (u) is a real function of a real variable. Solution: w(x. 2◦ . where C1 . Solution: w(x. t) = u(x)ei(C1 t+C2 ) .

where k. −U (β + λVz ) + k 2 Uzz − k 2 U (Vz )2 − kf (U )U Vz = 0. and the functions U = U (z) and θ = θ(z) are determined by the system of ordinary differential equations aUzz − aU (θz )2 − bU θzz − 2bUz θz − λUz + f (|U |)U = 0. ϕt + 4ϕ2 = 0. t) = U (z) exp iC1 t + iθ(z) . . C4 are arbitrary real constants. A.136 3◦ . t) = U (z) exp[iβt + iV (z)]. © © where C1 . . Berman and Yu. Integrating yields C2 . C2 t + C4 ). Suppose w(x. . 2 f (|u|)u − aC1 u 4◦ . . Solution: w(x. z = x + λt. t) . t) = u(t) exp iϕ(x. Danilov (1981). whose general solution can be represented in implicit form as du = t + C3 . 5. t) = ϕ(t)x2 + ψ(t)x + χ(t). S. 2 where u = u(t) is determined by the ordinary differential equation ut = f (|u|)u − aC1 u. aU θzz − bU (θz )2 + bUzz + 2aUz θz − λU θz − C1 U + g(|U |)U = 0. ∂t 1◦ . χt + ψ 2 + ψf (u) = 0. The plus or minus signs are chosen arbitrarily. and the functions U = U (z) and V = V (z) are determined by the system of ordinary differential equations λUz + k 2 (U Vz )z + k 2 Uz Vz + k[f (U )U ]z = 0. Then the functions 2 w1 = eiC1 w( C2 x + C3 . where C1 . A. 2 ϕ(x. ¨ ¡§ References: V. 4(t + C1 ) ψ = −2ϕ f (u) dt + C3 ϕ. ψt + 4ϕψ + 2ϕf (u) = 0. i ∂2 ∂w + f (|w|)w = 0. where the functions u = u(t). Zaitsev (2002). © 2004 by Chapman & Hall/CRC . and λ are arbitrary real constants. Solution: 4. ψ = ψ(t). Solution: ϕ= 1 . Polyanin and V. t)]. ∂t ∂x2 1◦ . . C4 are arbitrary real constants. F. where C1 and λ are arbitrary real constants. β. D. w(x. u= √ t + C1 2◦ . . χ=− ψ 2 + ψf (u) dt + C4 . i ∂w + ∂2w ∂x2 +i ∂ ∂x f (|w|)w = 0. t) = C1 x − C1 bt + g(|u|) dt + C2 . are also solutions of the equation. Solution: PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE w(x. w(x. z = kx + λt. ϕ = ϕ(t). t) is a solution of this equation. . v(x. t) = u(t) exp[iv(x. . and χ = χ(t) are determined by the system of ordinary differential equations ut + 2ϕu = 0.

and the function u = u(x) is determined by the ordinary differential equation 2 uxx − C1 u − C2 u−3 + f (x. C2 . C2 . t) . t) = V (ξ). ψt + 4ϕψf (u) = 0. ϕ = ϕ(t). where k. ¡ Reference: A. Solution: w(x. Solution: w(x. f (x. and the function u = u(x) is determined by the ordinary differential equation uxx − C1 u + f (x. t)]. Multiplicative separable solution: w(x. i 2 = −1. 2 χ = −C2 u4 f (u) dt + C3 .7. 2◦ . u2 (x) ∂w ∂2w where C1 . 1. and C3 are arbitrary real constants. w is a complex function of real variables x and t. Equations of General Form Involving Arbitrary Functions of Two Arguments Throughout this subsection. where the functions u = u(t). o + 1◦ . t) = ϕ(t)x2 + ψ(t)x + χ(t). u3 f (u) 3◦ . χt + ψ 2 f (u) = 0. 4◦ . ψ = ψ(t). F. i + f (x. t) = U (z) exp[iβt + iV (z)]. D. and the functions U = U (z) and V = V (z) are determined by an appropriate system of ordinary differential equations (which is not written out here). There is a self-similar solution of the form w(x. Polyanin and V. v(x. © 2004 by Chapman & Hall/CRC . ϕt + 4ϕ2 f (u) = 0.6. β. and C3 are arbitrary real constants.7. |u|)u = 0. There is a solution of the form w(x. z = kx + λt. t) = u(t) exp[iv(x. where ξ = x2/t. u) is a real function of two real variables. t) = C1 t + C2 dx + C3 .¨ 1. NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 137 2◦ . and λ are arbitrary real constants. Integrating yields ϕ = C 1 u2 . ϕ(x. where C1 and C2 are arbitrary real constants. ψ = C 2 u2 . ∂t ∂x2 Schr¨ dinger equation of general form. 1. |u|)u = 0. and the function u = u(t) is deﬁned implicitly as (C4 is an arbitrary constant) du + 2C1 t + C4 = 0. and χ = χ(t) are determined by the system of ordinary differential equations ut + 2uϕf (u) = 0. t) = u(x) exp iϕ(x. where C1 . t) = u(x)ei(C1 t+C2 ) . |w|)w = 0. Zaitsev (2002).

2◦ . Zaitsev (2002). |u|)u = 0. t) = u(x)ei(C1 t+C2 ) . i + xn xn 1◦ . ϕ(x. ∂w + g(x) + Φ(x. Solutions: w(x. t) = C1 t− n+1 2 ∂w 1 ∂ ∂w exp iϕ(x. and C3 are arbitrary real constants. i ∂w + f (x) © 2004 by Chapman & Hall/CRC ∂2w . o ¨ 1◦ . Reference: A. + f (t. where C1 and C2 are arbitrary real constants. C2 . ∂w 1 ∂ ∂w where C1 . o Solution: 4. The case o g(x) = fx (x) corresponds to an anisotropic medium.138 2. t) . and the function u = u(x) is determined by the ordinary differential equation x−n (xn ux )x − C1 u + f (x. 2 ϕ(x. where C1 and C2 are arbitrary real constants. and λ are arbitrary real constants. u) is a real function of two real variables. |C1 |t−1/2 dt + C3 . and C3 are arbitrary real constants. D. i + xn xn w(x. F. t). u) is a real function of two real variables. t) . t) = (x + C2 )2 + 4t where C1 . F. and the function u = u(x) is determined by the ordinary differential equation 2 x−n (xn ux )x − C1 u − C2 x−2n u−3 + f (x. ¡ ¡ Reference: A. t) = C1 t + C2 dx xn u2 (x) + C3 . f t. t) is a solution of the Schrodinger equation in question. ϕ(x. i PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂ 2 w + f (t. w(x. |w|)w = 0. t) = u(x) exp iϕ(x. a three-dimensional Schr¨ dinger equation with central symmetry. t) = C2 x − C2 t + 2 t+C1 ) w(x + 2λt + C2 . Zaitsev (2002). |C1 |) dt + C3 . Solution: w(x. Polyanin and V. + f (x. + ∂t ∂x2 Schr¨ dinger equation of general form. f (t. f (t. |w|)w = 0. Φ(x. f (t. |w|)w = 0. t) . where C1 . D. o 3. t) = x2 + 4t f t. ∂t ∂x ∂x Schr¨ dinger equation of general form. Suppose w(x. ∂t ∂x ∂x Schr¨ dinger equation of general form. ∂t ∂x2 ∂x Schr¨ dinger equation of general form. To n = 1 o ¨ there corresponds a two-dimensional Schrodinger equation with axial symmetry and to n = 2. ¡ ψ(x. t) = C1 exp iϕ(x. Polyanin and V. t) = C1 t−1/2 exp iψ(x. Zaitsev (2002). |u|)u = 0. t) . |w|)w = 0. F. f (x. is also a solution of the equation. Reference: A. C2 . D. u) is a real function of two real variables. 5. |C1 |t− n+1 2 dt + C2 . Multiplicative separable solution: w(x. Then the function w1 = e−i(λx+λ 2◦ . C2 . Polyanin and V. u) is a real function of two real variables.

Multiplicative separable solution: w(x. ∂t ∂x2 Solution: w(x. t) = u(x)ei(C1 t+C2 ) . 7. where C1 and C2 are arbitrary real constants. n = 1. © 2004 by Chapman & Hall/CRC ψ 2 + ψf (t. where the functions u = u(x) and θ = θ(x) are determined by the system of ordinary differential equations f1 uxx − f1 u(θx )2 − f2 uθxx − 2f2 ux θx + g1 (|u|)u = 0. |w|) + if2 (t. ϕt + 4ϕ2 = 0. t) . . Solution: w(x. and χ = χ(t) are determined by the system of ordinary differential equations ut + 2ϕu = 0. t) = ϕ(t)x2 + ψ(t)x + χ(t). and the function u = u(t) is determined by the ordinary differential equation 2 ut = ug1 (t. where C1 and C2 are arbitrary real constants. ϕ= . |U |)U = 0. C4 are arbitrary real constants. ϕ(x. f (x) where C1 . |u|)u = 0. and the function u = u(x) is determined by the ordinary differential equation f (x)uxx + g(x)ux − C1 u + Φ(x. Here. t) = u(t) exp[iv(x. . t) = U (x) exp iϕ(x. ϕ(x. .7. |w|) + g1 (t. |u|). |u|). fn = fn (x. |u|) − C1 uf1 (t. . t) = u(t) exp iϕ(x. |u|) − C1 f2 (t. ϕ = ϕ(t). t) . t) . 8. w(x. . where the functions u = u(t). |w|) w. u) dt + C3 ϕ. Integrating yields 1 C2 . ∂t ∂x2 Solution: w(x. |u|). and the function U = U (x) is determined by the ordinary differential equation 2 f (x)Uxx + g(x)Ux − C1 U − C2 f (x)R2 (x)U −3 + Φ(x. χt + ψ 2 + ψf (t. ψt + 4ϕψ + 2ϕf (t. ∂2w ∂w = f1 (t.¨ 1. |w|)w = 0. U 2 (x) R(x) = exp − g(x) dx . gn = gn (x. ψ = −2ϕ f (t. 2. t) = C1 t + θ(x). |w|) + g1 (x. v(x. |w|) + ig2 (t. |w|) + ig2 (x. f1 uθxx − f2 u(θx )2 + f2 uxx + 2f1 ux θx − C1 u + g2 (|u|)u = 0. NONLINEAR SCHRODINGER EQUATIONS AND RELATED EQUATIONS 139 1◦ . 2◦ . u) = 0. t)]. i ∂w ∂t Solution: + ∂2w ∂x2 +i ∂ ∂x f (t. χ = − u= √ 4(t + C1 ) t + C1 where C1 . t) = u(x) exp iϕ(x. ∂2w ∂w = f1 (x. u) = 0. t) = C1 t + C2 R(x) dx + C3 . |w|) + if2 (x. u) dt + C4 . C2 . t) = C1 x + 2 g2 (t. |u|) dt + C2 . and C3 are arbitrary real constants. ψ = ψ(t). ϕ(x. |w|) w. 6.

ψt + 4ϕψf (t. t) = U (x) exp[iβt + iV (x)]. C2 . |w|)w = 0. ψ = ψ(t). u) = 0. ∂t ∂x ∂x There is a solution of the form 12. t) = u(t) exp[iv(x. u) = 0. u) dt + C3 . where the functions u = u(t). t) = ϕ(t)x2 + ψ(t)x + χ(t). i PARABOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w ∂ 2 w +i + f (x. and χ = χ(t) are determined by the system of ordinary differential equations ut + 2uϕf (t. t) = u(t) exp[iv(x. and the real functions U = U (x) and V = V (x) are determined by the system of ordinary differential equations (U Vx )x + Ux Vx + [f (x. |w|) ∂w = 0. ϕt + 4ϕ2 f (t. u) = 0. where β is an arbitrary real constant. i ∂w + ∂ f (x. ϕt + 4ϕ2 f (t. χt + ψ 2 f (t. u) = 0. t)]. and χ = χ(t) are determined by the system of ordinary differential equations ut + 2uϕf (t. and u = u(t) is determined by the ordinary differential equation ut + 2C1 u3 f (t. and u = u(t) is determined by the ordinary differential equation ut + 2C1 u3 f (t. −βU + Uxx − U (Vx )2 − f (x. v(x. ψt + 4ϕψf (t. 2 χ = −C2 u4 f (t. ψ = C 2 u2 . u) = 0. 2 ∂t ∂x ∂x Solution: w(x. 2 χ = −C2 u4 f (t. w(x. Integrating yields ϕ = C 1 u2 . u) = 0. w(x. where C1 . © 2004 by Chapman & Hall/CRC . u) = 0. ψ = ψ(t). where β is an arbitrary real constant. ψ = C 2 u2 . U )U ]x = 0. u) = 0. i ∂w + ∂ ∂x f (t. Integrating yields ϕ = C 1 u2 . χt + ψ 2 f (t. where the functions u = u(t). w(x. u) = 0. |w|) ∂w ∂x = 0. and C3 are arbitrary real constants.140 9. i ∂w ∂t Solution: + ∂2 ∂x2 f (t. v(x. ∂t Solution: 11. where C1 . and C3 are arbitrary real constants. t)]. ϕ = ϕ(t). u) dt + C3 . 10. C2 . U )U Vx = 0. and the functions U = U (z) and V = V (z) are determined by an appropriate system of ordinary differential equations (which is not written out here). |w|)w = 0. t) = U (x) exp[iβt + iV (x)]. ϕ = ϕ(t). t) = ϕ(t)x2 + ψ(t)x + χ(t). u) = 0.

∂w ∂t = ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y + cwp . t) = t 1−p V (z1 .1. z2 ).Chapter 2 Parabolic Equations with Two or More Space Variables 2. C12−m y. y. is also a solution of the equation. 1 r2 = e−λy x2−n + . Then the function w1 = C1 w C12−n x. 1◦ .1. y. y + p−1 1−p p−1 ln C1 . y. There are “two-dimensional” solutions of the following forms: w(x. There are “two-dimensional” solutions of the following forms: w(x. 1 w(x.3 with f (w) = cw p . t). z2 ). Suppose w(x. y.2. 2◦ . C1 t + C2 . Then the function p−1 w1 = C1 w C12−n x. t) = t 1−p V (z1 . t) is a solution of the equation in question. ∂w ∂t = ∂ ∂x axn ∂w ∂x + ∂ ∂y beλy ∂w ∂y z2 = yt m−2 .2. t) is a solution of the equation in question.4.1. a(2 − n)2 bλ2 z1 = xt n−2 . 2◦ . t) = U (r. 2. 1 This is a special case of equation 2. Equations with Two Space Variables Involving Power-Law Nonlinearities 2. 1◦ . λ where C1 and C2 are arbitrary constants. y. C1 t + C2 . λ © 2004 by Chapman & Hall/CRC . 1 r2 = y 2−m x2−n + . is also a solution of the equation. + cwp .1 with f (w) = cw p . 1 w(x. Equations of the Form ∂w = ∂ f (x) ∂w + ∂ g(y) ∂w +aw p ∂t ∂x ∂x ∂y ∂y 1. t). z2 = y + 1 ln t.4. Suppose w(x. t) = U (r. y. 2 a(2 − n) b(2 − m)2 z1 = xt n−2 . This is a special case of equation 2. p−1 p−1 where C1 and C2 are arbitrary constants.

y. This is a special case of equation 2. ∂2w ∂ ∂w ∂w =a +b w . The solution of this equation can be written out in implicit form: ξ=B+ where A and B are arbitrary constants.3. y. aβ 2 bλ2 1 1 w(x. y + ln C1 . It arises in nonlinear heat conduction theory and the theory of unsteady ﬂows through porous media with a free surface (see Polubarinova–Kochina. B. Suppose w(x.2. There are “two-dimensional” solutions of the following forms: w(x.4.1. 1◦ .1. e−βx e−λy + .2. © 2004 by Chapman & Hall/CRC 2 2 k1 + k2 λw − A ln |A + λw| . C1 t + C2 . y. Solution linear in all independent variables: w(x.4 with n = 1. 2 ∂t ∂x ∂y ∂y + b ∂ w k ∂w ∂y ∂y This is a special case of equation 2. z1 = x + ln t. Boussinesq equation. t) = U (r. 1962). t) is a solution of the equation in question. Then the function w1 = C1 w x + 1−p 1−p p−1 ln C1 . λ 2. y. β λ where C1 and C2 are arbitrary constants. and C are arbitrary constants. β r2 = z2 = y + 1 ln t. ∂w ∂t = ∂ ∂x w ∂w ∂x + ∂ ∂y w ∂w ∂y .1. 3◦ . t) = Ax + By + (A2 + B 2 )t + C. t) = f (t)x2 + g(t)xy + h(t)y 2 . 2◦ . k2 .2. Equations of the Form ∂w = a ∂ w n ∂w ∂t ∂x ∂x 1. Traveling-wave solution (k1 . z2 ). t).3. This is a special case of equation 2. and λ are arbitrary constants): w = w(ξ).2 with f (w) = cw p . where A. ξ = k1 x + k2 y + λt. 1◦ . y. Generalized separable solution quadratic in the space variables: w(x. is also a solution of the equation. ∂w ∂t = ∂ ∂x aeβx ∂w ∂x + ∂ ∂y beλy ∂w ∂y + cwp . 2. where the function w(ξ) is determined by the ordinary differential equation 2 2 λwξ = (k1 + k2 )(wwξ )ξ .1 with c = 0. 2◦ . t) = t 1−p V (z1 . λ2 .

g(t). where the functions f (t). gt = gΦ(g). On solving this equation with the change of variable u(g) = (gt )2 . h(t). y). g(t).4 with n = −1. g. gt = 6(f + h)g.4 with n = α = 1. gt = 6(f + h)g. With this relation. I. Φ(g) = 1 1 1 1 h = 12 Φ(g) − 2 Ag. D. A. V. Polyanin. Vyazmin. t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t).2. and.1. Ustinov (1985). using (2) and assuming g 0. and χ(t) are determined by the system of ordinary differential equations ft = 6f 2 + 2f h + g 2 . and D. A. A. and the function Θ is determined by the stationary equation written out in Item 4◦ of equation 2. S.1. 6t 2 5◦ . the solution can be represented in explicit form (C is an arbitrary constant): ¡ f (t) = µ+A . hence.where the functions f (t). Further. V. Kazenin (1998). A. and h(t) are determined by the autonomous system of ordinary differential equations ft = 6f 2 + 2f h + g 2 . £ ¤¢ References: S. we obtain (B is an arbitrary constant) Bg 4/3 + 36(1 + A2 )g 2 . we eliminate h from (2) and (3) to obtain a nonlinear ordinary differential equation for g(t): 3ggtt − 5gt2 − 36(1 + A2 )g 4 = 0. where A and B are arbitrary constants. its solution can be written out in implicit form. Example. ∂t ∂x w ∂x ∂y w ∂y This is a special case of equation 2. C − µt h(t) = µ−A . ht = 6h2 + 2f h + g 2 . In the special case B = 0. ψ(t). Pukhnachov (1995). Solution: w(x. ht = 6h + 2f h + g . © 2004 by Chapman & Hall/CRC . ψt = 2gϕ + 2(f + 3h)ψ. and h can be solved independently (see Item 3 ◦ ). t) = (At + B)−1 Θ(x. 3. y. There is a “two-dimensional” solution in multiplicative separable form: w(x. 2 2 ϕt = 2(3f + h)ϕ + 2gψ. 3 y2 + Cxt−1/3 + C 2 t1/3 . ϕ(t). The ﬁrst three equations for f . we ﬁnd that f = h + Ag. where A is an arbitrary constant. ¡ (4) where the ﬁrst equation is separable. Titov and V. Generalized separable solution (generalizes the solution of Item 3 ◦ ): w(x.2. χt = ϕ2 + ψ 2 + 2(f + h)χ. t) = − where C is an arbitrary constant. y. V. ∂ 1 ∂w ∂ 1 ∂w ∂w =α + . A. Zhurov. (1) (2) (3) It follows from (1) and (3) that ft − ht = 6(f + h)(f − h). 4◦ . 2(C − µt) µ= √ 1 + A2 . 2(C − µt) g(t) = 1 . f = 12 Φ(g) + 2 Ag.

3◦ . © 2004 by Chapman & Hall/CRC . D. t) = − A sinh θ(t) A 1 − cosh θ(t) 1 + cosh θ(t) cosh(µx + ξ0 ) + sA sin(µy + η0 ) 2 sinh θ(t) 2 sinh θ(t) . ∆Θ − AeΘ = 0. where A. Ae−x sin y + B 2A2 αt + C . w(x. Kazenin (1998). and the function Θ(x. y. y. and λ are arbitrary constants. Vyazmin. −1 A sinh θ(t) 1 + cosh θ(t) 1 − cosh θ(t) cosh(µx + ξ0 ) + sA sin(µy + η0 ) 2 sinh θ(t) 2 sinh θ(t) . y. R. y.2. −1 w(x. y. Dorodnitsyn. t) = w(x. Knyazeva. λ(ξ + A) ξ = k1 x + k2 y + λt. see 5. k1 . A. B. Svirshchevskii (1983). B. Solutions: 2 2 α(k1 + k2 ) . sinh2 (αt + C) + (y + B)2 cosh2 (αt + C) sin(µy + η0 ) A + αµ2 t § 1 + B(A + αµ2 t)eµx A + αµ2 t § −1 . −1 w(x. t) = A coth θ(t) + B sinh θ(t) eµx § A sin(µy + η0 ) sinh θ(t) −1 . y. Other exact solutions: w(x. e2x cos2 Ae−x sin y + B where A. t) = (x + A)2 2 sinh(αt + C) cosh(αt + C) . t) = A cot θ(t) + B sin θ(t) eµx § A sin(µy + η0 ) sin θ(t) w(x. t) = 2αt + B . . where A and B are arbitrary constants. A. A. Traveling-wave solutions: w(ξ) = − w(ξ) = 2◦ . V. Zhurov. Polyanin. and C are arbitrary constants. t) = e2x cosh2 C − 2A2 αt . t) = w(x. For solutions of this equation. (sin y + Aex )2 2A2 αt + C . and D. y) is a solution of the stationary equation ∂2 ∂2 + . ∆= ∂x2 ∂y 2 which is encountered in combustion theory. y. t) = A cos θ(t) § A A § 1 + sin θ(t) 1 − sin θ(t) cosh(µx + ξ0 ) + sA sin(µy + η0 ) 2 cos θ(t) 2 cos θ(t) . I. −1 w(x. 4◦ . λξ 2 2 αA(k1 + k2 ) −1 A + B exp . y. A. y. −1 w(x. t) = A cos θ(t) A § 1 − sin θ(t) 1 + sin θ(t) cosh(µx + ξ0 ) + sA sin(µy + η0 ) 2 cos θ(t) 2 cos θ(t) . y. y. V.y) . t) = (Aαt + B)eΘ(x. A. I. y. w(x. The exact solutions speciﬁed in Item 2◦ are special cases of a more general solution having the form of the product of two functions with different arguments: w(x.1.144 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 1◦ . and S. ¦ ¤¥ References: V. k2 . t) = e2x sinh2 Ae−x sin y + B w(x. y.1. t) = − w(x.

t) = C1 + rϕ (ϕ ln r − 1) w(r. this equation can be reduced to ∆u + A(n + 1)u n+1 = 0. k1 . and λ are arbitrary constants. where C1 . ∂x ∂x ∂y ∂y If n ≠ −1.2.3. t) = where r = © ¤¨ ϕ= λ .3 with β = 0: ∂U = αU ∆U − α ∂t 4. and τ0 are arbitrary constants. ◦ 6 . in the above relations. x sin β + y cos β. This is a special case of equation 2.4. 1995). y.3 with f (w) = αw n .1. Then the functions w2 = w(x cos β − y sin β. . one can obtain another group of solutions (not written out here). are also solutions of the equation. λw + C1 where C1 . t). y). A and B are arbitrary constants. A. Multiplicative separable solution: w(x. y. Here. C2 t + C5 ). and s is a parameter that admits the values 1 or −1 (the ﬁrst solution was indicated by Pukhnachov. Aristov (1999). and λ are arbitrary constants. 4◦ .4. w(r. t) = f (t)Θ(x. 1 ∆= ∂2 ∂2 + . t) is a solution of the equation in question. αt + C2 x2 + y 2 . ξ0 . 2◦ . Suppose w(x. By swapping the variables. and λ are arbitrary constants. y) is a solution of the two-dimensional stationary equation ∂Θ ∂ ∂Θ ∂ Θn + Θn + AΘ = 0. N. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 145 where θ(t) = αµ2 At + τ0 . 3◦ . The plus or minus signs can be chosen arbitrarily. ∂w =α ∂ wn ∂w + ∂ wn ∂w ∂U ∂x 2 + ∂U ∂y 2 . where n can be integer. η0 . Reference: S. µ. w1 = C1 −2/n C2 1/n w( C1 x + C3 . and the function Θ(x. C1 y + C4 . © 2004 by Chapman & Hall/CRC . f (t) = (Aαnt + B)−1/n . Traveling-wave solution in implicit form (generalizes the solution of Item 2 ◦ ): 2 2 α(k1 + k2 ) wn dw = k1 x + k2 y + λt + C2 . . x y. where C. C1 . k2 . . and negative. C. ∂x2 ∂y 2 u = Θn+1 . ∂t ∂x ∂x ∂y ∂y This is a two-dimensional heat and mass transfer equation with power-law temperature-dependent thermal conductivity (diffusion coefﬁcient). Traveling-wave solution: w= nλ(k1 x + k2 y + λt + C) 2 2 α(k1 + k2 ) 1/n . k2 . rλ + Ceαt λϕrϕ−2 . . 5◦ . B.1. C5 and β are arbitrary constants. Solutions with axial symmetry: λ2 rλ−2 . fractional. The transformation w = 1/U leads to an equation of the form 2. 1◦ . k1 . . C2 . C2 .

5. η). z2 ) is determined by the differential equation 2βV − βn1 z1 ∂V ∂ ∂V ∂V − βn2 z2 = a1 V n1 ∂z1 ∂z2 ∂z1 ∂z1 + a2 ∂ ∂V V n2 . y. and S. λw + C1 where C1 . ξ = xt− 2 (kn1 +1) . There are “two-dimensional” solutions of the following forms: w(x. y. η2 = yt− nβ+1 2 . A. 1 where k is an arbitrary constant and the function U (ξ. Svirshchevskii (1983). and β are arbitrary constants. y. k2 . z2 = ye−βn2 t . t) = tk U (ξ. are also solutions of the equation (the plus or minus signs can be chosen arbitrarily). Titov and V. and S. Ibragimov (1994). M. A. S. R. t) = H(ξ1 . k2 .5. References: V.5. where β is an arbitrary constant and the function V (z1 . βi . C1 . η) is determined by the differential equation kU − 1 (kn1 + 1)ξ 2 ∂ ∂U ∂U 1 ∂U − 2 (kn2 + 1)η = a1 U n1 ∂ξ ∂η ∂x ∂x + a2 ∂ ∂U U n2 ∂y ∂y . k1 . Then the functions w1 = A2 w( A−n1 Bx + C1 . V. t).4. y. I. Dorodnitsyn. Samarskii and I. ξ2 ). ξ2 = k2 y + λ2 t. C2 . Dorodnitsyn. t). I. λ2 . Pukhnachov (1995). ζ1 = xe−βnt . V. λ1 . ¤ ¤ ¤ References: V. w(x. z = k1 x + k2 y. B 2 t + C3 ). where the αi . © 2004 by Chapman & Hall/CRC . S. R. where A. Reference: A. Suppose w(x. where k1 .3. C2 . η2 ). ζ2 ). 1 η = yt− 2 (kn2 +1) . and λ are arbitrary constants. y. t) = e2βt V (z1 . “Two-dimensional” solution: w(x. r = x2 + y 2 . w(x. y. 5. J. ◦ 3 . 4◦ . Knyazeva. Knyazeva. Traveling-wave solution in implicit form: 2 2 a 1 k 1 w n1 + a 2 k 2 w n2 dw = k1 x + k2 y + λt + C2 . Sobol’ (1963). t) = F (ξ1 . t) = G(r. V.5 and 2. A. z2 ). Ustinov (1985). ∂z2 ∂z2 5◦ . w(x. ∂ ∂w ∂ ∂w ∂w = a1 w n1 + a2 w n2 . ξ1 = α1 x + β1 y + γ1 t. V.6 for the case of two space variables. y. y. and γi are arbitrary constants. ξ1 = k1 x + λ1 t. See also equations 2. and C3 are arbitrary constants. 6◦ . H. King (1993). A−n2 By + C2 .5. N. 1◦ . 2◦ . There is a “two-dimensional” solution of the form w(x. R. ζ2 = ye−βnt . t) = e2βt V (ζ1 . Svirshchevskii (1983). t) = F (z. z1 = xe−βn1 t . t) is a solution of the equation in question.146 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 5◦ . t) = tβ U (η1 .4 with f (w) = a1 wn1 and g(w) = a2 wn2 . η1 = xt− nβ+1 2 . w(x. ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2. ξ2 = α2 x + β2 y + γ2 t. “Two-dimensional” solution: w(x. B. A. ξ2 ). y.

C2 t + C5 ) + 2 c(1 − C1 ) . 3◦ . k2 . 2◦ . and h = h(x. t)y + G(x. t). the subscripts denote partial derivatives. g = g(x. are also solutions of the equation (the plus or minus signs can be chosen arbitrarily). . z = y + bC1 x2 + bC2 x + C3 t. k1 . gt = agxx + 6bf g. Solution: 2 w = u(z) − 4abC1 x2 − 4abC1 C2 x. t)y + h(x. t). Generalized separable solution quadratic in y: w = f (x. With appropriate translations in both variables. Then the functions 2 −2 w1 = C1 w( C2 x + C3 . 4◦ . and λ are arbitrary constants.3. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 147 2. C1 C2 y + C4 . Given F = F (x. ∂t ∂x (1) (2) Equation (1) is a linear homogeneous heat equation. one can reduce this equation to a homogeneous one.1. Generalized separable solution linear in y (a degenerate solution): w = F (x. Suppose w(x. which can be integrated by quadrature. where C1 . C2 .1. equation (2) can be treated as a linear nonhomogeneous heat equation. C5 are arbitrary constants. t) is a solution of this equation. . ht = ahxx + bg 2 + 2bf h + 2cf . Equations of the Form ∂w = ∂ f (w) ∂w + ∂ g(w) ∂w +h(w) ∂t ∂x ∂x ∂y ∂y 1. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). and C3 are arbitrary constants and the function u(z) is determined by the ﬁrst-order ordinary differential equation 2 2 (bu + c + ab2 C2 )uz + (2abC1 − C3 )u = 8a2 bC1 z + C4 . t). For these equations. © 2004 by Chapman & Hall/CRC . where C1 . C2 . Here. where the functions F and G are determined by solving the one-dimensional equations ∂2F ∂F − a 2 = 0. 1◦ . Traveling-wave solution in implicit form: 2 2 2 2 bk2 w + (ak1 + ck2 − C1 bk2 ) ln |w + C1 | = λ(k1 x + k2 y + λt) + C2 . 2 bC1 where C1 . y. ∂w ∂t =a ∂2w ∂x2 + ∂ ∂y (bw + c) ∂w ∂y . . 5◦ . t).2. t) are determined by the system of differential equations ft = afxx + 6bf 2 . ∂t ∂x ∂2G ∂G − a 2 = bF 2 . . where the functions f = f (x. t)y 2 + g(x. t).

β where C is an arbitrary constant. ∂w = ∂ 1 ∂w + ∂ 1 ∂w ∂w = ∂ (αw + β) ∂w + ∂ (αw + β) ∂w 1◦ . In Zhuravlev (2000).1. a nonlinear superposition principle is presented that allows the construction of complicated multimodal solutions of the original equation. y. t). b where the function θ(x.148 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 6◦ . τ ).3: ∂ ∂u =α ∂τ ∂x 1 ∂u u ∂x + ∂ ∂y 1 ∂u u ∂y .2 for U = U (x.1. t) − c . t) − 2 2 aC1 + cC2 . © 2004 by Chapman & Hall/CRC .2. y.1. some exact solutions are also speciﬁed there.2. “Two-dimensional” solution: 2 w = V (η. t) is determined by a differential equation of the form 1.2. “Two-dimensional” solution: w = |y + C|1/2 θ(x. ∂t ∂x 7◦ . t). t) = eβt u(x. The substitution U = αw + β leads to an equation of the form 2. 2◦ . . ∂t ∂x ∂x ∂y ∂y This is a two-dimensional heat and mass transfer equation with a linear temperature-dependent thermal conductivity (diffusion coefﬁcient). where C1 and C2 are arbitrary constants and the function V (η. “Two-dimensional” solution: w = U (ξ. t) − 4abC1 x2 − 4abC1 C2 x. The transformation w(x.1. The substitution U = αw + β leads to an equation of the form 2. η = y + bC1 x2 + bC2 x. t) is determined by the linear heat equation ∂2θ ∂θ =a 2. . y.1: ∂U ∂U 2 ∂ U . 3. y. τ =C− 1 −βt e . t) is determined by the differential equation ∂V ∂ ∂V 2 ∂V 2 = (bV + c + ab2 C2 ) + 2abC1 − 8a2 bC1 . leads to a simpler equation of the form 2. ∂ ∂w =α ∂t ∂x 1 ∂w w ∂x + ∂ ∂y 1 ∂w w ∂y + βw. 4. 2 bC2 ξ = C1 x + C2 y. = bC2 ∂t ∂ξ ∂ξ 8◦ .3 for U = U (x.10. ∂t ∂η ∂η ∂η 2. where C1 and C2 are arbitrary constants and the function U (ξ. ∂t ∂x αw + β ∂x ∂y αw + β ∂y This is a two-dimensional heat and mass transfer equation with a hyperbolic temperature-dependent thermal conductivity (diffusion coefﬁcient).

y. V. Ibragimov (1994). Suppose w(x. Then the functions 1. leads to a simpler equation of the form 2. and B3 are arbitrary constants. N. A. ∂w ¤ References: V. “Two-dimensional” solution: w(x. N. y. ∂w =α ∂ wn ∂w + ∂ wn ∂w + βw. Galaktionov and S. . 2◦ . t). N. ∂w 6. B2 . Kurdyumov. V. References: V. η) is determined by the differential equation n1 − k + 1 ∂F n2 − k + 1 ∂F ∂ ∂F α F +α ξ +α η = a1 F n1 1−k 2(k − 1) ∂ξ 2(k − 1) ∂η ∂ξ ∂ξ ¤ + a2 ∂ ∂F F n2 ∂η ∂η + bF k . t) is a solution of this equation. where the function F = F (ξ. P.1. A. where λ1 and λ2 are arbitrary constants. are also solutions of the equation.2. Knyazeva. 5. H. A. η). ∂w = aw ∂2w + ∂2w where C1 . I. ∂t ∂x ∂x ∂y ∂y The transformation (C is an arbitrary constant) w(x. y. Dorodnitsyn. t). A. and S. Dorodnitsyn. Then the functions w1 = A2 w( Ak−n1 −1 x + B1 . A. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 149 ∂ 1 ∂w ∂ 1 ∂w =α + + βw2 . t) = (αt + β) 1−k F (ξ. where A. C1 C2 t + C5 ). y. ∂t ∂x ∂x ∂y ∂y 1◦ . t) = u(z). 2 w1 = C1 w(C2 x + C3 . .1. τ= 1 βnt e +C βn + ∂ ∂U Un ∂y ∂y . Svirshchevskii (1988). Suppose w(x. R. © 2004 by Chapman & Hall/CRC . Dimova. . ∂t ∂x w ∂x ∂y w ∂y The substitution w = 1/U leads to an equation of the form 2. Ibragimov (1994). t) = eβt U (x. t) is a solution of this equation. S. C2 y + C4 . Samarskii. A. I.4. and S. H. Traveling-wave solution: w(x. R. are also solutions of the equation (the plus or minus signs can be chosen arbitrarily). x sin β + y cos β. z = t − λ1 x − λ2 y. τ ). y. 1 2 3◦ . M. 1 n1 −k+1 2(k−1) n2 −k+1 2(k−1) ξ = x(αt + β) .1. y.1. C5 and β are arbitrary constants. B1 . η = y(αt + β) .4: ∂ ∂U ∂U =α Un ∂τ ∂x ∂x 7. . A2k−2 t + B3 ).3 for U = U (x. 2. w2 = w(x cos β − y sin β. ∂t ∂x2 ∂y 2 1◦ . Svirshchevskii (1983). and the function u = u(z) is determined by the ordinary differential equation uz = [(a1 λ2 un1 + a2 λ2 un2 )uz ]z + buk . Bakirova.4.2. Ak−n2 −1 y + B2 . ∂ ∂w ∂ ∂w = a1 w n1 + a2 w n2 + bwk . S. Other Equations . y. Posashkov (1989).

one can solve the second equation in (2). y.150 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 2◦ . C2 . y. The equation has “two-dimensional” solutions with following the forms: w(x. Multiplicative separable solution: w(x. The ﬁrst equation in (2) is independent of g(t) and is separable. y. © 2004 by Chapman & Hall/CRC . β z = k1 x + k2 y. The functions f (t) and g(t) are determined by the autonomous system of ordinary differential equations ft = γf 2 + δf + ε. . t) = f (t) + g(t)Θ(x. t) = t U (η1 . C7 − 2a(C1 + C3 )t dw = k1 x + k2 y + λt + C2 . y). r= x2 + y 2 . t). k1 . η2 = y 2 t−β−1 . Θ(x. t) = G(r. t) = H(ξ1 . w(x. C7 are arbitrary constants. ξ2 = k2 y + λ2 t. ζ2 ). y. η1 = x2 t−β−1 . and the function Θ = Θ(x. t). . λ2 . y. y. and λ are arbitrary constants. w(x. ξ1 = k1 x + λ1 t. k2 . y) satisﬁes the two-dimensional Poisson equation ∂2 ∂2 + . Traveling-wave solution in implicit form: 2 2 a(k1 + k2 ) ϕ(t) = 1 . t) = F (z. η2 ). ∂x2 ∂y 2 (1) where κ = γ/β (β ≠ 0). w(x. which is linear in g(t). ζ1 = xe−t . w(x. On ﬁnding f (t). k2 . The equation admits a more general solution in the form of the product of functions with different arguments: Θ(x. For solutions of the linear equation (1). w(x. t) = e2t V (ζ1 . 3◦ . ∆= ∂x2 ∂y 2 For solutions of this linear equation. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). λ ln |w| + C1 where C1 . y. t) = A + Bt where A and B are arbitrary constants. y. where C1 . λ1 . 4◦ . ∆= ∂2 ∂2 + . . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). a∆Θ + B = 0. where k1 . Here. w(x. 5◦ . ξ2 ). . ζ2 = ye−t . 2. (2) gt = (γf + δ − ακ)g. y) . y) is any solution of the two-dimensional Helmholtz equation ∆Θ + κΘ = 0. ∂w ∂t ∂x2 ∂y 2 Generalized separable solution: = (α + βw) ∂2w + ∂2w + γw2 + δw + ε. t) = ϕ(t)(C1 x2 + C2 xy + C3 y 2 + C4 x + C5 y + C6 ). and β are arbitrary constants.

−δ µ . A. g(t) = exp f (t) = − 2γ C1 + γt C1 + γt 5◦ . For γ ≠ 0 and δ 2 − 4γε = µ2 > 0 (µ > 0). y. C1 . δ g(t) = C2 e−ακt . A w(x. θ(t) = αµ2 At + τ0 . A w(x. . w(x. A. Ibragimov (1994). sin(µy + η0 ) . t). For γ ≠ 0 and δ 2 − 4γε = 0. and C2 are arbitrary constants. 1 C2 δ − . C4 and β are arbitrary constants. . η0 . one can obtain another group of solutions (not written out here). y. Posashkov (1989). t) = A cot θ(t) + B sin θ(t) eµx A sin θ(t) 1 − sin θ(t) 1 + sin θ(t) A cosh(µx + ξ0 ) + sA sin(µy + η0 ). t) = sinh θ(t) 2 sinh θ(t) 2 sinh θ(t) 1 − cosh θ(t) 1 + cosh θ(t) A cosh(µx + ξ0 ) + sA sin(µy + η0 ).1. Solutions for β = 0: 1 + B(A + αµ2 t)eµx w(x. t) = A + αµ2 t . ν. are also solutions of the equation. where µ. f (t) = ¤ s1. y. References: V. where C1 . 2γ 2 g(t) = C2 exp cos 1 2 δ − ακ 1 2 µt + C1 t . Suppose w(x. y. H. A w(x.2. 2◦ . g(t) = C2 e(δ−ακ)t . f (t) = 1 + C1 eµt 1 + C1 eµt 4◦ . For γ = δ = 0. © 2004 by Chapman & Hall/CRC 3◦ . y. and τ0 are arbitrary constants. y. y. ε . C1 y + C3 . ξ0 . For γ = 0 and δ ≠ 0. t) = − cos θ(t) 2 cos θ(t) 2 cos θ(t) 1 + cosh θ(t) 1 − cosh θ(t) A cosh(µx + ξ0 ) + sA sin(µy + η0 ). By swapping the variables. t) is a solution of this equation. ∂w = αw ∂2w –α ∂w + ∂w 2 2◦ . C2 s1 + s2 C1 eµt . B. Generalized separable solution: 1 w(x. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 151 The functions f (t) and g(t) have different forms depending on the values of the equation parameters. f (t) = C1 eδt − f (t) = C1 + εt. y. . y. g(t) = e−(γs2 +ακ)t . in the above relations. 1◦ . C1 t + C4 ). w(x. µ tan 2γ ∂2w 1 2 µt + C1 − δ . Below are ﬁve possible cases. For γ ≠ 0 and δ 2 − 4γε = −µ2 < 0 (µ > 0). + −1 w1 = C1 w(C1 x + C2 . and s is a parameter that can assume the values 1 or −1. t) = − sinh θ(t) 2 sinh θ(t) 2 sinh θ(t) where A. – β. x sin β + y cos β. C1 and C2 are arbitrary constants. w2 = w(x cos β − y sin β. N. 2γ 3◦ . t) = cos θ(t) 2 cos θ(t) 2 cos θ(t) 1 + sin θ(t) 1 − sin θ(t) A cosh(µx + ξ0 ) + sA sin(µy + η0 ).2 = 1 2δ − ακ t . . t) = C1 − βt + C2 exp α(µ2 + ν 2 ) C1 t − 2 βt2 eµx+νy . Galaktionov and S. A w(x. A + αµ2 t sin(µy + η0 ) . Then the functions 3. ∂t ∂x2 ∂y 2 ∂x ∂y 1◦ . µ. x y. t) = A coth θ(t) + B sinh θ(t) eµx A sinh θ(t) sin(µy + η0 ) .

If D2 = 0 (D1 ≠ 0). 1 2 gt = ανf g. g(t). if ϕxx = νϕ and ψyy = −νψ. A2 . y. The order of this system can be reduced by 2. where ν is an arbitrary constant. then a4 = −B1 . C2 . If D1 = 0 (D2 ≠ 0). y. and u(t) are determined by the following system of ordinary differential equations (s = sign ν): 2 2 ft = −4αν(A2 − sA2 )g 2 + 4αν(B1 + sB2 )h2 − β. C1 C2 ≠ 0. ψyy = −4νψ. (ν = −µ2 < 0). we have ϕ(x) = A1 cosh µx + A2 sinh µx. (1) Here. and a4 are deﬁned by 2 2 2 2 2 D2 − sD2 C 2 − sC2 D2 + sD2 C1 + sC2 . a2 = 1 . C1 + (A2 − sA2 ) 1 2 5◦ . B1 ≠ 0. and D1 D2 ≠ 0. then a2 = D1 D2 /B2 . A2 . then a3 = −A1 . The functions f (t). and B2 are arbitrary constants. then one should set a1 = C1 C2 /A2 . θxx = νθ. ht = −ανhΦ(h). a1 = © 2004 by Chapman & Hall/CRC ! . then the system becomes f = Φ(h). (2) For ϕxx = 4νϕ. A1 . solutions in explicit form may be obtained in some cases (see Item 3◦ ). In particular. B2 . If A1 = 0 (A2 ≠ 0). The arbitrary constants A1 . ht = −ανf h. 2 1 2 2 gt = −4ανf g + ανa1 (D1 + sD2 )u2 . where Φ(h) = g = C2 /h. then a4 = B1 . If B1 = 0 (B2 ≠ 0). 2 2 2B1 D1 D2 = B2 (D1 − sD2 ). h(t). and D2 are related by the two constraints 2 2 2A1 C1 C2 = A2 (C1 + sC2 ). then a3 = A1 . t) = f (t) + g(t)ϕ(x) + h(t)ψ(y). t) = f (t) + g(t)ϕ(x) + h(t)ψ(y) + u(t)θ(x)χ(y). a3 . If C2 = 0 (C1 ≠ 0). 2A1 2B1 C1 C2 D1 D2 provided A1 ≠ 0. There is a generalized separable solution of the form w(x. There is a generalized separable solution of the form w(x. a 4 = B2 1 . 2 h αν C1 and C2 are arbitrary constants. ψ(y) = B1 cos µy + B2 sin µy ϕ(x) = A1 cos µx + A2 sin µx. B1 . If C1 = 0 (C2 ≠ 0). g(t). C1 . where s = sign ν. 2 2β C2 2 2 + ln |h| + (B1 + sB2 )h2 . For β = 0. ψ(y) = B1 cosh µy + B2 sinh µy (ν = µ2 > 0). and χyy = −νχ.152 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 4◦ . where ν is an arbitrary constant. D1 . The coefﬁcients a1 . 2 2 ht = 4ανf h − ανa2 (C1 − sC2 )u2 . one can set in (2) for ν = µ2 > 0 ϕ(x) = A1 cosh 2µx + A2 sinh 2µx ψ(y) = B1 cos 2µy + B2 sin 2µy θ(x) = C1 cosh µx + C2 sinh µx χ(y) = D1 cos µy + D2 sin µy for ν = −µ2 < 0 ϕ(x) = A1 cos 2µx + A2 sin 2µx ψ(y) = B1 cosh 2µy + B2 sinh 2µy θ(x) = C1 cos µx + C2 sin µx χ(y) = D1 cosh µy + D2 sinh µy The functions f (t). and h(t) in (1) are determined by the system of ordinary differential equations 2 2 ft = αν(A2 − sA2 )g 2 − αν(B1 + sB2 )h2 − β. a2 . a 3 = A2 1 . B1 . ut = −2αν(a3 g − a4 h)u.

A. 3◦ .1. Zaitsev (2002). α References: V. C1 − βt (C1 − βt)2 where C1 and C2 are arbitrary constants. © 2004 by Chapman & Hall/CRC . and D. w(x. and the function F (x. y. A. where k1 . ∂x ∂y 1◦ . C1 t + C4 ). w2 = w(x cos β − y sin β. y + C3 . y. B − βt g(t) = β . There is a generalized separable solution of the form w(x. Galaktionov and S. . Here. . y. x sin β + y cos β. B. κ = sign C2 .2. |∇w|2 = ∂w ∂w + . are also solutions of the equation (the plus or minus signs can be chosen arbitrarily). y. # ¤" ht = α(2f h − 2h2 − g 2 ).3: V. t) = f (t) + g(t)Θ(x. (B − βt)[A + C(B − βt)] where A. −1 η = C4 x + C5 y + C6 t. y). the functions f (t) and g(t) are given by f (t) = 1 . η). Posashkov (1989). The equation admits a traveling-wave solution: w = w(k1 x + k2 y + λt). N. “Two-dimensional” generalized separable solution: w(x. t) is a solution of the equation in question. y). . h(t). and λ are arbitrary constants. Here. Zhurov. ξ = C1 x + C1 y + C3 t. χt = 2α(f + h)χ − β. In the special case ϕ(t) = ψ(t) ≡ 0. A. ∂x ∂x ∂y ∂y α 4◦ . Polyanin and V. t) = (C1 t + C2 ) V (x. References for equation 2. y). I. D. Vyazmin. α µ= A . the functions f (t). Galaktionov and S. y) is any solution of the stationary equation ∂F ∂ ∂F β ∂ |∇F | + |∇F | + κF 2 = 0. A. A. “Two-dimensional” generalized separable solution: w(x. k2 . A. gt = −2αg(f + h). Then the functions w1 = C1 w( x + C2 . and C are arbitrary constants. t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t). A.4. g(t). Polyanin. D. Posashkov (1989). A. Suppose w(x.1. A. There are “two-dimensional” solutions of the following forms: w(x. and the function Θ(x. H. t) = U (ξ. . t). 2◦ . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 153 6◦ . κ= β . V. 4. Ibragimov (1994). y) is any solution of the stationary equation $ ∂Θ ∂ |∇Θ| ∂x ∂x # ¤" + ∂ ∂Θ |∇Θ| ∂y ∂y κΘ2 = µΘ. and χ(t) are determined by the autonomous system of ordinary differential equations ft = α(2f h − 2f 2 − g 2 ). C4 and β are arbitrary constants. ∂w ∂t =α ∂ ∂x |∇w| 2 ∂w ∂x + 2 ∂ ∂y |∇w| ∂w ∂y + βw2 . 7◦ . y. t) = C2 1 + F (x. Kazenin (1998). $ $ where C1 . which can be completely integrated. F. y.

154

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

**2.2. Equations with Two Space Variables Involving Exponential Nonlinearities
**

2.2.1. Equations of the Form ∂w = ∂ f (x) ∂w + ∂ g(y) ∂w +aeλw ∂t ∂x ∂x ∂y ∂y

1. ∂ ∂w ∂w = axn ∂t ∂x ∂x + ∂ ∂w by m ∂y ∂y + ceλw .

This is a special case of equation 2.4.2.1 with f (w) = ceλw . 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the function 1 ln C1 , λ where C1 and C2 are arbitrary constants, is also a solution of the equation. w1 = w C12−n x, C12−m y, C1 t + C2 + 2◦ . There are “two-dimensional” solutions of the following forms: w(x, y, t) = U (r, t), r2 = y 2−m x2−n + ; 2 a(2 − n) b(2 − m)2 z1 = xt n−2 , + ceβw .

1 1 1

w(x, y, t) = V (z1 , z2 ) − 2. ∂w ∂t = ∂ ∂x axn ∂w ∂x + ∂ ∂y

1 ln t, λ ∂w ∂y

z2 = yt m−2 .

1

beλy

This is a special case of equation 2.4.2.3 with f (w) = ceβw . 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the function w1 = w C12−n x, y −

1

1 1 ln C1 , C1 t + C2 + ln C1 , λ β

where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . There are “two-dimensional” solutions of the following forms: w(x, y, t) = U (r, t), r2 = e−λy x2−n + ; 2 a(2 − n) bλ2 z1 = xt n−2 ,

1

w(x, y, t) = V (z1 , z2 ) − 3. ∂w = ∂ aeβx ∂w + ∂

1 ln t, β

z2 = y +

1 ln t. λ

+ ceµw . ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2.4.2.2 with f (w) = ceµw . beλy 1 1 1 ln C1 , y − ln C1 , C1 t + C2 + ln C1 , β λ µ

∂w

1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the function w1 = w x −

where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . There are “two-dimensional” solutions of the following forms: w(x, y, t) = U (r, t), r2 = e−βx e−λy + ; aβ 2 bλ2 z1 = x + 1 ln t, β z2 = y + 1 ln t. λ

w(x, y, t) = V (z1 , z2 ) −

1 ln t, µ

© 2004 by Chapman & Hall/CRC

2.2. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES

155

**2.2.2. Equations of the Form ∂w = a ∂ eβw ∂w +b ∂ eλw ∂w +f (w) ∂t ∂x ∂x ∂y ∂y
**

1. . ∂t ∂x ∂x ∂y ∂y This is a two-dimensional nonstationary heat (diffusion) equation with exponential temperaturedependent thermal diffusivity (diffusion coefﬁcient). This is a special case of equation 2.4.3.3 with f (w) = αeµw . eµw eµw C2 1 ln 2 , µ C1 ∂w =α ∂ ∂w + ∂ ∂w

1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions w1 = w(C1 x + C3 , C1 y + C4 , C2 t + C5 ) +

w2 = w(x cos β − y sin β, x sin β + y cos β, t), where C1 , . . . , C5 and β are arbitrary constants, are also solutions of the equation. 2◦ . Traveling-wave solution in implicit form:

2 2 α(k1 + k2 )

eµw dw = k1 x + k2 y + λt + C2 , λw + C1

where C1 , C2 , k1 , k2 , and λ are arbitrary constants. 3◦ . “Two-dimensional” additive separable solution: w(x, y, t) = f (t) + 1 ln Θ(x, y), µ f (t) = − 1 ln(Aαt + B). µ

Here, A, B, and µ are arbitrary constants, and the function Θ(x, y) is any solution of the Poisson equation ∂2 ∂2 + . ∆Θ + A = 0, ∆= 2 ∂x ∂y 2 For solutions of this linear equation, see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). 4◦ . There are “two-dimensional” solutions of the following forms: w(x, y, t) = F (z, t), w(x, y, t) = G(r, t), w(x, y, t) = G(ξ1 , ξ2 ), z = k1 x + k2 y; r= x2 + y 2 ; η2 = y 2/t; ζ2 = ye−t , ξ1 = k1 x + λ1 t, ξ2 = k2 y + λ2 t;

**w(x, y, t) = H(η1 , η2 ), η1 = x2/t, 2 w(x, y, t) = t + U (ζ1 , ζ2 ), ζ1 = xe−t , µ where k1 , k2 , λ1 , and λ2 are arbitrary constants.
**

& ¤%

Reference: V. A. Dorodnitsyn, I. V. Knyazeva, and S. R. Svirshchevskii (1983).

**. ∂t ∂x ∂x ∂y ∂y 1◦ . Suppose w(x, y, t) is a solution of this equation. Then the functions 2. eβw eλw
**

β λ 2 w1 = w(C1 C2 x + C3 , C1 C2 y + C4 , C1 t + C5 ) − 2 ln |C2 |,

'

∂w

=a

∂

∂w

+b

∂

∂w

**where C1 , . . . , C5 are arbitrary constants, are also solutions of the equation.
**

© 2004 by Chapman & Hall/CRC

156

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

2◦ . “Two-dimensional” solution: w(x, y, t) = u(θ, t), θ = c1 x + c2 y, where c1 and c2 are arbitrary constants, and the function u = u(θ, t) is determined by the differential equation ∂ ∂w ∂u = ac2 eβw + bc2 eλw . 1 2 ∂t ∂θ ∂θ 3◦ . “Two-dimensional” solution: w(x, y, t) = u(ζ1 , ζ2 ), ζ1 = k1 x + m1 t, ζ2 = k2 y + m2 t, where k1 , k2 , m1 , and m2 are arbitrary constants, and the function u = u(ζ1 , ζ2 ) is determined by the differential equation m1 ∂u ∂w ∂u 2 ∂ + m2 = ak1 eβw ∂ζ1 ∂ζ2 ∂ζ1 ∂ζ1 w(x, y, t) = U (ξ, η) + 2kt,

2 + bk2

∂ ∂w eλw . ∂ζ2 ∂ζ2 η = ye−kλt ,

4◦ . “Two-dimensional” solution: ξ = xe−kβt , where k is an arbitrary constant, and the function U = U (ξ, η) is determined by the differential equation ∂U ∂ ∂U ∂ ∂U ∂U − kλη =a eβU +b eλU . 2k − kβξ ∂ξ ∂η ∂ξ ∂ξ ∂η ∂η 5◦ . “Two-dimensional” solution: y + C2 x + C1 , z2 = √ , w(x, y, t) = V (z1 , z2 ), z1 = √ t + C3 t + C3 where C1 , C2 , and C3 are arbitrary constants, and the function V = V (z1 , z2 ) is determined by the differential equation 1 ∂V ∂ ∂w 1 ∂V − z2 =a eβw − z1 2 ∂z1 2 ∂z2 ∂z1 ∂z1 6◦ . “Two-dimensional” solution (β = 1): w(x, y, t) = u(z, t) + 2 ln |x|, ∂ 2u ∂u = aλ2 z 2 eu + beλu + λ aλz 2 eu + beλu ∂t ∂z 2

) ¤(

+b

∂ ∂w eλw . ∂z2 ∂z2

z = x−λ y,

2

where the function u = u(z, t) is determined by the differential equation ∂u ∂z + aλ(λ − 3)zeu ∂u + 2aeu . ∂z

References: V. A. Dorodnitsyn, I. V. Knyazeva, and S. R. Svirshchevskii (1983), N. H. Ibragimov (1994).

**∂ ∂w ∂ ∂w ∂w =a eλ1 w +b eλ2 w + ceβw . ∂t ∂x ∂x ∂y ∂y 1◦ . Suppose w(x, y, t) is a solution of this equation. Then the functions 3.
**

β−λ β−λ 2β w1 = w( C1 1 x + C2 , C1 2 y + C3 , C1 t + C4 ) + 2 ln |C1 |,

where C1 , . . . , C4 are arbitrary constants, are also solutions of the equation (the plus or minus signs can be chosen arbitrarily). 2◦ . There is a “two-dimensional” solution of the form w(x, y, t) = U (ξ, η) −

) ¤(

1 ln t, β

References: V. A. Dorodnitsyn, I. V. Knyazeva, and S. R. Svirshchevskii (1983), N. H. Ibragimov (1994).

© 2004 by Chapman & Hall/CRC

0

0

ξ = xt

λ1 −β 2β

,

η = yt

λ2 −β 2β

.

2.3. OTHER EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS

157

4.

∂w ∂t

=α

∂ ∂x

eµw

∂w ∂x

+

∂ ∂y

eµw

∂w ∂y

+ βeµw + γ + δe–µw .

The substitution w =

1 ln U leads to an equation of the form 2.1.4.2: µ ∂U = αU ∂t ∂ 2U ∂ 2 U + ∂x2 ∂y 2 + βµU 2 + µγU + µδ.

w2 = w(x cos β − y sin β, x sin β + y cos β, t), where C1 , . . . , C4 and β are arbitrary constants, are also solutions of the equation. 2◦ . “Two-dimensional” multiplicative separable solution: w(x, y, t) = exp(C1 e−kt )Θ(x, y), where the function Θ(x, y) is a solution of the stationary equation a ∂ 2Θ ∂ 2 Θ + ∂x2 ∂y 2 − kΘ ln Θ = 0.

This equation has a particular solution of the form Θ = exp(A1 x2 + A2 xy + A3 y 2 + A4 x + A5 y + A6 ), where the coefﬁcients Ak are determined by an algebraic system of equations. 3◦ . “Two-dimensional” solution with incomplete separation of variables (the solution is separable in the space variables x and y, but is not separable in time t): w(x, y, t) = ϕ(x, t)ψ(y, t), where the functions ϕ(x, t) and ψ(y, t) are determined by solving two independent one-dimensional nonlinear parabolic differential equations: ∂2ϕ ∂ϕ = a 2 − kϕ ln ϕ, ∂t ∂x ∂2ψ ∂ψ = a 2 − kψ ln ψ. ∂t ∂y For solutions of these equations, see 1.6.1.4 with f (t) = 0. 4◦ . There are exact solutions in the form of the product of functions representing two independent traveling waves: w(x, y, t) = ϕ(ξ)ψ(η), ξ = a1 x + b1 t, η = a2 y + b2 t, where a1 , b1 , a2 , and b2 are arbitrary constants. This solution is a special case of the solution presented in Item 3◦ .

2 ¤1

References: V. A. Dorodnitsyn, I. V. Knyazeva, and S. R. Svirshchevskii (1983), V. K. Andreev, O. V. Kaptsov, V. V. Pukhnachov, and A. A. Rodionov (1999), A. D. Polyanin and V. F. Zaitsev (2002).

© 2004 by Chapman & Hall/CRC

3

2 ¤1

References: V. A. Galaktionov and S. A. Posashkov (1989), N. H. Ibragimov (1994).

**2.3. Other Equations with Two Space Variables Involving Arbitrary Parameters
**

2.3.1. Equations with Logarithmic Nonlinearities

– kw ln w. ∂t ∂x2 ∂y 2 This is a nonstationary equation with a logarithmic source arising in heat and mass transfer theory and combustion theory. This is a special case of equation 2.4.1.1 with f (w) = −kw ln w. 1. 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions w1 = exp(C1 e−kt )w(x + C2 , y + C3 , t + C4 ), ∂w =a ∂2w + ∂2w

158 2. ∂w

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

∂ ∂w ∂ ∂w = axn + by m + cw ln w + w. ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2.4.2.1 with f (w) = cw ln w + sw and a special case of equation 2.4.2.4 in which a should be renamed c and b renamed s and then the functions f (x) and g(y) should be substituted by axn and by m , respectively. There are “two-dimensional” solutions of the following forms: 4 4 x2−n + y 2−m ; 2 a(2 − n) b(2 − m)2 w(x, y, t) = exp(Aect )G(x, y); w(x, y, t) = H1 (x, t)H2 (y, t), w(x, y, t) = F (ξ, t), ξ2 = where A is an arbitrary constant. + cw ln w + w. ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2.4.2.2 with f (w) = cw ln w + sw and a special case of equation 2.4.2.4 in which a should be renamed c and b renamed s and then the functions f (x) and g(y) should be substituted by eβx and beλy , respectively. There are exact solutions of the following forms: 4 4 −βx e + 2 e−λy ; aβ 2 bλ w(x, y, t) = exp(Aect )G(x, y); w(x, y, t) = F (ξ, t), ξ2 = w(x, y, t) = H1 (x, t)H2 (y, t), where A is an arbitrary constant.

4 4 4

3.

∂w

=

∂

aeβx

∂w

+

∂

beλy

∂w

4.

∂ ∂w ∂ ∂w ∂w = axn + beλy + cw ln w + w. ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2.4.2.3 with f (w) = cw ln w + sw and a special case of equation 2.4.2.4 in which a should be renamed c and b renamed s and then the functions f (x) and g(y) should be substituted by axn and beλy , respectively. There are exact solutions of the following forms: 4 4 x2−n + 2 e−λy ; a(2 − n)2 bλ w(x, y, t) = exp(Aect )G(x, y); w(x, y, t) = H1 (x, t)H2 (y, t), w(x, y, t) = F (ξ, t), ξ2 =

where A is an arbitrary constant.

**2.3.2. Equations with Trigonometrical Nonlinearities
**

4

1.

+ c sin(kw + ). ∂t ∂x ∂x ∂y ∂y This is a special case of equation 2.4.2.1 with f (w) = c sin(kw + s). There is an exact solution of the form axn by m w(x, y, t) = U (ξ, t), ξ2 = 4 4 x2−n + y 2−m . 2 a(2 − n) b(2 − m)2

∂w

=

∂

∂w

+

∂

∂w

© 2004 by Chapman & Hall/CRC

2.4. EQUATIONS INVOLVING ARBITRARY FUNCTIONS

159

∂t

∂x

∂x

∂y

∂y

This is a special case of equation 2.4.2.2 with f (w) = c sin(kw + s). There is an exact solution of the form w(x, y, t) = U (ξ, t), 3. ∂ ∂w ∂w = axn ∂t ∂x ∂x + ξ2 = 4 4 −βx e + 2 e−λy . aβ 2 bλ

This is a special case of equation 2.4.2.3 with f (w) = c sin(kw + s). There is an exact solution of the form w(x, y, t) = U (ξ, t), ξ2 = 4 4 x2−n + 2 e−λy . a(2 − n)2 bλ

**2.4. Equations Involving Arbitrary Functions
**

2.4.1. Heat and Mass Transfer Equations in Quiescent or Moving Media with Chemical Reactions

1. ∂w ∂t =a ∂2w ∂x2 + ∂2w ∂y 2 + f (w).

This is a two-dimensional equation of unsteady heat/mass transfer or combustion in a quiescent medium. 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions w1 = w( x + C1 , y + C2 , t + C3 ), w2 = w(x cos β − y sin β, x sin β + y cos β, t), where C1 , C2 , C3 , and β are arbitrary constants, are also solutions of the equation (the plus or minus signs in w1 can be chosen independently of each other). 2◦ . Traveling-wave solution: w = w(ξ), ξ = Ax + By + λt,

6 6

where A, B, and λ are arbitrary constants, and the function w(ξ) is determined by the autonomous ordinary differential equation a(A2 + B 2 )wξξ − λwξ + f (w) = 0. For solutions of this equation, see Polyanin and Zaitsev (1995, 2003). 3◦ . There are “two-dimensional” solutions of the following forms: w(x, y, t) = F (z, t), z = k1 x + k2 y;

**w(x, y, t) = G(r, t), r = x2 + y 2 ; w(x, y, t) = H(ξ1 , ξ2 ), ξ1 = k1 x + λ1 t, ξ2 = k2 y + λ2 t, where k1 , k2 , λ1 , and λ2 are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

5

∂ ∂w beλy ∂y ∂y

+ c sin(kw + ).

5

2.

∂w

=

∂

aeβx

∂w

+

∂

beλy

∂w

+ c sin(kw + ).

160 2.

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

∂w ∂w ∂2w ∂2w ∂w + – f (w). +a +b = ∂t ∂x ∂y ∂x2 ∂y 2 This is a two-dimensional equation of unsteady heat/mass transfer with a volume chemical reaction in a steady translational ﬂuid ﬂow. The transformation w = U (ξ, η, t), ξ = x − at, η = y − bt leads to a simpler equation of the form 2.4.1.1: ∂2U ∂2U ∂U = + − f (U ). ∂t ∂ξ 2 ∂η 2 – f (w). ∂t ∂x ∂y ∂y 2 This is a two-dimensional equation of unsteady heat/mass transfer with a volume chemical reaction in a steady translational-shear ﬂuid ﬂow. 3. ∂x2 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions w1 = w x + Cb1 eλt , y + C(λ − a1 )eλt , t , where C is an arbitrary constant and λ = λ1,2 are roots of the quadratic equation λ2 − (a1 + b2 )λ + a1 b2 − a2 b1 = 0, are also solutions of the equation. 2◦ . Solutions: w = w(z), z = a2 x + (λ − a1 )y + Ceλt , (1) ∂w + (a1 x + b1 y + c1 ) ∂w + (a2 x + b2 y + c2 ) ∂w = ∂2w + ∂2w

where λ = λ1,2 are roots of the quadratic equation (1), and the function w(z) is determined by the ordinary differential equation λz + a2 c1 + (λ − a1 )c2 wz = a2 + (λ − a1 )2 wzz − f (w). 2 3◦ . “Two-dimensional” solutions: w = U (ζ, t), ζ = a2 x + (λ − a1 )y, where λ = λ1,2 are roots of the quadratic equation (1), and the function U (ζ, t) is determined by the differential equation ∂U ∂2U ∂U + λζ + a2 c1 + (λ − a1 )c2 = a2 + (λ − a1 )2 − f (U ). 2 ∂t ∂ζ ∂ζ 2 Remark. In the case of an incompressible ﬂuid, the equation coefﬁcients must satisfy the condition a1 + b2 = 0. 4. – g(w). ∂t ∂x ∂y ∂x2 ∂y 2 This equation describes mass transfer with volume chemical reaction in an unsteady translational ﬂuid ﬂow. The transformation w = U (ξ, η, t), ξ =x− f1 (t) dt, η=y− f2 (t) dt, ∂w + f1 (t) ∂w + f2 (t) ∂w = ∂2w + ∂2w

leads to a simpler equation of the form 2.4.1.1: ∂2U ∂2U ∂U = + − g(U ). ∂t ∂ξ 2 ∂η 2

© 2004 by Chapman & Hall/CRC

2.4. EQUATIONS INVOLVING ARBITRARY FUNCTIONS

161

**2.4.2. Equations of the Form ∂w = ∂ f (x) ∂w + ∂ g(y) ∂w +h(w) ∂t ∂x ∂x ∂y ∂y
**

1. ∂w ∂ ∂w ∂ ∂w axn + by m + f (w). = ∂t ∂x ∂x ∂y ∂y This is a two-dimensional equation of unsteady heat (mass) transfer or combustion in an anisotropic case with power-law coordinate-dependent principal thermal diffusivities (diffusion coefﬁcients). Solution for n ≠ 2 and m ≠ 2: w = w(ξ, t), ξ2 = 4 4 x2−n + y 2−m , 2 a(2 − n) b(2 − m)2 4 − nm . (2 − n)(2 − m)

where the function w(ξ, t) is determined by the one-dimensional nonstationary equation ∂w ∂ 2 w A ∂w + = + f (w), ∂t ∂ξ 2 ξ ∂ξ A=

For solutions of this equation with A = 0 and various f (w), see Subsections 1.1.1 to 1.1.3 and equations 1.2.1.1 to 1.2.1.3, 1.4.1.2, 1.4.1.3, 1.4.1.7, and 1.4.1.8. 2. + f (w). ∂t ∂x ∂x ∂y ∂y This is a two-dimensional equation of unsteady heat (mass) transfer or combustion in an anisotropic case with exponential coordinate-dependent principal thermal diffusivities (diffusion coefﬁcients). Solution for β ≠ 0 and λ ≠ 0: aeβx beλy w = w(ξ, t), ξ2 = 4 4 −βx e + 2 e−λy , aβ 2 bλ ∂w = ∂ ∂w + ∂ ∂w

where the function w(ξ, t) is determined by the one-dimensional nonstationary equation ∂w ∂ 2 w 1 ∂w = − + f (w). ∂t ∂ξ 2 ξ ∂ξ 3. ∂w = ∂ axn ∂w + ∂ ∂y beλy ∂w ∂y + f (w).

∂t ∂x ∂x Solution for n ≠ 2 and λ ≠ 0:

w = w(ξ, t),

ξ2 =

4 4 x2−n + 2 e−λy , a(2 − n)2 bλ

where the function w(ξ, t) is determined by the one-dimensional nonstationary equation n 1 ∂w ∂w ∂ 2 w = + + f (w). ∂t ∂ξ 2 2 − n ξ ∂ξ 4. + aw ln w + bw. ∂t ∂x ∂x ∂y ∂y This is a two-dimensional equation of unsteady heat (mass) transfer or combustion in an anisotropic case with arbitrary coordinate-dependent principal thermal diffusivities (diffusion coefﬁcients) and a logarithmic source. g(y) 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the function w1 = exp(C1 eat )w(x, y, t + C2 ), where C1 and C2 are arbitrary constants, is also a solution of the equation.

© 2004 by Chapman & Hall/CRC

∂w

=

∂

f (x)

∂w

+

∂

∂w

162

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

2◦ . “Two-dimensional” multiplicative separable solution: w(x, y, t) = exp(C1 eat )U (x, y), where the function U (x, y) is determined by the stationary equation ∂U ∂ ∂U ∂ f (x) + g(y) + aU ln U + bU = 0. ∂x ∂x ∂y ∂y 3◦ . “Two-dimensional” solution with incomplete separation of variables (the solution is separable in the space variables x and y, but is not separable in time t): w(x, y, t) = ϕ(x, t)ψ(y, t), where the functions ϕ(x, t) and ψ(y, t) are determined from the two independent one-dimensional nonlinear parabolic differential equations ∂ϕ ∂ ∂ϕ f (x) + aϕ ln ϕ + C(t)ϕ, = ∂t ∂x ∂x ∂ ∂ψ ∂ψ = g(y) + aψ ln ψ + bψ − C(t)ψ, ∂t ∂y ∂y and C(t) is an arbitrary function.

8 ¤7

Reference: A. D. Polyanin (2000).

**2.4.3. Equations of the Form ∂w = ∂ f (w) ∂w + ∂ g(w) ∂w +h(t,w) ∂t ∂x ∂x ∂y ∂y
**

1. + f (t)w. ∂t ∂x ∂x ∂y ∂y 1◦ . Suppose w(x, y, t) is a solution of this equation. Then the functions n n −2 w1 = C1 w C1 x + C2 , C 1 y + C 3 , t , wn wn

9 9

∂w

=a

∂

∂w

+

∂

∂w

w2 = w(x cos β − y sin β, x sin β + y cos β, t), where C1 , C2 , C3 , and β are arbitrary constants, are also solutions of the equation (the plus or minus signs in w1 can be chosen independently of each other). 2◦ . Multiplicative separable solution: w(x, y, t) = exp f (t) dt Θ(x, y)

1 n+1

,

(1)

where the function Θ(x, y) is a solution of the Laplace equation ∂2 ∂2 + . ∆Θ = 0, ∆≡ ∂x2 ∂y 2 For solutions of this linear stationary equation, see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). 3◦ . Multiplicative separable solution: w(x, y, t) = ϕ(t) Θ(x, y) n+1 , where the function ϕ(t) is a solution of the Bernoulli equation ϕt − f (t)ϕ + Aaϕn+1 = 0, A is an arbitrary constant, and the function Θ(x, y) is determined by the stationary equation 1 ∂2 ∂2 + . ∆≡ ∆Θ + A(n + 1)Θ n+1 = 0, ∂x2 ∂y 2 The general solution of equation (3) is given by ϕ(t) = exp F (t) Aan exp nF (t) dt + B

−1/n 1

(2) (3)

,

F (t) =

f (t) dt,

**where B is an arbitrary constant.
**

© 2004 by Chapman & Hall/CRC

2.4. EQUATIONS INVOLVING ARBITRARY FUNCTIONS

163

4◦ . The transformation w(x, y, t) = F (t)U (x, y, τ ), τ= F n (t) dt, F (t) = exp f (t) dt

leads to a simpler equation of the form 2.1.2.4: ∂ ∂U ∂U =a Un ∂τ ∂x ∂x 2. ∂w =a ∂ eµw ∂w + ∂ eµw ∂w + ∂ ∂U Un ∂y ∂y .

**+ f (t). ∂t ∂x ∂x ∂y ∂y 1◦ . Suppose w(x, y, t) is a solution of this equation. Then the functions 2 ln |C1 |, µ w2 = w(x cos β − y sin β, x sin β + y cos β, t), w1 = w C1 x + C2 , C 1 y + C 3 , t −
**

@

where C1 , C2 , C3 , and β are arbitrary constants, are also solutions of the equation. 2◦ . Additive separable solution: w(x, y, t) = ϕ(t) + 1 ln Θ(x, y), µ (1)

where the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt + A(a/µ) exp(µϕ) − f (t) = 0, and the function Θ(x, y) is a solution of the two-dimensional Poisson equation ∆Θ + A = 0, ∆≡ ∂2 ∂2 + . ∂x2 ∂y 2 (2)

The general solution of equation (1) is given by ϕ(t) = F (t) − 1 ln B + Aa µ exp µF (t) dt , F (t) = f (t) dt. (3)

For solutions of the linear stationary equation (2), see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). Note that equations (1), (2) and relation (3) involve arbitrary constants A and B. 3◦ . The transformation w(x, y, t) = U (x, y, τ ) + F (t), τ= exp[µF (t)] dt, F (t) = f (t) dt,

leads to a simpler equation of the form 2.2.2.1: ∂ ∂U ∂U =a eµU ∂τ ∂x ∂x 3. ∂w = ∂ f (w) ∂w + ∂ f (w) ∂w + ∂ ∂U eµU ∂y ∂y .

**. ∂t ∂x ∂x ∂y ∂y This is a two-dimensional nonlinear heat and mass transfer equation for an anisotropic medium.
**

2 w1 = w(C1 x + C2 , C1 y + C3 , C1 t + C4 ), w2 = w(x cos β − y sin β, x sin β + y cos β, t),

1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions

**where C1 , . . . , C4 and β are arbitrary constants, are also solutions of the equation.
**

© 2004 by Chapman & Hall/CRC

164

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

2◦ . Traveling-wave solution in implicit form: f (w) dw 2 2 = k1 x + k2 y + λt + C2 , (k1 + k2 ) λw + C1 where C1 , C2 , k1 , k2 , and λ are arbitrary constants. x2 + y 2 , t where the function U = U (ξ) is determined by the ordinary differential equation w(x, y, t) = U (ξ), ξ= [ξf (U )Uξ ]ξ + 1 ξUξ = 0. 4 4◦ . “Two-dimensional” solutions (for the axisymmetric problems): w(x, y, t) = V (r, t), r= x2 + y 2 , where the function V = V (r, t) is determined by the differential equation ∂V 1 ∂ ∂V rf (V ) . = ∂t r ∂r ∂r 5◦ . For other “two-dimensional” solutions, see equation 2.4.3.4 with g(w) = f (w).

B ¤A

3◦ . Solution:

Reference: V. A. Dorodnitsyn, I. V. Knyazeva, and S. R. Svirshchevskii (1983).

. ∂t ∂x ∂x ∂y ∂y This is a two-dimensional unsteady heat and mass transfer equation in an anisotropic case with arbitrary coordinate-dependent principal thermal diffusivities (diffusion coefﬁcients). 4. g(w) 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions

2 w1 = w(C1 x + C2 , C1 y + C3 , C1 t + C4 ),

∂w

=

∂

f (w)

∂w

+

∂

∂w

where C1 , . . . , C4 are arbitrary constants, are also solutions of the equation. 2◦ . Traveling-wave solution in implicit form:

2 2 k1 f (w) + k2 g(w) dw = k1 x + k2 y + λt + C2 , λw + C1 where C1 , C2 , k1 , k2 , and λ are arbitrary constants.

**3◦ . “Two-dimensional” solution: w(x, y, t) = U (z, t), ∂ ∂U ∂U = ϕ(U ) , ∂t ∂z ∂z w(x, y, t) = V (ζ1 , ζ2 ), 5◦ . “Two-dimensional” solution: z = k1 x + k2 y,
**

2 2 ϕ(U ) = k1 f (U ) + k2 g(U ).

where the function U = U (z, t) is determined by a differential equation of the form 1.6.15.1:

4◦ . There are more complicated “two-dimensional” solutions of the form ζ1 = a1 x + a2 y + a3 t, ζ2 = b1 x + b2 y + b3 t.

y x ξ= √ , η= √ , at at where a ≠ 0 is any number and the function W = W (ξ, η) is determined by the differential equation w(x, y, t) = W (ξ, η), ∂W ∂ f (W ) ∂ξ ∂ξ + ∂ ∂W g(W ) ∂η ∂η + a ∂W a ∂W ξ + η = 0. 2 ∂ξ 2 ∂η

**6◦ . For group classiﬁcation of the equation in question, see Dorodnitsyn, Knyazeva, and Svirshchevskii (1983).
**

© 2004 by Chapman & Hall/CRC

C

2.4. EQUATIONS INVOLVING ARBITRARY FUNCTIONS

165

**2.4.4. Other Equations Linear in the Highest Derivatives
**

∂2w ∂2w ∂w + + f (w) = ∂t ∂x2 ∂y 2 The substitution 1. U= F (w) dw, ∂w ∂x

2

+

∂w ∂y

2

.

where F (w) = exp

f (w) dw ,

∂ 2U ∂ 2 U ∂U = + . ∂t ∂x2 ∂y 2 For solutions of this equation, see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). 2. ∂2w ∂2w ∂w = aw + f (t) + + bw2 + g(t)w + h(t), ∂t ∂x2 ∂y 2 “Two-dimensional” generalized separable solution: w(x, y, t) = ϕ(t) + ψ(t)Θ(x, y), where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕt = bϕ2 + g(t)ϕ + h(t), ψt = bϕ − βf (t) + g(t) ψ, ∆Θ + βΘ = 0, ∆≡ β = b/a, (1) (2) a ≠ 0.

leads to the linear heat equation

and the function Θ(x, y) is any solution of the two-dimensional Helmholtz equation ∂2 ∂2 + . (3) 2 ∂x ∂y 2 The ﬁrst equation (1) is independent of ψ and is a Riccati equation for ϕ. In Polyanin and Zaitsev (2003), many exact solutions of equation (1) for various g(t) and h(t) are presented. Solving equation (1) followed by substituting the expression of ϕ = ϕ(t) into (2), we arrive at a linear equation for ψ = ψ(t), which is easy to integrate. In the special case B = 0, a solution of system (1), (2) is given by ϕ(t) = exp G(t) A+ h(t) exp −G(t) dt , F (t) = f (t) dt, G(t) = g(t) dt,

ψ(t) = B exp G(t) − βF (t) ,

where A and B are arbitrary constants. For solutions of the linear stationary equation (3), see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). 3. ∂w ∂t ∂x2 ∂y 2 ◦ 1 . Generalized separable solution: = aw ∂2w + ∂2w –a ∂w ∂x

2

+

∂w ∂y

2

+ f (t).

w(x, y, t) = ϕ(t) + ψ(t)eβx+γy , where β and γ are arbitrary constants and the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕt = f (t), Solving this system yields the solution w(x, y, t) = ϕ(t) + A exp βx + γy + a(β 2 + γ 2 ) where A and B are arbitrary constants.

© 2004 by Chapman & Hall/CRC

ψt = a(β 2 + γ 2 )ϕψ. ϕ(t) dt , ϕ(t) = f (t) dt + B,

166

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

2◦ . There are generalized separable solutions of the following forms: w(x, y, t) = ϕ(t) + ψ(t)(A1 cosh µx + A2 sinh µx) + χ(t)(B1 cos µy + B2 sin µy), w(x, y, t) = ϕ(t) + ψ(t)(A1 cos µx + A2 sin µx) + χ(t)(B1 cosh µy + B2 sinh µy), where A1 , A2 , B1 , B2 , and µ are arbitrary constants, and the functions ϕ(t), ψ(t), and χ(t) are determined by a system of ordinary differential equations (not written out here). 3◦ . There is a generalized separable solution of the form w(x, y, t) = ϕ(t) + ψ(t)F (x) + χ(t)G(y) + η(t)H(x)P (y), where F (x) = A1 cos 2µx + A2 sin 2µx, H(x) = C1 cos µx + C2 sin µx, G(y) = B1 cosh 2µy + B2 sinh 2µy, P (y) = D1 cosh µy + D2 sinh µy,

where the constants A1 , A2 , B1 , B2 , C1 , C2 , D1 , D2 , and µ are related by two constraints, and the functions ϕ(t), ψ(t), χ(t), and η(t) are determined by a system of ordinary differential equations (not written out here). . ∂t ∂x ∂y ∂x ∂x ∂y ∂y This equation describes unsteady anisotropic heat/mass transfer in a steady translational-shear ﬂuid ﬂow. 4. g(w) 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the functions w1,2 = w x + Cb1 eλt , y + C(λ − a1 )eλt , t , where C is an arbitrary constant, and λ = λ1,2 are roots of the quadratic equation λ2 − (a1 + b2 )λ + a1 b2 − a2 b1 = 0, are also solutions of the equation. 2◦ . Solutions: w = w(z), z = a2 x + (λ − a1 )y + Ceλt , (2) (1) ∂w + (a1 x + b1 y + c1 ) ∂w + (a2 x + b2 y + c2 ) ∂w = ∂ f (w) ∂w + ∂ ∂w

where λ = λ1,2 are roots of the quadratic equation (1), and the function w(z) is determined by the ordinary differential equation λz + a2 c1 + (λ − a1 )c2 wz = [ϕ(w)wz ]wz , 3◦ . “Two-dimensional” solutions: w = U (ζ, t), ζ = a2 x + (λ − a1 )y, (3) ϕ(w) = a2 f (w) + (λ − a1 )2 g(w). 2

where λ = λ1,2 are roots of the quadratic equation (1), and the function U (ζ, t) is determined by the differential equation ∂U ∂ ∂U ∂U + [λζ + a2 c1 + (λ − a1 )c2 ] = ϕ(U ) , ∂t ∂ζ ∂ζ ∂ζ ϕ(U ) = a2 f (U ) + (λ − a1 )2 g(U ). 2

Remark 1. A more general equation, with an additional term h(w) on the right-hand side, where h is an arbitrary function, also has solutions of the forms (2) and (3). Remark 2. In the case of an incompressible ﬂuid, the equation coefﬁcients must satisfy the condition a1 + b2 = 0.

© 2004 by Chapman & Hall/CRC

2.4. EQUATIONS INVOLVING ARBITRARY FUNCTIONS

167

5.

∂w = f (w)L [w] + g(t)w + h(t). ∂t Here, L is an arbitrary linear differential operator with respect to the space variables x, y (the operator is independent of t). “Two-dimensional” generalized separable solution: w(x, y, t) = ϕ(t) + ψ(t)Θ(x, y), where the functions ϕ(t) and ψ(t) are given by ϕ(t) = eG(t) A + h(t)e−G(t) dt , ψ(t) = BeG(t) , G(t) = g(t) dt,

A is an arbitrary constant, and the function Θ(x, y) is a solution of the linear stationary equation L [Θ] = 0. Remark 1. In the equation under consideration, the order of the linear operator L and the number of space variables can be any. The coefﬁcients of L can be dependent on the space variables. Remark 2. The above remains valid if f (w) in the equation is substituted by a function f (x, y, t, w). In the special case f (x, y, t, w) = f1 (t) + αw, L [w] = ∆w + βw, where ∆ is the Laplace operator, α and β are some constants, we obtain an equation of the form 2.4.4.2. + kw ln w. ∂t ∂x ∂x ∂y ∂y This is an equation of unsteady heat (mass) transfer or combustion in an anisotropic case with arbitrary coordinate-dependent principal thermal diffusivities (diffusion coefﬁcients) and a logarithmic source. 6. g(x, y) 1◦ . Suppose w(x, y, t) is a solution of the equation in question. Then the function w1 = exp(C1 ekt )w(x, y, t + C2 ), where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . Multiplicative separable solution: w(x, y, t) = exp(C1 ekt )Θ(x, y), where the function U (x, y) satisﬁes the stationary equation ∂U ∂ f (x, y) ∂x ∂x 7. ∂w = ∂ f1 (x, t) ∂w + ∂ + ∂ ∂U g(x, y) ∂y ∂y ∂w + kU ln U = 0. ∂w = ∂ f (x, y) ∂w + ∂ ∂w

+ g1 (x, t) + g2 (y, t) w + h(t)w ln w. ∂t ∂x ∂x ∂y ∂y Exact solution with incomplete separation of variables (the solution is separable in the space variables x and y, but is not separable in time t): w(x, y, t) = ϕ(x, t)ψ(y, t).

f2 (y, t)

Here, the functions ϕ(x, t) and ψ(y, t) are determined from the two one-dimensional nonlinear parabolic differential equations ∂ ∂ϕ ∂ϕ = f1 (x, t) + g1 (x, t)ϕ + h(t)ϕ ln ϕ + C(t)ϕ, ∂t ∂x ∂x ∂ ∂ψ ∂ψ = f2 (y, t) + g2 (y, t)ψ + h(t)ψ ln ψ − C(t)ψ, ∂t ∂y ∂y where C(t) is an arbitrary function.

© 2004 by Chapman & Hall/CRC

168 8.

PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES

**∂2w ∂2w ∂2w ∂w + f2 (x, y) = f1 (x, y) + f3 (x, y) ∂t ∂x2 ∂x∂y ∂y 2 ∂w ∂w + g2 (x, y) + h(x, y) + (t) w + kw ln w. + g1 (x, y) ∂x ∂y Multiplicative separable solution: w(x, y, t) = exp Aekt + ekt e−kt s(t) dt Θ(x, y),
**

D

where A is an arbitrary constant, and the function Θ(x, y) is a solution of the stationary equation f1 (x, y) ∂2Θ ∂2Θ ∂2Θ + f2 (x, y) + f3 (x, y) 2 ∂x2 ∂x∂y ∂y ∂Θ ∂Θ + g2 (x, y) + h(x, y)Θ + kΘ ln Θ = 0. + g1 (x, y) ∂x ∂y f (x, t) ∂w + ∂ g(x, t)w + h(x, t) ∂w

9.

. ∂t ∂x ∂x ∂y ∂y There are “two-dimensional” generalized separable solutions linear and quadratic in y: w(x, y, t) = ϕ(x, t)y + ψ(x, t), w(x, y, t) = ϕ(x, t)y 2 + ψ(x, t)y + χ(x, t).

∂w

=

∂

**2.4.5. Nonlinear Diffusion Boundary Layer Equations
**

∂w ∂w ∂ ∂w ∂w + f (x, t) + g(x, t)y = h(x, t) k(w) . ∂t ∂x ∂y ∂y ∂y This equation arises in nonlinear problems of the unsteady diffusion boundary layer (mass exchange of drops and bubbles with a ﬂow, convective diffusion in ﬂuid ﬁlms), where the coordinates x and y are longitudinal and normal to the interphase surface, respectively. The transformation 1. w = U (ζ, τ , ψ), ζ = yϕ(x, t), τ = τ (x, t), ψ = ψ(x, t), where the functions ϕ(x, t), τ (x, t), and ψ(x, t) are determined by the system of ﬁrst-order partial differential equations ∂ϕ ∂ϕ + f (x, t) = −g(x, t)ϕ, ∂t ∂x ∂τ ∂τ (1) + f (x, t) = h(x, t)ϕ2 , ∂t ∂x ∂ψ ∂ψ + f (x, t) = 0, ∂t ∂x leads to a simpler equation of the form 1.6.15.1: ∂ ∂U ∂U = k(U ) . ∂τ ∂ζ ∂ζ (2)

The cyclic variable ψ does not appear in equation (2); however, it can be involved in the transformed initial and boundary conditions as a parameter. Integrating system (1) is reduced to solving a single ordinary differential equation: x t = f (x, t). In particular, if the functions f , g, and h are only dependent on x, the general solution of system (1) is given by ϕ = Φ1 (z)E(x), τ = Φ2 (z) 1 h(x) 2 E (x) dx + Φ2 (z), f (x) ψ = Φ3 (z),

© 2004 by Chapman & Hall/CRC

2.5. EQUATIONS WITH THREE OR MORE SPACE VARIABLES

169

**where Φ1 (z), Φ2 (z), and Φ3 (z) are arbitrary functions, and z =t−
**

F ¤E

dx , f (x)

E(x) = exp −

g(x) dx . f (x)

Reference: A. D. Polyanin (1982).

2. f (x, y)z n–1

∂w ∂x

+ g(x, y)z n–1

∂w ∂y

+ h(x, y)z n

∂w ∂z

=

∂ ∂z

k(w)

∂w ∂z

.

This equation arises in nonlinear problems of the steady three-dimensional diffusion boundary layer (mass exchange of solid particles, drops, and bubbles with a ﬂow, convective diffusion in ﬂuid ﬁlms), where z is a normal coordinate to the particle surface. To a solid particle there corresponds n = 2 and to drops and bubbles, n = 1. The transformation w = U (ζ, τ , ψ), ζ = zϕ(x, y), τ = τ (x, y), ψ = ψ(x, y),

where the functions ϕ(x, y), τ (x, y), and ψ(x, y) are determined by the system of ﬁrst-order partial differential equations ∂ϕ ∂ϕ + g(x, y) = −h(x, y)ϕ, ∂x ∂y ∂τ ∂τ + g(x, y) = ϕ2 , f (x, y) ∂x ∂y ∂ψ ∂ψ + g(x, y) = 0, f (x, y) ∂x ∂y f (x, y) leads to a simpler equation of the form 1.6.17.16: ∂ ∂U ∂U = ζ 1−n k(U ) . ∂τ ∂ζ ∂ζ (4) (1) (2) (3)

The cyclic variable ψ does not enter into equation (4); however, it can appear in the transformed initial and boundary conditions as a parameter. Suppose an integral of the ordinary differential equation f (x, y)y x = g(x, y) has the form Ξ(x, y) = C. Then the general solution of equation (3) is given by ψ = F (Ξ), where F is an arbitrary function. On passing in (1)–(2) from x, y to the new variables x, Ξ, one arrives at ordinary differential equations with independent variable x where Ξ appears as a parameter.

**2.5. Equations with Three or More Space Variables
**

2.5.1. Equations of Mass Transfer in Quiescent or Moving Media with Chemical Reactions

1. ∂2w ∂2w ∂2w ∂w = + + – f (w). ∂t ∂x2 ∂y 2 ∂z 2

This equation describes unsteady mass or heat transfer with a volume reaction in a quiescent medium. The equation admits translations in any of the variables x, y, z, t. 1◦ . There is a traveling-wave solution, w = w(k1 x + k2 y + k3 z + λt).

© 2004 by Chapman & Hall/CRC

where A. and C are arbitrary constants equation ∂u = (A2 + B 2 + C 2 ) ∂t 2. B. ∂t ∂x ∂y ∂z ∂x2 ∂y 2 ∂z 2 This equation describes unsteady mass transfer with a volume chemical reaction in a steady translational ﬂuid ﬂow. t) = Φ1 (x. y. t). y. ∂t ∂ξ 2 ∂η 2 ∂ζ 2 3. For f (t) = b = const. ζ = z − a3 t leads to a simpler equation of the form 2. The solution speciﬁed in Item 3◦ can be used to obtain other “three-dimensional” solutions by means of the cyclic permutations of the space variables. t). There is a incomplete separable solution of the form w(x. ρ = x2 + y 2 . t) = exp n. t) = ϕ(t)Θ(x. “Three-dimensional” solution: x w = u(ξ. For axisymmetric case.1. + f (t)w ln w + g(t)w.5. x2 = y. y. and Θ(x. η. ζ. η. z. t)Φ2 (y. “Three-dimensional” solution: w = u(ξ. There is a functional separable solution of the form 3 3 w(x. z) is a solution of the stationary equation ∂ 2Θ ∂ 2 Θ ∂ 2 Θ + + + bΘ ln Θ = 0. ∂x2 ∂y 2 ∂z 2 ρ ∂ρ ∂ρ ∂z 2 1 ∂ ∂w 1 ∂ ∂w ∂ 2w ∂ 2 w ∂ 2 w + + = 2 r2 + 2 sin θ . r = x2 + y 2 + z 2 . t) is determined by the ∂ 2 u ∂ 2 u 1 ∂u + + ∂ξ 2 ∂η 2 η ∂η − f (u).1: ∂2U ∂2U ∂2U ∂U = + + − f (U ). a ∂x2 ∂y 2 ∂z 2 – f (w). η. η = y − a2 t. the Laplace operator on the right-hand side of the equation takes the following forms in cylindrical and spherical coordinates. η = (C 2 − 1)x2 − 2Cxy + C 2 z 2 . The transformation w = U (ξ. x3 = z. z. z). 2◦ . 4◦ . A is an arbitrary constant. ∂w + a1 ∂w + a2 ∂w + a3 ∂w = ∂2w + ∂2w + ∂2w © 2004 by Chapman & Hall/CRC . 2 ∂t C ∂ξ ∂ξ∂η ∂η ∂η Remark. t). x1 = x. where ϕ(t) is given by ϕ(t) = exp Aebt + ebt e−bt g(t) dt . and the function u = u(ξ. respectively: ∂w ∂2w ∂ 2w ∂ 2 w ∂ 2 w 1 ∂ + + = ρ + . ∂w =a ∂2w + ∂2w + ∂2w η = (Bx − Ay)2 + (Cy − Bz)2 + (Az − Cx)2 . ξ = x − a1 t. ◦ 3 . the equation also has a multiplicative separable solution of the form w(x. t).170 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 2◦ . ξ = Ax + By + Cz. η. ξ = y + . and the function u = u(ξ. y. η. y. t)Φ3(z. 2 2 2 ∂x ∂y ∂z r ∂r ∂r r sin θ ∂θ ∂θ 3◦ . ∂t ∂x2 ∂y 2 ∂z 2 1◦ .m=1 ϕnm (t)xn xm + n=1 ψn (t)xn + χ(t) . z. t) is determined by the equation 1 ∂2u ∂2u ∂2u ∂u ∂u = 1+ 2 − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) − f (u). C where C is an arbitrary constant (C ≠ 0).

The transformation 4. “Two-dimensional” solution: w = U (ζ. λ is a root of the cubic equation (1). t) is a solution of the equation in question.1. ζ =z− f3 (t) dt leads to a simpler equation of the form 2. is also a solution of the equation. 1◦ . ∂t ∂ξ 2 ∂η 2 ∂ζ 2 ∂y ∂2w ∂2w ∂2w ∂w + + – f (w). 1 2 3 ∂t ∂ζ ∂ζ Remark. A2 . η. λ is a root of the cubic equation (1). where C is an arbitrary constant. t) is determined by the differential equation ∂U ∂2U ∂U + (λζ + A1 d1 + A2 d2 + A3 d3 ) = (A2 + A2 + A2 ) 2 − f (U ). z. and A3 are the corresponding solution of the algebraic system (2). y + A2 Ceλt . Solution: w = w(ξ). ξ = A1 x + A2 y + A3 z + Ceλt . = + (a3 x + b3 y + c3 z + d3 ) ∂z ∂x2 ∂y 2 ∂z 2 This equation describes unsteady mass transfer with a volume chemical reaction in a threedimensional steady translational-shear ﬂuid ﬂow. One of these equations can be omitted. ξ =x− f1 (t) dt. ∂w ∂t + (a1 x + b1 y + c1 z + d1 ) ∂w ∂x + (a2 x + b2 y + c2 z + d2 ) ∂w (2) . z + A3 Ceλt . t). Let λ be a root of the cubic equation a1 − λ a2 a3 b1 b2 − λ b3 = 0. A2 . w = U (ξ. Then the function w1 = w x + A1 Ceλt . and the function w(ξ) is determined by the ordinary differential equation (λξ + A1 d1 + A2 d2 + A3 d3 )wξ = (A2 + A2 + A2 )wξξ − f (w). 1 2 3 3◦ . and A3 be the corresponding solution of the algebraic system (2). © 2004 by Chapman & Hall/CRC 5. since it is a consequence of the other two. EQUATIONS WITH THREE OR MORE SPACE VARIABLES 171 ∂w ∂w ∂w ∂2w ∂2w ∂2w ∂w + + – g(w).1: ∂2U ∂2U ∂2U ∂U = + + − g(U ). Let λ be a root of the cubic equation (1) and let A1 . A2 . where C is an arbitrary constant.2. In the case of an incompressible ﬂuid. and A3 solve the degenerate system of linear algebraic equations (a1 − λ)A1 + a2 A2 + a3 A3 = 0. A 1 . η=y− f2 (t) dt. t . the equation coefﬁcients must satisfy the condition a1 + b2 + c3 = 0. and A 1 . 2◦ . A2 . t). ζ = A1 x + A2 y + A3 z. Suppose w(x. c1 A1 + c2 A2 + (c3 − λ)A3 = 0. b1 A1 + (b2 − λ)A2 + b3 A3 = 0. + f1 (t) + f2 (t) + f3 (t) = ∂t ∂x ∂y ∂z ∂x2 ∂y 2 ∂z 2 This equation describes unsteady mass transfer with a volume chemical reaction in an unsteady translational ﬂuid ﬂow.5. where the function U (ζ. (1) c1 c2 c3 − λ and let the constants A1 . and A3 are the corresponding solution of the algebraic system (2).5. y.

Suppose w(x.172 6. z. c3 − λ (1) and the constants A1 . since it is a consequence of the other two. ϕ(w) = A2 f1 (w) + A2 f2 (w) + A2 f3 (w). A2 . and A3 are the corresponding solution of the algebraic system (2). Then the function w1 = w x + A1 Ceλt . PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w ∂w ∂w ∂w + (a1 x + b1 y + c1 z + d1 ) + (a2 x + b2 y + c2 z + d2 ) + (a3 x + b3 y + c3 z + d3 ) ∂t ∂x ∂y ∂z ∂w ∂ ∂w ∂ ∂w ∂ f1 (w) + f2 (w) + f3 (w) . ∂t ∂ζ ∂ζ ∂ζ ϕ(U ) = A2 f1 (U ) + A2 f2 (U ) + A2 f3 (U ). 1 2 3 Remark 1. © 2004 by Chapman & Hall/CRC . 1 2 3 3◦ . (3) where C is an arbitrary constant. and the function w(ξ) is determined by the ordinary differential equation (λξ + A1 d1 + A2 d2 + A3 d3 )wξ = [ϕ(w)wξ ]ξ . and A 1 . with an additional term g(w) on the right-hand side. One of these equations can be omitted. Let λ be a root of the cubic equation a1 − λ b1 c1 a2 b2 − λ c2 a3 b3 = 0. t) is determined by the differential equation ∂U ∂ ∂U ∂U + (λζ + A1 d1 + A2 d2 + A3 d3 ) = ϕ(U ) . also has solutions of the forms (3) and (4). t . λ is a root of the cubic equation (1). 2◦ . y. ξ = A1 x + A2 y + A3 z + Ceλt . A more general equation. (4) (2) where the function U (ζ. A2 . y + A2 Ceλt . c1 A1 + c2 A2 + (c3 − λ)A3 = 0. “Two-dimensional” solution: w = U (ζ. t). and A3 are the corresponding solution of the algebraic system (2). ζ = A1 x + A2 y + A3 z. Remark 2. the equation coefﬁcients must satisfy the condition a1 + b2 + c3 = 0. 1◦ . is also a solution of the equation. b1 A1 + (b2 − λ)A2 + b3 A3 = 0. and A3 solve the degenerate system of linear algebraic equations (a1 − λ)A1 + a2 A2 + a3 A3 = 0. where g is an arbitrary function. In the case of an incompressible ﬂuid. and A 1 . A2 . = ∂x ∂x ∂y ∂y ∂z ∂z This equation describes unsteady anisotropic mass or heat transfer in a three-dimensional steady translational-shear ﬂuid ﬂow. t) is a solution of the equation in question. Solution: w = w(ξ). where C is an arbitrary constant. and A3 be the corresponding solution of the algebraic system (2). z + A3 Ceλt . λ is a root of the cubic equation (1). A2 . Let λ be a root of the cubic equation (1) and let A1 .

(2) where the function ϕ(t) is determined by the Bernoulli equation (3) . 1. z. spherical. 3◦ . ∂t 1◦ . z) is a solution of the stationary equation ∆Θ + A(n + 1)Θ n+1 = 0. Heat Equations with Power›Law or Exponential Temperature›Dependent Thermal Diffusivity Throughout this subsection. y.5. one arrives at the simpler equation 3. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). z) 1 n+1 .2.5. t) = ϕ(t) Θ(x.6.5. ∇. and Laplace operator in Cartesian coordinates x. and other three-dimensional orthogonal systems of coordinates can be used instead of the Cartesian coordinates). Here. ∂t The substitution U = w n leads to a special case of equation 2. the symbols div. For solutions of this linear equation. y. z. z (cylindrical. y. y. y. y. τ ). gradient operator. t) = F (t)U (x. 2◦ . = ∆(wm ). The general solution of equation (3) is given by ϕ(t) = exp F (t) Aαn exp nF (t) dt + B −1/n 1 1 n+1 . y.5.4.4: α ∂U = αU ∆U + |∇U |2 + nf (t)U + ng(t). w(x. ∂w τ= F n (t) dt. t) = exp f (t) dt Θ(x. Multiplicative separable solution: w(x. where B is an arbitrary constant. ∂U = α div(U n ∇U ). z) satisﬁes the Laplace equation ∆Θ = 0. ∂τ = α div(wn ∇w) + f (t)w + g(t)w1–n .2. y. = α div wn ∇w + f (t)w. ∂t n © 2004 by Chapman & Hall/CRC . and ∆ stand for the divergence operator. z) ϕt − f (t)ϕ + Aαϕn+1 = 0. and Θ(x. z. EQUATIONS WITH THREE OR MORE SPACE VARIABLES 173 2. F (t) = exp f (t) dt . Multiplicative separable solution: 2. z. F (t) = f (t) dt. ∂w ∂w (1) where the function Θ(x. A is an arbitrary constant. ∂t This is a special case of equation 2. Using the transformation w(x.5. y.

1 to 1.1. w(x.3 and equations 1. t) = ln[ϕ(t) + ψ(t)Θ(x. (2) and relation (3) contain arbitrary constants A and B. y. the original equation was studied in Galaktionov and Posashkov (1989) and Ibragimov (1994). ∂w =a w = w(ξ. Additive separable solution: ∂U = α div(eµU ∇U ).7. ∂t 1◦ .1. (1) Here. ξ2 = © 2004 by Chapman & Hall/CRC . with g(t) ≡ const and h(t) ≡ const.4.4. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).4. y. t). t) = f (t) dt + 2◦ . one arrives at the simpler equation 5.2. F (t) = f (t) dt. τ ) + F (t). z) is a solution of the Poisson equation ∆Θ + A = 0.174 4.1 to 1.4.8. µ Note that. y. Using the transformation w(x. see Subsections 1. 1. F (t) = f (t) dt. µ For solutions of the linear stationary equation (2).5 for U = U (x. ∂w = α div(eµw ∇w + f (t). y. z).1.1. and the function Θ = Θ(x. ∂t ∂ξ 2 ξ ∂ξ (2 − m)(2 − n) For solutions of this equation with A = 0 and various f (w). t) is determined by the one-dimensional nonstationary equation 2(4 − m − n) ∂w ∂ 2 w A ∂w = + + f (w). t): ∂U = aU ∆U + bµU 2 + µg(t)U + µh(t). z.1. Note that equations (1). z. and 1. PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 1 ln Θ(x. ∂τ = a div eµw ∇w + beµw + g(t) + h(t)e–µw . y. 1. A= . Equations of Heat and Mass Transfer in Anisotropic Media ∂ ∂w + by m ∂t ∂x2 ∂y ∂y ◦ 1 . ∂t Hence. z)].4.5. Additive separable solution: 1 ln Θ(x. A is an arbitrary constant. µ where the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt + A(α/µ) exp(µϕ) − f (t) = 0.1. ∂t The substitution U = eµw leads to an equation of the form 2. z. 2.2.2. (2) The general solution of equation (1) is given by 1 (3) ϕ(t) = F (t) − ln B + Aα exp µF (t) dt . 4 y 2−m 4 z 2−n x2 + + . 2 a b(2 − m) c(2 − n)2 where the function w(ξ. Solution for m ≠ 2 and n ≠ 2: 1. z). t) = U (x. y. y. y.3. the original equation has solutions of the form 1 w(x.1.5. z. t) = ϕ(t) + 3◦ . z. z. ∂2w + ∂ ∂z cz n ∂w ∂z + f (w).3. ∂w τ= exp[µF (t)] dt. µ where the function Θ = Θ(x. w(x. 1. y. y. z) is any solution of the Laplace equation ∆Θ = 0.1.3. y.

4.7. e−λx e−µy e−νz + + .8. m ≠ 2.1. see Subsections 1. t) is determined by the one-dimensional nonstationary equation ∂w ∂ 2 w 1 ∂w = − + f (w). and l ≠ 2: ∂w w = w(ξ.1. µ ≠ 0. t). 1.3 and equations 1.1. EQUATIONS WITH THREE OR MORE SPACE VARIABLES 175 2◦ . ∂t ∂x ∂x ∂y Solution for n ≠ 2. 1. ∂ ∂w ∂ ∂w ∂w = aeλx + beµy ∂t ∂x ∂x ∂y ∂y Solution for λ ≠ 0. ∂t ∂x ∂ξ 2 ξ ∂ξ 2.3. ∂w = ∂ axn ∂w + ∂ beµy ∂w ∂y + ∂ ∂z ceνz ∂w ∂z + f (w). Solution for m ≠ 2 and n ≠ 2: 4 z 2−n 4 y 2−m + .1. 1. ∂t ∂ξ 2 ξ ∂ξ (2 − n)(2 − m) For solutions of this equation with A = 0 and various f (w).1 to 1. b(2 − m)2 c(2 − n)2 where the function w(x.3 and equations 1.1 to 1. A=2 + + ∂t ∂ξ 2 ξ ∂ξ 2−n 2−m 2−l For solutions of this equation with A = 0 and various f (w). t).1.4.4. 1.4. t). = + f (w). m ≠ 2. see Subsections 1.5.7.1. A= . 1. e−µy e−νz x2−n + + .4.2.2.1 to 1.2.1. 2 ∂t ∂ξ 2 − n ξ ∂ξ © 2004 by Chapman & Hall/CRC . aλ2 bµ2 cν 2 where the function w(ξ.8. ξ2 = 4 + ∂ ∂z A= cz l 4 − mn .4. a(2 − n)2 bµ2 cν 2 where the function w(ξ. and 1.1. and ν ≠ 0: w = w(ξ.3. t). y 2−m e−νz x2−n + + . (2 − m)(2 − n) + f (w). 2 2 a(2 − n) b(2 − m) c(2 − l)2 where the function w(ξ.1. ∂w ∂z y 2−m z 2−l x2−n + + .2. 3.2. ∂t ∂ξ 2 ξ ∂ξ ∂t ∂x ∂x ∂y Solution for n ≠ 2. ξ) is determined by the two-dimensional nonstationary equation w = w(x.3. and 1. ξ2 = 4 + ∂ ∂w ceνz ∂z ∂z + f (w). ∂ ∂w ∂ ∂w = axn + by m ∂t ∂x ∂x ∂y ∂y Solution for n ≠ 2.1.1.1. 4. t) is determined by the one-dimensional nonstationary equation ξ2 = 4 n 1 ∂w ∂w ∂ 2 w = + + f (w).1. a(2 − n)2 b(2 − m)2 cν 2 where the function w(ξ. ξ. and ν ≠ 0: w = w(ξ.4.2.3.1. t) is determined by the one-dimensional nonstationary equation 1 1 1 ∂w ∂ 2 w A ∂w + − 1.1.1. 1. µ ≠ 0. and ν ≠ 0: w = w(ξ.2.4. t). ξ2 = ∂ 2 w ∂ 2 w A ∂w ∂w =a 2 + + + f (w). t) is determined by the one-dimensional nonstationary equation ξ2 = 4 4 − nm ∂w ∂ 2 w A ∂w = + + f (w).1 to 1. 5. ∂w = ∂ axn ∂w + ∂ by m ∂w ∂y + ∂ ∂z ceνz ∂w ∂z + f (w).

with an additional term g(w) on the right-hand side. the equation coefﬁcients must satisfy the condition a1 + b2 + c3 = 0. and A3 are the corresponding solution of the algebraic system (2). Solution: w = w(ξ). ∂t ∂x ∂x ∂y ∂y ∂z ∂z For group classiﬁcation and exact solutions of this equation for some f n (w) and g(w). ζ = A1 x + A2 y + A3 z. is also a solution of the equation. ξ = A1 x + A2 y + A3 z + Ceλt . Let λ be a root of the cubic equation a1 − λ b1 c1 a2 b2 − λ c2 a3 = 0. Then the function w1 = w x + A1 Ceλt . t). see Dorodnitsyn. A more general equation. and the function w(ξ) is determined by the ordinary differential equation (λξ + A1 d1 + A2 d2 + A3 d3 )wξ = [ϕ(w)wξ ]ξ . and A3 be the corresponding solution of the algebraic system (2). y + A2 Ceλt . = ∂x ∂x ∂y ∂y ∂z ∂z This equation describes unsteady anisotropic heat or mass transfer in a three-dimensional steady translational-shear ﬂuid ﬂow. In the case of an incompressible ﬂuid. One of these equations is redundant and can be omitted. A2 . “Two-dimensional” solutions: w = U (ζ. ∂t ∂ζ ∂ζ ∂ζ ϕ(U ) = A2 f1 (U ) + A2 f2 (U ) + A2 f3 (U ). 1 2 3 3◦ . t . 2◦ . t) is determined by the differential equation ∂U ∂ ∂U ∂U + (λζ + A1 d1 + A2 d2 + A3 d3 ) = ϕ(U ) .176 6. 1 2 3 Remark 1. 7. and Svirshchevskii (1983). A2 . z + A3 Ceλt . Let λ be a root of the cubic equation (1) and let A1 . c1 A1 + c2 A2 + (c3 − λ)A3 = 0. where g is an arbitrary function. λ is a root of the cubic equation (1). t) is a solution of the equation in question. Remark 2. 1◦ . y. z. also has solutions of the forms (3) and (4). and A 1 . ϕ(w) = A2 f1 (w) + A2 f2 (w) + A2 f3 (w). b1 A1 + (b2 − λ)A2 + b3 A3 = 0. A2 . ∂w ∂w ∂w ∂w + (a1 x + b1 y + c1 z + d1 ) + (a2 x + b2 y + c2 z + d2 ) + (a3 x + b3 y + c3 z + d3 ) ∂t ∂x ∂y ∂z ∂w ∂ ∂w ∂ ∂w ∂ f1 (w) + f2 (w) + f3 (w) . (4) where the function U (ζ. where C is an arbitrary constant. and A3 are the corresponding solution of the algebraic system (2). and A3 solve the degenerate system of linear algebraic equations (a1 − λ)A1 + a2 A2 + a3 A3 = 0. λ is a root of the cubic equation (1). Suppose w(x. and A 1 . Knyazeva. b3 c3 − λ (1) and let the constants A1 . A2 . ∂w PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂ ∂w = f1 (w) + f2 (w) + f3 (w) + g(w). © 2004 by Chapman & Hall/CRC . (3) (2) where C is an arbitrary constant.

See 2. where F (w) = exp f (w) dw . and other three-dimensional orthogonal systems of coordinates can be used instead of the Cartesian coordinates. EQUATIONS WITH THREE OR MORE SPACE VARIABLES 177 2. The more general equation ∂w = ∂t 3 anm (t)w + bnm (t) n. C. Solutions: w(x. t) = k. = αw∆w – α|∇w|2 – β.4. 3. w(x. spherical. x2 . t) = A − βt + B exp α(κ 2 + µ2 + ν 2 ) At − 1 βt2 eκx+µy+νz . x3 .5.l=1 ϕkl (t)xk xl + k=1 ψk (t)xk + χ(t). = a∆w + f (t)|∇w|2 + g(t)w + h(t). ∇. 2 where A. the symbols div. ∂t There are generalized separable solutions of the form 3 3 ∂w ∂w w(x1 . y. κ.4 with f (t) = −α. gradient operator. Remark.2. 4. 2◦ . = ∆w + f (w)|∇w|2 . g(t) = 0. z. ∂t 1◦ . x3 .m=1 ∂2w + ∂xn ∂xm 3 cn (t) n=1 ∂w ∂xn 2 3 + n=1 sn (t) ∂w + g(t)w + h(t) ∂xn has solutions of the same form. and Laplace operator in Cartesian coordinates x. Remark.4. ∂t There is a generalized separable solution of the form 3 3 ∂w w(x1 . = αw∆w + f (t)|∇w|2 + g(t)w + h(t).l=1 ϕkl (t)xk xl + k=1 ψk (t)xk + χ(t).5. Other Equations with Three Space Variables Throughout this subsection. ∂t The substitution U= F (w) dw.5. ∂w leads to the linear heat equation ∂U = ∆U .m=1 ∂2w + ∂xn ∂xm 3 bn (t) n=1 ∂w ∂xn 2 3 + n=1 cn (t) ∂w + g(t)w + h(t) ∂xn has solutions of the same form. 2. y. t) = k. x2 . and h(t) = −β. ∂t For solutions of this equation. µ. z. z. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). y. t) = Ax + By + Cz − [α(A2 + B 2 + C 2 ) + β]t + D. and ν are arbitrary constants. cylindrical. The more general equation ∂w = ∂t 3 anm (t) n. B. 1. and ∆ stand for the divergence operator. © 2004 by Chapman & Hall/CRC . D.

with f (t) ≡ const. This is a special case of equation 2. the solution of system (1). (1) . τ ). x2 = y. x2 . ∂τ which has a multiplicative separable solution U = ϕ(τ )Θ(x. y. x2 . PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w = [aw + f (t)]∆w + bw2 + g(t)w + h(t). 6. Nβ [w] is an arbitrary homogeneous nonlinear differential operator of degree β with respect to w (i. y. g(t). this equation was studied in Galaktionov and Posashkov (1989) and Ibragimov (1994). where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕt = bϕ2 + g(t)ϕ + h(t).m=1 pnm (x) ∂2w + ∂xn ∂xm 3 qn (x) n=1 ∂w . z). On solving (1) and substituting the resulting ϕ = ϕ(t) into (2). one obtains a linear equation for ψ = ψ(t). and h(t) are arbitrary functions.5. ∂t Here. ψt = bϕ − βf (t) + g(t) ψ. A large number of exact solutions to equation (1) for various g(t) and h(t) can be found in Polyanin and Zaitsev (2003). where A and B are arbitrary constants. x3 = z only and satisﬁes the condition L [const] ≡ 0: 3 L [w] ≡ n. α = const) that depends on the space variables x.4. one arrives at the simpler equation ∂U = Nβ [U ]. x3 .e.. There is a generalized separable solution of the form w(x1 . ψ(t) = B exp G(t) − βF (t) . G(t) = g(t) dt. F (t) = f (t) dt. t) = ϕ(t) + ψ(t)Θ(x1 . a and b are arbitrary parameters (a ≠ 0). β = b/a. L [Θ] + βΘ = 0. z. In the special case b = 0. f (t). ∂xn x = {x1 . x2 . z only (and is independent of t).6 with L [w] ≡ ∆w. g(t) ≡ const. Using the transformation w(x. In the special case L ≡ ∆. = f (t)Nβ [w] + g(t)w. t) = G(t)U (x. (1) (2) (3) and the function Θ(x1 . and h(t) are arbitrary functions. x3 }. x2 . x3 ) is a solution of the linear stationary equation Equation (1) is independent of ψ and represents a Riccati equation for ϕ. ∂w = [aw + f (t)]L [w] + bw2 + g(t)w + h(t). ∂t Here. y. x3 ). y. and h(t) ≡ const. 7. see Tikhonov and Samarskii (1990) and Polyanin (2002) for solutions of the linear stationary equation (3). G(t) = exp g(t) dt . Nβ [αw] = αβ Nβ [w]. © 2004 by Chapman & Hall/CRC ∂w τ= f (t)Gβ−1 (t) dt. g(t). ∂t Here. L [w] is an arbitrary linear differential operator of the second (or any) order that depends on the space variables x1 = x. z.178 5. (2) is given by ϕ(t) = exp G(t) A+ h(t) exp −G(t) dt . f (t). Note that. a and b are arbitrary parameters (a ≠ 0). which is easy to integrate.

∂xk 1. ∂w + (w ⋅ ∇)w = a∆w. The coefﬁcients of N β can be dependent on the space variables. ∂w = ∆w + f (w)|∇w|2 .5. w = {w1 . ∂θ = a∆θ. Nerney. ∂w + (v ⋅ ∇)w = a∆w + a|∇w|2 + f (x.5. Solution: 2a w = − ∇θ. leads to the linear heat equation The substitution U = For solutions of this equation.8 with n = 3. θ where θ is a solution of the linear heat equation 10. b. Equations with n Space Variables Notation: x = (x1 . + (v ⋅ ∇)w = ∆w + f (w)|∇w|2 . where ξ = k1 x + k2 y + k3 z + λτ . ∆w = ∇⋅v = n k=1 n k=1 ∂vk . x2 . . The order of the nonlinear operator Nβ (with respect to the derivatives) and the number of the space variables in the original equation can be any. ∂t This is a special case of equation 2. U = U (ξ). . E. (v ⋅ ∇)w = n vk k=1 ∂w .5. 2. |∇w|2 = ∂x2 k n k=1 ∂w ∂xk 2 . where a. w3 } and wn = wn (x1 . c. Remark 2. . Musielak (1996). ∂xk ∂ 2w . EQUATIONS WITH THREE OR MORE SPACE VARIABLES 179 Remark 1. and Z. Below are two examples of such operators: Nβ [w] = a div(wβ−1 ∇w) + b|∇w|β + cwβ . H ¤G Reference: S. ∂t For solutions of this equation. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).5. ∂t F (w) dw. Nβ [w] = a div(|∇w|β−1 ∇w) + bwµ |∇w|β−µ . x3 ). w2 . . where F (w) = exp ∂U = ∆U . J.5.5. The Hamilton operator ∇ and Laplace operator ∆ can be represented in any orthogonal system of coordinates. t). ∂t Vector Burgers equation. and µ are some constants. E. ∂w 9.5. Schmahl. © 2004 by Chapman & Hall/CRC .9 with n = 3. ∂t f (w) dw . 8. then equation (1) has also a traveling-wave solution. xn ). ∂t This is a special case of equation 2. If Nβ is independent explicitly of the space variables. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).2.

3.j=1 i=1 hi (t)xi + p(t). n 4. Kurdyumov. . P. PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w = f (t)∆w + g(t)w ln w + h(t)w. xn . and h(t) = 0.j=1 ϕij (t)xi xj + i=1 n ψi (t)xi + χ(t).j=1 i=1 hi (t)xi + p(t). where n should be substituted by n − 1. Samarskii. ∂w ∂t = f1 (t)w∆w + f2 (t)|∇w|2 + f3 (t)w + gij (t)xi xj + i. Solutions in the radially symmetric case: n 1 2 − r + Bet . Galaktionov. . There are exact solutions of the following forms: n n w(x1 . = a∇ ⋅ (wm ∇w). A. n 1 2 4 n 1 ln(1 − Ae−t ) w = exp − r 2 (1 − Ae−t )−1 + et B − 4 2A w = exp . The ﬁrst solution is a special case of the second solution as A → 0. Reference: A. . . A > 1. Mikhailov (1995). n 1 ∂w Its exact solutions are given in 1. t) = exp i. S.j=1 ϕij (t)xi xj + i=1 ψi (t)xi + χ(t). this equation describes the ﬂow of a polytropic gas through a homogeneous porous medium (w is the gas density). g(t) = 1. ∂t r ∂r ∂r r= x 2 + · · · + x2 . A < 1. ∂t For m > 1. Let f (t) = 1. . . . . and h(t) = 0. . Solution of the instantaneous source type for a = 1: −n/(nm+2) r2 m 2 K0 − 2/(nm+2) t w= 2(nm + 2) t 0 © 2004 by Chapman & Hall/CRC 1/m if r ≤ K0 t1/(nm+2) . g(t) = −1. .180 2. Example 1. Let f (t) = 1. P. Solution in radially symmetric case: n 1 ln(Aet − 1) w = exp − r 2 (Aet − 1)−1 + e−t B − 4 2A where A and B are arbitrary constants. . t) = i. ∂t There is a functional separable solution of the form n n w(x1 .j=1 ϕij (t)xi xj + i=1 ψi (t)xi + χ(t) . t) = i. A. V. xn . . if r > K0 t1/(nm+2) . 5. where A and B are arbitrary constants. 1◦ .15. r = x2 + · · · + x2 . P ¤I . 2◦ . Example 2. and A.1. . In the radially symmetric case the equation is written as a ∂ ∂w ∂w = n−1 rn−1 wm . xn . There are exact solutions of the following forms: n n w(x1 . ∂w ∂t = f1 (t)∆w + f2 (t)|∇w|2 + f3 (t)w + n n gij (t)xi xj + i.9.

in which m should be substituted by m + 1. t) dx = E0 = const > 0. King (1993). ∂t For m > 1. V. Rn Q 1/m m/(nm+2) . S. S ¤R S ¤R References: J. Rudykh and E. ∂t 1◦ . where A is an arbitrary constant (A > 0). S. Pukhnachov (1995). Ustinov (1985). I. Reference: G. King (1993). x Rn . S ¤R References: S. (2) where the Cj are arbitrary constants.5. other exact solutions are also given there. and the functions ϕk = ϕk (t) are determined by the system of ordinary differential equations ϕ1 dϕn 2m dϕ1 = · · · = ϕn = dt dt m−1 ϕ2 = ϕ2 + C j . © 2004 by Chapman & Hall/CRC . The function ϕn = ϕn (t) is deﬁned in implicit form by (the Cj are assumed to be positive) ϕn n−1 z B m j=1 (z + Cj ) 2 m−1 2 dz = 2mt . EQUATIONS WITH THREE OR MORE SPACE VARIABLES 181 where K0 = π −n/2 Γ(n/m + 1 + 1/m) 2(nm + 2) E0 m Γ(1/m + 1) This is the solution of the initial-value problem with initial function satisfying the condition of constant energy: w(x. B. Titov and V.j=1 aij (t)xi xj + i=1 bi (t)xi + c(t) . E0 = const . Semenov (2000). A. n ∂w w= k=1 −1 n ϕk A− k=1 x2 k ϕ2 k 1 m−1 . R.5: ∂w = m∇ ⋅ (wm−1 ∇w). Kompaneets (1950). J. ◦ −1 1 m−1 2 . V. See also equation 2. A. Solution for m > 1: 6. m−1 where B is an arbitrary constant. I. V. .2. j n n 1−m ϕk k=1 . and the functions ϕk = ϕk (t) are determined by the system of ordinary differential equations (1). where A is an arbitrary constant. . w(x. 2. Pukhnachov (1995). and the remaining ϕj (t) are determined by the positive roots of the quadratic equations (2). There is an exact solution of the form n n w= i. n − 1. Solution for 0 < m < 1: n n w= k=1 ϕk A+ k=1 x2 k ϕ2 k . 3◦ .5. = ∆(wm ). R. It can be rewritten in the form of equation 2.5. Zel’dovich and A. 0) = E0 δ(x). G. V. . (1) System (1) admits n − 1 ﬁrst integrals: j = 1. S ¤R References: Ya.6. Barenblatt (1952). this equation describes the ﬂow of a polytropic gas through a homogeneous porous medium (w is the gas density). .5.5. 1 m−1 3◦ .

+ (v ⋅ ∇)w = a∆w + a|∇w|2 + f (x. . xn . There is a generalized separable solution of the form w(x1 . .182 7. . The substitution 8. where the functions ϕ(t). v is a prescribed vector function dependent on the space coordinates and time (but independent of w). β = b/a. . + (v ⋅ ∇)w = ∆w + f (w)|∇w|2 . g(t). . A large number of exact solutions to equation (1) for various g(t) and h(t) can be found in Polyanin and Zaitsev (2003). one obtains a linear equation for ψ = ψ(t). which is easy to integrate. and h(t) are arbitrary functions. . ψt = bϕ − βf (t) + g(t) ψ. v is a prescribed vector function dependent on the space coordinates and time (but independent of w). t) = ϕ(t) + ψ(t)Θ(x1 . ∆Θ + βΘ = 0. . . . ∂t Here. ∂t 9. © 2004 by Chapman & Hall/CRC . Multiplicative separable solution: w(x. ψ(t) are determined by the system of ordinary differential equations ϕt = bϕ2 + g(t)ϕ + h(t). On solving (1) and substituting the resulting ϕ = ϕ(t) into (2). . PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w = [aw + f (t)]∆w + bw2 + g(t)w + h(t). see Tikhonov and Samarskii (1990) and Polyanin (2002). f (t). see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). . Θ= F (w) dw. t) = exp f (t) dt Θ(x) 1 m+1 . For solutions of this linear equation. ∂w (1) (2) leads a linear convective heat and mass transfer equation for Θ = Θ(x 1 . a and b are arbitrary parameters (a ≠ 0). ∂t 1◦ . . = α∇ ⋅ wm ∇w + f (t)w. . t). ∂t 10. xn . . . t)Θ. ∂w ∂w (1) where the function Θ(x) satisﬁes the Laplace equation ∆Θ = 0. ∂t Here. The substitution Θ = ew leads to the linear equation ∂Θ + (v ⋅ ∇)Θ = a∆Θ + f (x. ∂t Here. . xn ). t): ∂Θ + (v ⋅ ∇)Θ = ∆Θ. where F (w) = exp f (w) dw . xn ) is a solution of the Helmholtz equation (3) Equation (1) is independent of ψ and represents a Riccati equation for ϕ. and the function Θ(x1 . For solutions of the linear stationary equation (3).

The general solution of equation (3) is given by ϕ(t) = exp F (t) 3◦ . f (t) = −β < 0. [g(t)]−n/(nm+2) w= e 2/(nm+2) 2(nm + 2) [g(t)] 0 if r > r∗ (t). T0 ). where B is an arbitrary constant. t) = F (t)U (x. Zaitsev (2002). V = At2/(nm+2) − n 1 nm + 1 A is an arbitrary constant (A > 0). r = x2 + · · · + x2 . where (nm + 2)2 2 t − r 2 . where 11. we have ∂w = ∇ ⋅ (w m ∇w) − βw. The diameter of the support increases with t on the time interval (0. A is an arbitrary constant and the function Θ(x) satisﬁes the stationary equation ∆Θ + A(m + 1)Θ m+1 = 0. Aαm exp mF (t) dt + B −1/m 1 1 (2) (3) . A. = ∇ ⋅ (wm ∇w) – w1–m . F.5. g(t) = 1 + n 1 βm βm The diameter of the support of this solution is monotonically increasing but is bounded now by the constant L = lim |r∗ (t)| = η0 1 + t→∞ ∂U = α∇ ⋅ (U m ∇w). Multiplicative separable solution: w(x. r = x 2 + · · · + x2 . Martinson and K. where 1 − e−βmt 1/(nm+2) 1 − e−βmt . For α = 1. r∗ (t) = η0 1 + . F (t) = f (t) dt. . The solution has a compact support. Martinson (1979). D. ∂τ 1 βm 1/(nm+2) < ∞. where A(nm + 1) T0 = (nm + 2)2 nm+2 nm+1 ∂w nm+2 2(nm+1) . The transformation w(x. L. Kersner (1978). leads to a simpler equation: τ= F m (t) dt. Here. τ ). U ¤T References: L. ∂t Solution in the radially symmetric case: 1/m −βt/m r2 m 2 η0 − if r ≤ r∗ (t). A(nm + 1) t∗ = (nm + 2)3 and decreases on the interval (t∗ . B. Pavlov (1972). Polyanin and V. K. t∗ ). EQUATIONS WITH THREE OR MORE SPACE VARIABLES 183 2◦ . K. w= m 0 if V < 0. The perturbation is localized in a ball of radius L. where the function ϕ(t) is determined by the Bernoulli equation ϕt − f (t)ϕ + Aαϕm+1 = 0. Example.2. t) = ϕ(t) Θ(x) m+1 . The solution vanishes at t = T0 . F (t) = exp f (t) dt . © 2004 by Chapman & Hall/CRC U ¤T References: R. ∂t Solution in the radially symmetric case for 0 < m < 1: 2(nm + 2) −1/m 1/m t V if V ≥ 0.

13. 15. For details about this equation. A. Solution in implicit form: b . see Tikhonov and Samarskii (1990) and Polyanin (2002). a where the function U (x) is determined by the Helmholtz equation f (w) = eat U (x) − ∆U + acU = 0. where the function U (x) is determined by the Poisson equation ∆U + h(x) = 0. ∂w ∂t = ∆f (w) + af (w) + b f (w) + c[af (w) + b]. W ¤V Reference: V. see Kamke (1977) and Polyanin and Zaitsev (2003). λ ψ(t) = exp λ bϕ(t) + f (t) dt . ∂w ∂t = ∇ ⋅ [f (w)∇w] + g(t) f (w) + h(x). see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). where the function U (x) is determined by the Poisson equation ∆U + b = 0. © 2004 by Chapman & Hall/CRC . 14. (2) (1) For details about the Riccati equation (1). and the function Θ = Θ(x) is a solution of the Helmholtz equation a∆Θ + bλΘ = 0. For details about this equation. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). Galaktionov (1994). For solutions of the linear equation (2). W ¤V Reference: V. where the function ϕ(t) is determined by the Riccati equation ϕt = bλϕ2 + λf (t)ϕ + λg(t).184 12. Galaktionov (1994). t) = 1 ln ϕ(t) + ψ(t)Θ(x) . ∂w ∂t = ∇ ⋅ [f (w)∇w] + a f (w) + b. ∂t Functional separable solution: w(x. A. PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w = a∇ ⋅ (eλw ∇w) + beλw + f (t) + g(t)e–λw . Solution in implicit form: f (w) dw = at + U (x). Solution in implicit form: f (w) dw = g(t) dt + U (x).

20. t) = ϕ(t) + exp 2bϕ(t) + f (t) dt Θ(x). the operator satisﬁes the condition L [const] = 0. w) + b where the function ϕ(t) is determined by the Riccati equation ϕt = bϕ2 + f (t)ϕ + g(t). ∂t f (w) Here. L [U ] + ag(x)U = 0. © 2004 by Chapman & Hall/CRC . 18. w) = g(t) dt + U (x). and fw stands for the partial derivative of f with respect to w. ∂w = L [f (w)] + af (w) + b where the function U (x) is determined by the linear equation L [U ] + h(x) = 0.5. ∂t Generalized separable solution: w(x. Solution in implicit form: b f (w) = eat U (x) − . Solution in implicit form: f (x. + g(x)[af (x. where the function U (x) is determined by the linear equation L [U ] + h(x) = 0. w) Here.2. g(t) ∂w = L [f (x. ∂t fw (x. EQUATIONS WITH THREE OR MORE SPACE VARIABLES 185 16. ∂t f (w) Here. Solution in implicit form: ∂w f (w) = g(t) dt + U (x). L is an arbitrary linear differential operator of the second (or any) order with respect to the space variables with coefﬁcients independent of t. w) + b]. a where the function U (x) is determined by the linear equation 17. the operator satisﬁes the condition L [const] = 0. + g(x)[af (w) + b]. and the function Θ = Θ(x) is a solution of the stationary equation a∇ ⋅ (|∇Θ|∇Θ) + bΘ2 = 0. ∂t fw (x. w) = eat U (x) − . w) Here. ∂w ∂w = L [f (x. L is an arbitrary linear differential operator of the second (or any) order with respect to the space variables with coefﬁcients independent of t. = a∇ ⋅ (|∇w|∇w) + bw 2 + f (t)w + g(t). g(t) = L [f (w)] + + h(x). L [U ] + ag(x)U = 0. Solution in implicit form: b f (x. L is an arbitrary linear differential operator of the second (or any) order with respect to the space variables with coefﬁcients independent of t. w)] + af (x. w)] + + h(x). L is an arbitrary linear differential operator of the second (or any) order with respect to the space variables with coefﬁcients independent of t. a where the function U (x) is determined by the linear equation 19. the operator satisﬁes the condition L [const] = 0. the operator satisﬁes the condition L [const] = 0.

h2 + 2g2 h1 + 4g3 h2 − 2Af2 f3 = 0. 1 2 The prime denotes a derivative with respect to t. t) = where C1 . w3 = w(x cos β − y sin β. . gk = gk (t). y. y). y. C1 t + C4 ). 2 2 2 g1 + 4g1 + g2 − Af1 = 0. where the functions fk = fk (t). are also solutions of the equation.6. 2 h3 + h2 + h2 − Af3 = 0. Nonlinear Schrodinger Equations 2. t) is a solution of the Schrodinger equation in question. and β are arbitrary real constants.186 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ¨ 2. © 2004 by Chapman & Hall/CRC X X X . 1. t). “Two-dimensional” solution: w(x. ∂w ∂2w ∂2w w2 = e−i[λ1 x+λ2 y+(λ1 +λ2 )t+C5 ] w(x + 2λ1 t. Two›Dimensional Equations + A|w|2 w = 0. i + + ¨ 1◦ . y. y. Suppose w(x. y.1. . ∂t ∂x2 ∂y 2 Two-dimensional Schr¨ dinger equation with a cubic nonlinearity.1. 2 (x + C2 )2 + (y + C3 )2 − 4AC1 C1 exp i + iC4 . y) is determined by the stationary equation ∂ 2 u ∂ 2u + + Au3 − C1 u = 0. t) = C1 exp i [C2 x + C3 y + (AC1 − C2 − C3 )t + C4 ] . t) = ei(C1 t+C2 ) u(x. . ∂x2 ∂y 2 4◦ . . t 4t 2 2 w(x. h1 + 4g1 h1 + 2g2 h2 − 2Af1 f3 = 0. The plus or minus signs in the expression for w1 are chosen arbitrarily. f2 + 2(g1 + 3g3 )f2 + 2f1 g2 = 0. C4 are arbitrary real constants. where C1 and C2 are arbitrary real constants. x sin β + y cos β. 2◦ . λ1 . This is a special case of equation o 2. f3 + 2(g1 + g3 )f3 + 2(f1 h1 + f2 h2 ) = 0. λ2 . and hk = hk (t) are determined by the autonomous system of ordinary differential equations f1 + 2(3g1 + g3 )f1 + 2f2 g2 = 0. Solution: w(x. . 2 2 2 g3 + g2 + 4g3 − Af2 = 0. C5 . y + 2λ2 t. . where C1 . Then the functions 2 w1 = C1 w( C1 x + C2 . C1 y + C3 . . t) = (f1 x + f2 y + f3 ) exp i(g1 x2 + g2 xy + g3 y 2 + h1 x + h2 y + h3 ) . 3◦ . Solutions: 2 2 2 w(x.6.3 with f (u) = Au2 . t). and the function u = u(x. g2 + 4(g1 + g3 )g2 − 2Af1 f2 = 0.6. .

w(x. t) = √ C1 t + C 2 x + C3 ξ= √ . C1 y + C3 . . 2n (x + C2 )2 + (y + C3 )2 AC1 1−2n C1 exp i +i t + iC4 . η). λ1 . Suppose w(x. see equation 2. and the function U = U (ξ1 .1. ξ2 ) is determined by a differential equation of the form 5. k2 . .1. a. C5 . where k1 . and b are arbitrary constants. and b are arbitrary constants. Then the functions n n 2n w1 = C1 w( C1 x + C2 . The plus or minus signs in the expression for w1 are chosen arbitrarily.3 with f (w) = Aw 2n . t) = C1 exp i [C2 x + C3 y + (A|C1 |2n − C2 − C3 )t + C4 ] . . and λ2 are arbitrary real constants. η) is determined by the differential equation ∂u ∂u ∂2u ∂2u 1 + − iC1 ξ +η + u + A|u|2 u = 0. 3◦ . . . z2 ) is determined by a differential equation of the form 5. where k1 . y. NONLINEAR SCHRODINGER EQUATIONS 187 5◦ .6. y. “Two-dimensional” solution: 1 u(ξ. z2 ) exp i(k1 xt + k2 yt − 3 k1 t3 − 3 k2 t3 + at + b) . β. . C1 t + C 2 y + C4 η= √ .1: ∂2Φ ∂2Φ + + A|Φ|2 Φ − (k1 z1 + k2 z2 + a)Φ = 0. t) = where C1 . For other exact solutions. . x sin β + y cos β. ¨ 1◦ . k2 . Solutions: 2 2 w(x. and the function u = u(ξ. t) is a solution of the Schrodinger equation in question. C4 are arbitrary real constants.1. z 1 = x − k 1 t2 .3 with f (u) = Au2n . ∂ξ 2 ∂η 2 2 ∂ξ ∂η 2. y. . C4 are arbitrary constants. y. where C1 . y.6. are also solutions of the equation. t). ∂t ∂x2 ∂y 2 Two-dimensional Schr¨ dinger equation with a power-law nonlinearity.4. C1 t + C 2 where C1 .4.1: ∂2U ∂2U 2 2 + + A|U |2 U − (k1 + k2 + a)U = 0. and the function Φ = Φ(z1 . .1. C1 t + C4 ). o This is a special case of equation 2. i ∂w ∂ 2 w ∂ 2 w + + + A|w|2n w = 0. ξ2 = y − 2k2 t. y + 2λ2 t.6. Y Y Y w2 = e−i[λ1 x+λ2 y+(λ1 +λ2 )t+C5 ] w(x + 2λ1 t. © 2004 by Chapman & Hall/CRC . t). . A and n are real numbers. z 2 = y − k 2 t2 . ξ1 = x − 2k1 t. y. w3 = w(x cos β − y sin β.¨ 2. ξ2 )ei(k1 x+k2 y+at+b) . t) = U (ξ1 . “Two-dimensional” solution: w(x. 2 2 ∂ξ1 ∂ξ2 6◦ . 2 2 ∂z1 ∂z2 7◦ . . “Two-dimensional” solution: 2 2 2 2 w(x. a. . 2◦ . t) = Φ(z1 . t 4t 1 − 2n 2 2 w(x.

ξ = x − 2A1 t. y. y. η)ei(A1 x+A2 y+Bt+C) .1. and C are arbitrary constants.1. t) = U (ξ. and the function U = U (ξ. z 1 = x − k 1 t2 . y) is determined by a stationary equation of the form 5. and β are arbitrary real constants. “Two-dimensional” solution: w(x.188 3. t) is a solution of the Schrodinger equation in question. and the function Φ = Φ(z1 .7. C3 . where A. z2 ) exp i(k1 xt + k2 yt − 3 k1 t3 − 3 k2 t3 + at + b) . 2 2 ∂z1 ∂z2 7◦ . Then the functions w1 = e−i[λ1 x+λ2 y+(λ1 +λ2 )t+A] w(x + 2λ1 t + C1 .2 with n = 1. where A and B are arbitrary real constants. 4◦ . z 2 = y − k 2 t2 . 2◦ . © 2004 by Chapman & Hall/CRC . x sin β + y cos β. t) = C1 exp i (C2 x + C3 y + λt + C4 ) . A2 . ∂x2 ∂y 2 5◦ . y. There is a “two-dimensional” solution of the form w(x. 3◦ . see Gagnon and Winternitz (1988) and Ibragimov (1995). t) = Φ(z1 . . Exact solutions depending only on the radial variable r = x2 + y 2 and time t are determined by the equation ∂w ∂w 1 ∂ + r + f (|w|)w = 0. are also solutions of the equation. η = y − 2A2 t. i PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES ∂w ∂ 2 w ∂ 2 w + + f (|w|)w = 0. Suppose w(x. y).5. t) = ei(At+B) u(x. 2 2 where A1 . “Two-dimensional” solution: 2 2 2 2 w(x. z2 ).1: ∂ 2u ∂ 2u + + f (|u|)u − Au = 0. y. where k1 . . “Two-dimensional” solution: w(x. t). λ1 . and the function u = u(x. C4 are arbitrary real constants. η) is determined by a differential equation of the form 5. y.4. ¨ 1◦ . λ2 . y + 2λ2 t + C2 . t) = U (z1 . z2 ) is determined by the differential equation ∂2Φ ∂2Φ + + f (|Φ|)Φ − (k1 z1 + k2 z2 + a)Φ = 0.1: ∂2U ∂2U + + f (|U |)U − (A2 + A2 + B)U = 0. a. Traveling-wave solution: w(x. . and b are arbitrary constants. i ∂t r ∂r ∂r which is a special case of equation 1. y. C1 . For group classiﬁcation of the original equation. k2 . t + C3 ). + ∂t ∂x2 ∂y 2 ¨ Two-dimensional nonlinear Schrodinger equation of general form. w2 = w(x cos β − y sin β. z1 = a1 x + b1 y + c1 t. B. 8◦ . C2 . 1 2 ∂ξ 2 ∂η 2 6◦ .4. z2 = a2 x + b2 y + c2 t. where C1 . . 2 2 λ = f (|C1 |) − C2 − C3 .

where C1 . Krasil’shchik (1997). C1 t + C 2 y + C4 η= √ . ζ) is determined by the differential equation ∂u ∂u ∂u ∂2u ∂2u ∂2u 1 + + − iC1 ξ +η +ζ + u + A|u|2 u = 0. Ibragimov (1995). y + 2λ2 t. C1 . Vinogradov and I. k2 . ¨ 1◦ . C1 t + C5 ). ∂t ∂x2 ∂y 2 ∂z 2 ¨ Three-dimensional nonlinear Schrodinger equation of general form. t) = U (ξ1 . 2◦ . are also solutions of the equation. z. There is an exact solution of the form w = (f1 x + f2 y + f3 z + f4 ) exp i(g1 x2 + g2 y 2 + g3 z 2 + g4 xy + g5 xz + g6 yz + h1 x + h2 y + h3 z + h4 ) . . . y. z.2. t) is a solution of the Schrodinger equation in question. ∂t ∂x2 ∂y 2 ∂z 2 Three-dimensional Schr¨ dinger equation with a cubic nonlinearity. y. gk = gk (t). Exact solutions depending only on the radial variable r = x2 + y 2 + z 2 and time t are determined by the equation 1 ∂ ∂w ∂w + 2 r2 + f (|w|)w = 0.2 with n = 2. Three and n›Dimensional Equations 1. η. where k1 . λ2 . 3◦ . . Gagnon and P. ξ3 = z − 2k3 t. ξ2 . C1 t + C 2 z + C5 ζ= √ . 2. . . Winternitz (1988. NONLINEAR SCHRODINGER EQUATIONS 189 2. y. w2 = e−i[λ1 x+λ2 y+λ3 z+(λ1 +λ2 +λ3 )t+C6 ] w(x + 2λ1 t. ξ3 )ei(k1 x+k2 y+k3 z+at+b) . λ1 . C5 are arbitrary constants. 2◦ . 1989). C1 t + C 2 2 2 2 where C1 . and hk = hk (t). i ∂w ∂ 2 w ∂ 2 w ∂ 2 w + + + + A|w|2 w = 0.6. ξ3 ) is determined by the differential equation ∂2U ∂2U ∂2U 2 2 2 + + + A|U |2 U − (k1 + k2 + k3 + a)U = 0. where A. t) = √ C1 t + C 2 x + C3 ξ= √ . C1 y + C3 .5. and λ3 are arbitrary real constants.6.2 with f (u) = Au2 . ξ2 = y − 2k2 t. i ∂t r ∂r ∂r which is a special case of equation 1. S. z + 2λ3 t + C3 . Solution: w(x. is also a solution of the equation. A. It admits translations in any of the independent variables. k3 . and λ3 are arbitrary real constants. N. 2 2 2 ∂ξ1 ∂ξ2 ∂ξ3 4◦ . η. C4 . and the function u = u(ξ. . C6 . ∂w ∂ 2 w ∂ 2 w ∂ 2 w + + + + f (|w|)w = 0. . a. M. . C1 z + C4 . . Suppose w(x. z. . z + 2λ3 t. and b are arbitrary constants. . y + 2λ2 t + C2 . ξ1 = x − 2k1 t. i ¨ 1◦ . Then the function w1 = e−i[λ1 x+λ2 y+λ3 z+(λ1 +λ2 +λ3 )t+A] w(x + 2λ1 t + C1 . This is a special case of equation o 2.7. Then the functions 2 w1 = C1 w( C1 x + C2 .2. t). λ1 . The plus or minus signs in the expression for w1 are chosen arbitrarily. λ2 . where fk = fk (t). “Three-dimensional” solution: 1 u(ξ. w(x. H. ζ). y. .6. ξ2 .¨ 2. t + C4 ). and the function U = U (ξ1 . ∂ξ 2 ∂η 2 ∂ζ 2 2 ∂ξ ∂η ∂ζ b ¤a References: L. Suppose w(x. © 2004 by Chapman & Hall/CRC 2 2 2 ` ` ` ` . t) is a solution of the Schrodinger equation in question.

z). “Three-dimensional” solution: w = V (ξ. t). “Three-dimensional” solution: w = U (ξ. © 2004 by Chapman & Hall/CRC . ξ=y+ x . B. = ∆w + |w|2 w. H. z. 1989). t) is determined by the equation ∂2V ∂2V 1 ∂V ∂V + (A2 + B 2 + C 2 ) + + + f (|V |)V = 0. and C are arbitrary constants and the function V = V (ξ. t) = ei(At+B) u(x. Axisymmetric solutions in cylindrical and spherical coordinates are determined by equations where the Laplace operator has the form ∂w ∂ 2w ∂ 2 w ∂ 2 w 1 ∂ + + = ρ 2 2 2 ∂x ∂y ∂z ρ ∂ρ ∂ρ 1 ∂ ∂w ∂ 2w ∂ 2 w ∂ 2 w + + = 2 r2 ∂x2 ∂y 2 ∂z 2 r ∂r ∂r respectively. y. Reference: A. “Three-dimensional” solution: w(x. The bar over a symbol denotes the complex conjugate. C η = (C 2 − 1)x2 − 2Cxy + C 2 z 2 . η. r= r2 sin θ ∂θ ∂θ where C is an arbitrary constant (C ≠ 0). i ∂t ∂ξ 2 ∂η 2 η ∂η d ¤c d ¤c References: L. Winternitz (1988. Vinogradov and I. η. ∂t This is an n-dimensional Schr¨ dinger equation with a cubic nonlinearity. x + i∇ ⋅ (∆w + |w|2 w)∇w − (∆w + |w|2 w)∇w ¯ ¯ ¯ = 0. ∂z 2 + ρ= x2 + y 2 . + ∂2w . y. η. N. 4◦ . Ibragimov (1995). M. and the function U = U (ξ. 5◦ . Krasil’shchik (1997). where A. z) is determined by the stationary equation ∆u + f (|u|)u − Au = 0. y. where A and B are arbitrary real constants. Gagnon and P. S. 1 ∂ ∂w sin θ . t) is determined by the differential equation i 1 ∂U + 1+ 2 ∂t C ∂2U ∂U ∂2U ∂ 2U − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) + f (|U |)U = 0. and the function u = u(x.190 PARABOLIC EQUATIONS WITH TWO OR MORE SPACE VARIABLES 3◦ . 2 ∂ξ ∂ξ∂η ∂η ∂η 6◦ . x2 + y 2 + z 2 . η. o Conservation laws: 3. ξ = Ax + By + Cz. i |w|2 |∇w| − t 2 ∂w + i∇ ⋅ w∇w − w∇w ¯ ¯ 4 1 2 |w| t x = 0. t). η= (Bx − Ay)2 + (Cy − Bz)2 + (Az − Cx)2 .

C2 . Equations with Power-Law Nonlinearities The general properties of equations of this type are outlined in 3. w(x. t) = k(t + C1 )2 − k(x + C2 )2 where C1 and C2 are arbitrary constants. 1 k = 4 a(1 − n)2 . x + C2 .4. Then the functions w2 = w(x cosh λ + t sinh λ. C3 . ξ = (t + C1 )2 − (x + C2 )2 .1. Suppose w(x. 1. are also solutions of the equation (the plus or minus signs are chosen arbitrarily).1. t) is a solution of the equation in question. 2 z = x + C 1 t + C2 . t) = F (z). t + C1 ξ= where the function u(ξ) is determined by the ordinary differential equation (1 − ξ 2 )uξξ + The transformation 2(1 + n) 2(1 + n) ξuξ − u + aun = 0. 2◦ . t) = b x + C1 t + C2 2 1−n 2 2(1 + n)(C1 − 1) b= a(1 − n)2 1 1−n 1 n−1 2 2 3. ∂t2 ∂x2 This is a special case of equation 3. traveling-wave solutions and some other solutions are also presented there.1. w(x. . . t) = (t + C1 ) 1−n u(ξ). 3◦ . x sinh λ + t cosh λ). and λ are arbitrary constants. Solutions: w(x. 4◦ .1. Equations of the Form ∂ w = ∂ w + aw + bw n + cw 2n–1 2 ∂t ∂x2 ∂2w = ∂2w 2 w1 = C1 w n−1 n−1 C1 x + C 2 . 1◦ . 1−n (1 − n)2 2 u = (cosh θ) n−1 U (θ). t) = G(ξ). Self-similar solution: w(x.1 with f (w) = aw n . + awn .1.1. where C1 .Chapter 3 Hyperbolic Equations with One Space Variable 3.4. The solutions of Item 2◦ are special cases of solutions of the following forms: w(x. . C 1 t + C 3 . © 2004 by Chapman & Hall/CRC ξ = tanh θ .

z= 1 2 √ a (1 − n)(x sinh C1 t cosh C1 ) + C2 t cosh C1 ) + C2 t cosh C1 ) + C2 .4. ∂2w = ∂2w 1 1−n + awn + bw2n–1 . t) = λ + C1 exp z 1 1−n . Traveling-wave solutions for a > 0: 2b sinh2 z w(x. z= √ a (1 − n)(x sinh C2 where λ = λ1. if b(n + 1) < 0. 4. Solutions: a(1 − n)2 (x sinh C1 w(x. .1 with f (w) = aw n + bw2n−1 . t cosh C2 ). 2◦ . © 2004 by Chapman & Hall/CRC ¡ . z= 1 2 √ a (1 − n)(x sinh C1 ¡ . + aw + bwn .4. if b(n + 1) > 0. t) = ¡ b(n + 1) t cosh C1 + C2 ) − 2an 2 1 1−n b 1 a(1 − n)2 (t + C1 )2 − (x + C2 )2 − 4 an where C1 and C2 are arbitrary constants.1. 2. t) = a(n + 1) 2b cosh2 z w(x. Traveling-wave solutions: = ¡ ∂2w ∂2w w(x. ∂t2 ∂x2 This is a special case of equation 3.2 are roots of the quadratic equation aλ2 − a(n + 1)λ + b = 0. 1◦ . See also equation 3. Traveling-wave solutions for a < 0 and b(n + 1) > 0: 2b cos2 z w(x. t) = − a(n + 1) 1 1−n −1/2 ∂2w = ∂2w 1 1−n where C1 and C2 are arbitrary constants. ∂t2 ∂x2 This is a special case of equation 3. and C1 and C2 are arbitrary constants. z= 1 2 |a| (1 − n)(x sinh C1 ¡ . + aw – a(n + 1)wn + bw2n–1 . 2◦ .192 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE brings this equation to the autonomous form Uθθ − 4 U + aU n = 0. ∂t2 ∂x2 1◦ .5. 1 1−n .1. t) = − a(n + 1) 3.1. t) = 2(n + 1) w(x.1. (1 − n)2 Integrating yields the general solution in implicit form 2a n+1 4 U2 − U + C3 (n − 1)2 n+1 where C3 and C4 are arbitrary constants. in which b should be renamed −a(n + 1) and c renamed b.1 with f (w) = aw + bw n . . ¡ dU = C4 θ.

A=− 1 z= ¢ − 2 2 3.4. For b = 1. 2 1−k . w) = az n wk . B= − . © 2004 by Chapman & Hall/CRC ¢ ¢ ¢ A=− 1/2 |a| (1 − n)(x sinh C1 ∂u ∂x 2 = a(1 − n)u2 + b(1 − n)u + c(1 − n).3. B= − 2 . 3◦ .1. 2 1−k . = ∂2w ∂2w ¢ ∂2w ∂2w = + aeβt wk . 2 (n + 1)2 a(n + 1) a an where C1 and C2 are arbitrary constants. t) = (A + B sinh z) 1−n .1.1.1.4. a(n + 1) an a (n + 1)2 where C1 and C2 are arbitrary constants (the expressions in square brackets must be nonnegative).1. ∂t2 ∂x2 This is a special case of equation 3.1. t) = C exp 4(k + 1) β √ a β(k − 1) (t x) − (1 − k)eβt/2 w(x.13 and 3. ∂2w ∂2w = + aw + bwn + cw2n–1 . t) = (A + B cosh z) 1−n . ¢ w(x. + a(x2 – t2 )(xt)n wk . t) = C exp 4(k + 1) β where C is an arbitrary constant. The substitution u = w 1−n leads to an equation with a quadratic nonlinearity: u ∂2u ∂2u − ∂t2 ∂x2 + n 1−n ∂u ∂t 2 1/2 ¢ w(x.4.4.4. z = a (1 − n)(x sinh C1 t cosh C1 ) + C2 . ∂t2 ∂x2 This is a special case of equation 3. see also equations 3. 1/2 ¢ t cosh C1 ) + C2 . w) 2 ∂t ∂x2 1. w) = cz n wk . Traveling-wave solutions for a > 0: 1 √ w(x. ¢ A=− b . t. = ∂2w =a ∂2w ∂2w ∂2w + c(x + bt)n wk .7 with f (w) = aw k . 1◦ .4.4.2 with f (w) = aw k . Equations of the Form ∂ w = a ∂ w + f (x.5 with f (z. ∂t2 ∂x2 This is a special case of equation 3.2.1 with f (w) = aw + bw n + cw2n−1 .1. + a(x2 – t2 )wk . 4. Functional separable solutions: √ a β(k − 1) (t x) + (1 − k)eβt/2 w(x.1. a(n + 1) 1 B= b2 c b . ∂t2 ∂x2 This is a special case of equation 3.14 with f (ξ) = cξ n and g(w) = wk . 2 (n + 1)2 a an √ z = a (1 − n)(x sinh C1 t cosh C1 ) + C2 .1. c b2 − . 2. Traveling-wave solutions for a < 0: c b2 b . t) = (A + B cos z) 1−n .4 with f (z. 2◦ . 3. EQUATIONS WITH POWER-LAW NONLINEARITIES 193 5. . ∂t2 ∂x2 This is a special case of equation 3.

Zhurov. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w = + aw + beβt wk . ∂t2 = ∂x2 + k2 w(x. t) = £ £ £ ∂2w ∂2w β2 k ≠ 0. Zhurov. k where C is an arbitrary constant. . Polyanin.194 5. Polyanin. I. k 2a βt e +b . A. t) = w(x. V. Vyazmin. © 2004 by Chapman & Hall/CRC £ . ¥ ¦¤ Reference: A. w + a2 e2βt + abkeβt – b2 w–3 . A. A. Polyanin. A. ∂t2 ∂x2 Functional separable solutions: 6. t) = C exp 2 £ £ £ £ k−1 4β(k + 1) [β 2 − (k − 1)2 a][β 2 − (k + 3)2 a] x + (k − 1) b eβt/2 β 2 − (k − 1)2 a 2 1−k + [β + (k − 1)(k + 3)a] t w(x. t) = 1 β(k − 1) 1 (t x) + √ a + 2 (1 − k)beβt C exp 2(k + 1) β b 1 β(k − 1) 1 (t x) − √ a + 2 (1 − k)beβt C exp 2(k + 1) β b 2 1−k ∂2w = ∂2w . A. +4 Functional separable solutions: 7. D. k 2a βt e +b . w(x. and D. ¥ ¦¤ Reference: A. V. D. V. A. ∂t2 ∂x2 Functional separable solutions: w(x. Vyazmin. Zhurov. I. Kazenin (1998). A. Kazenin (1998). Reference: A. Kazenin (1998). and D. A. k−1 4β(k + 1) [β 2 − (k − 1)2 a][β 2 − (k + 3)2 a] x − (k − 1) b eβt/2 β 2 − (k − 1)2 a 2 1−k + [β + (k − 1)(k + 3)a] t where C is an arbitrary constant. Vyazmin. ¥ ¦¤ . 2 1−k where C is an arbitrary constant. and D. A. t) = w(x. + eβt (a + beβt )wk . D. βkx C exp βt + √ k2 + 4 βkx C exp βt − √ k2 + 4 βkx C exp βt + √ k2 + 4 βkx C exp βt − √ k2 + 4 + + − − √ k2 + 4 β √ k2 + 4 β √ k2 + 4 β √ k2 + 4 β 2a βt e +b . k 2a βt e +b . t) = C exp 2 . t) = w(x. I. t) = w(x.

Zhurov. A. ∂t2 ∂x2 k –4 |k| > 2.3. I. © 2004 by Chapman & Hall/CRC . k where C is an arbitrary constant. ∂t2 ∂x2 This is a special case of equation 3. EQUATIONS WITH POWER-LAW NONLINEARITIES 195 8. 9. Functional separable solutions: w(x. k 2a βt e +b . Kazenin (1998). k 2a βt e +b . © ¦¨ Reference: A. t) = w(x. t) = w(x. .6 with f (w) = −aw k . A.1. Functional separable solutions: § § w(x. V. ∂2w ∂2w = – aeβx wk .1. 10. ∂2w β2 ∂2w = – 2 w + a2 e2βt + abkeβt + b2 w–3 . t) = § § § § βkx C exp βt + √ k2 − 4 βkx C exp βt − √ k2 − 4 βkx C exp βt + √ k2 − 4 βkx C exp βt − √ k2 − 4 + + − − √ k2 − 4 β √ k2 − 4 β √ k2 − 4 β √ k2 − 4 β 2a βt e +b . ∂2w ∂t2 = ∂2w ∂x2 – aw – beβx wk . Vyazmin. Functional separable solutions: w(x. t) = w(x. t) = w(x.4. . k 2a βt e +b . and D. 2 1−k where C is an arbitrary constant. D. t) = β(k − 1) (x C exp 4(k + 1) § β(k − 1) (x C exp 4(k + 1) √ a (1 − k)eβx/2 t) + β √ a (1 − k)eβx/2 t) − β § 2 1−k . t) = C exp 2 . k−1 4β(k + 1) [β 2 − (k − 1)2 a][β 2 − (k + 3)2 a] t − (k − 1) b eβx/2 β 2 − (k − 1)2 a 2 1−k + [β + (k − 1)(k + 3)a] x where C is an arbitrary constant. t) = C exp 2 k−1 4β(k + 1) [β 2 − (k − 1)2 a][β 2 − (k + 3)2 a] t + (k − 1) b eβx/2 β 2 − (k − 1)2 a 2 1−k + [β + (k − 1)(k + 3)a] x w(x. A. Polyanin.

t) is a solution of the equation in question. 2003) give more than 20 exact solutions to equation (2) for speciﬁc values of the parameters n and m. b(1 − m)2 a(t + C)2 − x2 2a(2 + n − nm) 1 1−m . C1 t + C2 . 12. this equation describes nonlinear waves with axial and central symmetry.3. 4◦ .8 with f (w) = cw k . r Aa (2) The books by Polyanin and Zaitsev (1995. t) = where C is an arbitrary constant. ∂w 2 ∂t ∂x2 ∂x ∂2w = ∂2w 1. 2◦ .196 11.1. (1) where the sign of A must coincide with that of the expression in square brackets. t) = β(k − 1) (x C exp 2(k + 1) t) + β b 1 √ 1 √ a+ a+ 1 2 (1 − k)be − k)be βx 2 1−k . Solution (1) is a special case of the wider family of exact solutions w = w(r). ∂t2 ∂x2 This is a special case of equation 3. t. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). Then the functions 2 w1 = C1 w k−1 k−1 C1 x. t) = w(x. . 1◦ . ∂2w = a ∂ ∂w For n = 1 and n = 2. 2 2 3. + ceax+bt wk . There is a self-similar solution of the form w = t 1−m f (ξ). Suppose w(x. and the function w = w(r) is determined by the ordinary differential equation wrr + b m n+1 wr = w .4. Equations of the Form ∂ w = a ∂ w + f x. β(k − 1) (x C exp 2(k + 1) t) − β b 1 2 (1 βx 2 1−k where C is an arbitrary constant. a > 0. Functional separable solution: w(x. + bwm .1. © 2004 by Chapman & Hall/CRC 2 . w. 3◦ . ∂x ∂x This equation can be rewritten in the equivalent form ∂t2 xn xn ∂ 2 w n ∂w ∂2w =a + ∂t2 ∂x2 x ∂x + bwm . r2 = A a(t + C)2 − x2 . HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w = – eβx (a + beβx )wk . where C1 and C2 are arbitrary constants. ∂t2 ∂x2 Functional separable solutions: w(x. where ξ = x/t. respectively.

g(t) = k. ∂2w =a ∂2w +c ∂w 2 5. ∂t2 ∂x2 ∂x This is a special case of equation 3. 2 2 ∂t ∂x ∂x This is a special case of equation 3. Traveling-wave solution: w(x. + cw + xn .4.4. Then the functions m m w1 = C1 w C1 x + C2 .5 with f (x) = sxn . ∂t2 ∂x2 ∂x This is a special case of equation 3. (1) where C and λ are arbitrary constants. Suppose w(x. where C1 .2. see 3. © 2004 by Chapman & Hall/CRC 6. 3.4.4 with f (t) = s. and λ are arbitrary constants.2. . 2.2. t) = − U (t) = s 1 c 2 ( x + C 1 t + C 2 )2 + (a − C1 ) − + U (t). t) is a solution of this equation. C2 . where C1 . t + C2 ) + C3 cos(kt) + C4 sin(kt) ∂2w =a ∂2w +b ∂w 2 if c = −k 2 < 0. Generalized separable solution quadratic in x: w(x. C. . t + C2 ) + C3 cosh(kt) + C4 sinh(kt) if c = k 2 > 0. and C3 are arbitrary constants.3. 1◦ . w2 = w( x + C1 . 4. . 3◦ . ∂2w ∂w 2 ∂2w =a +b + cw + tn . EQUATIONS WITH POWER-LAW NONLINEARITIES 197 ∂2w ∂w ∂2w =a + bwm . t) = bm(x + λt + C) (m + 1)(a − λ2 ) −1/m . 3◦ .4. 2◦ . ∂t2 ∂x2 ∂x 1◦ . . C 1 t + C 3 . Solution (1) is a special case of the wider class of traveling-wave solutions x + λt + C dw = . ∂2w =a ∂2w +b ∂w 2 + bcw2 + kw + . Suppose w(x. Then the functions w1 = w( x + C1 . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). . m+1 A + bw (m + 1)(λ2 − a) where A.2. + cw + . ∂t2 ∂x2 ∂x This is a special case of equation 3. are also solutions of the equation. 2◦ .4.10 with f (t) = c.3 with f (w) = bw m . t) is a solution of the equation in question. C4 are arbitrary constants. C3 cos(kt) + C4 sin(kt) if c = −k 2 < 0.1. There is a self-similar solution of the form w = t−1/m f (x/t).4 with f (t) = stn .2. and h(t) = s. For other solutions. 4b 2b c C3 cosh(kt) + C4 sinh(kt) if c = k 2 > 0.

V. ∂w ∂t2 ∂x2 ∂x ∂2w =a ∂2w + btn ∂w 2 + ctk x ∂w 1. D. For more complicated solutions.8 with f (x) = bxn . ∂2w ∂w ∂2w =a + bxn 2 2 ∂t ∂x ∂x 2 + cxm + tk . A. ∂t2 ∂x2 This equation describes the propagation of nonlinear waves in an inhomogeneous medium. 2◦ .1. If 2Abc + k − ab = −σ 2 < 0.2. Zaitsev and A. + bwm . w.10. t.2. g(t) = stm .4.198 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE Let A be a root of the quadratic equation bcA2 + kA + s = 0. 2 2 3. Equations of the Form ∂ w = f (x) ∂ w + g x.4. t) = s a(2 − n)2 (t + C)2 − 4(x + β)2−n 1 1−m ∂2w = a(x + β)n ∂2w where C is an arbitrary constant. 2−n r k 4b(2 − n)2 4(1−n) m ξ n w . kn2 The substitution ξ = r 2−n leads to the Emden–Fowler equation wξξ = © 2004 by Chapman & Hall/CRC 7. g(x) = cxm . s= b(1 − m)2 . Then there are generalized separable solutions √ w(x. ∂2w ∂w ∂2w =a + ctn 2 2 ∂t ∂x ∂x 2 + bctn w2 + tm w + ptk . r2 = k (x + β)2−n 1 (t + C)2 − . Suppose that 2Abc + k − ab = σ 2 > 0. Polyanin (1996). Galaktionov (1995. This is a special case of equation 3. 8.4. Functional separable solution for n ≠ 2: w(x. and h(t) = stk . 2a(2 − n)(nm − 3n + 4) (1) (2) . 2◦ . see 3. t) = A + C1 cos(σt) + C2 sin(σt) exp x −b . F. and h(t) = ptk . 4 a(2 − n)2 where k and the expression in square brackets must have like signs. and the function w(r) is determined by the ordinary differential equation wrr + n 4b m 2(1 − n) 1 wr = w . 1◦ . ∂x ∂x There is a generalized separable solution quadratic in x: ∂t2 ∂x2 w = ϕ(t)x2 + ψ(t)x + χ(t). . .10 with f (t) = ctn .2. t) = A + C1 exp(σt) + C2 exp(−σt) exp x −b . 9. the case a = c was considered). 1◦ .4. where C1 and C2 are arbitrary constants. This is a special case of equation 3. ¦ References: V. a > 0. there are generalized separable solutions √ w(x. Functional separable solution (generalizes the solution of Item 1 ◦ ): w = w(r).

Polyanin (1996). wrr + 2−n r k The substitution ξ = r n−2 leads to the Emden–Fowler equation 4b(2 − n)2 − 4 m ξ nw . D. Zaitsev and A. a > 0. EQUATIONS WITH POWER-LAW NONLINEARITIES 199 The books by Polyanin and Zaitsev (1995. . and the function w = w(y) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wyy − aBwy + bwm = 0. For n = 1. + bwm . the general solution of equation (1) is written in explicit form as C1 + where C1 and C2 are arbitrary constants. and the function w(r) is determined by the ordinary differential equation 4b m 2 1 wr = w . √ aB 2 − A2 wy brings (3) to the Abel equation For A ≠ B a. Then the functions w1 = C1 w C12−n x. − 2 . 2003) give more than 20 exact solutions to equation (2) for speciﬁc values of the parameters n and m. 4◦ . 1 can be found in Polyanin and Zaitsev (2003). © 2004 by Chapman & Hall/CRC n 1 1−m ∂2w = where C1 and C2 are arbitrary constants. 2. Suppose w(x. s= .1. where C is an arbitrary constant. ¦ Reference: V. 3◦ . are also solutions of the equation. (3) √ The solution of equation (3) with A = B a is given by b(1 − m) y+C w(y) = aB 1 1−m . Solution for n = 2: w = w(y). Functional separable solution (generalizes the solution of Item 2 ◦ ): x2−n 1 . where A and B are arbitrary constants. 8b w m+1 k(m + 1) −1/2 dw = r + C2 . U Uw − U = a2 B 2 1 whose general solutions for m = −2. ∂t2 axn 1◦ . 0. where ξ = (x + β)t n−2 . the substitution U (w) = aB b(A2 − aB 2 ) m w . Special case. −1. (1) wξξ = kn2 The books by Polyanin and Zaitsev (1995. y = At + B ln |x + β|. Functional separable solution for n ≠ 2: w(x. F. b(1 − m)2 . ∂x ∂x This equation describes the propagation of nonlinear waves in an inhomogeneous medium. r2 = k (t + C)2 − 4 a(2 − n)2 where k and the expression in square brackets must have like signs. 2003) give more than 20 exact solutions to equation (1) for speciﬁc values of the parameters n and m. t) is a solution of the equation in question. C1 2◦ . w = w(r).3. t) = s a(2 − n)2 (t + C)2 − 4x2−n where C is an arbitrary constant. 3◦ . There is a self-similar solution of the form w = t 1−m f (ξ). 2a(2 − n)(4 − n − nm) 2 2 ∂ ∂w m−1 m−1 2 t + C2 .

√ A2 − aB 2 wz brings (2) to the Abel equation For A ≠ B a. 2◦ . C1 t + C2 ). where C1 and C2 are arbitrary constants. where A. # ¦" Reference: V. F. 3. ∂2w ∂w ∂2w = axn + bxn–1 wm . Zaitsev and A. − 2 .200 4◦ . 1 can be found in Polyanin and Zaitsev (1995. . and C are arbitrary constants. the substitution U (w) = aB b(A2 − aB 2 ) m w . There is a self-similar solution of the form w = U (ζ). where A and B are arbitrary constants. 2 w = w(ξ). z = At + B ln |x|.4. Functional separable solution for n ≠ 2: w = w(z). z = a(2 − n)2 (t + C)2 − 4x2−n 1/2 ! where C is an arbitrary constant. where C is an arbitrary constant. and the function w = w(ξ) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wξξ + B(bwm − a)wξ = 0. Solution for n = 2: ζ = xt n−2 . Suppose w(x. B. There is a self-similar solution of the form w = t 1−m f (ξ). 4◦ .3. Solution for n = 2: HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE w = w(z). √ The solution of equation (2) with A = B a is given by b(m − 1) z+C w(z) = aB ! (2) 1 1−m . z = At + B ln |x| + C. and the function w = w(z) is determined by the ordinary differential equation 2 a(1 − n) + bwm wz = 0. Then the functions 2−n 2 w1 = w(C1 x.5 with f (w) = bw m . 3◦ . −1. 0. and the function w = w(z) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wzz + aBwz + bwm = 0. t) is a solution of the equation in question. (1) wzz + a(2 − n)z The substitution u(w) = zwz leads equation (1) to a ﬁrst-order separable ordinary differential equation. D. are also solutions of the equation. U Uw − U = a2 B 2 1 whose exact solutions for m = −2. m+1 − a(m + 1)w + C bw (m + 1)(aB 2 − A2 ) 1 © 2004 by Chapman & Hall/CRC ! 2 2 . where ξ = xt n−2 . 1◦ . 2 2 ∂t ∂x ∂x This is a special case of equation 3. 2b a(2 − n) wm+1 + C1 anw − m+1 where C1 and C2 are arbitrary constants. Integrating yields Bξ dw =− . 5◦ . 2003). a > 0. Integrating yields a solution in implicit form: 1 dw = ln z + C2 . Polyanin (1996).

are also solutions of the equation. ∂t2 ∂x2 This is an equation of the propagation of nonlinear waves in an inhomogeneous medium. This is a special case of equation 3. aλ2 4 $ 1 1−m . 1◦ .1. where C1 and C2 are arbitrary constants. (1) wzz + aλ z z akλ2 © 2004 by Chapman & Hall/CRC $ z = 4ke−λx − akλ2 (t + C)2 1/2 .4. . Functional separable solution for λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). 2 ln |t|. 7. t) = |t| 1−m F (z). 4◦ . 2 2 ∂t ∂x ∂x Functional separable solution: w = w(z). 3◦ . r2 = 4k e−λx 1 − (t + C)2 . a > 0. 6. r2 = 4k e−λx 1 − (t + C2 )2 . 4.3.3.9 with b = aλ and f (w) = cw m .3. ∂t2 ∂x ∂x This is a special case of equation 3. Suppose w(x. + cwm .9 with b = 0 and f (w) = cw m .4. and k are arbitrary constants. λ where C1 and C2 are arbitrary constants. There is an exact solution of the form w(x. 2 2 ∂t ∂x ∂x This is a special case of equation 3. 5. t) is a solution of the equation in question. Integrating yields the general solution in implicit form $ C1 − 2c wm+1 k(m + 1) −1/2 dw = C2 r. a > 0. ∂2w = ∂ aeλx ∂w ∂2w = aeλx ∂2w + cwm . Functional separable solution for m ≠ 1 and λ ≠ 0: c(m − 1)2 (r + C1 )2 w= − 2k(1 + m) where C1 . 2◦ . C2 . k = 1.6 with f (w) = bw m and g(w) = cwk .4. aλ2 4 where the function w(r) is determined by the autonomous ordinary differential equation wrr + ck −1 wm = 0. λ where C is an arbitrary constant.3. and the function w = w(z) is determined by the ordinary differential equation c 2(aλ − b) 1 w + wm = 0. Then the functions $ w1 = C1 w x + m−1 1−m ln C1 . EQUATIONS WITH POWER-LAW NONLINEARITIES 201 ∂2w ∂w ∂2w = axn + bxn–1 wm + cwk . ∂2w ∂w ∂2w = aeλx + beλx + cwm . C1 2 t + C2 . 2 z =x+ a > 0.

.5. For b = aλ. . 2◦ . The ﬁrst solution is degenerate and the second one is a generalized separable solution.1. Solution: 2 w = U (z) + 4aC1 t2 + 4aC1 C2 t. where C1 and C2 are arbitrary constants and the function U (z) is determined by the autonomous ordinary differential equation 2 2 (U − aC2 )Uzz − 2C1 Uz = 8C1 . a > 0. w) 2 ∂t ∂x2 1. . . and the function f = f (t) is determined by the autonomous ordinary differential equation ftt = 2af 2 .202 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE This equation has the exact solution w(z) = 2kλ aλ(m − 3) + 2b(1 − m) c(1 − m)2 z 2 1 m−1 .5 with n = 1. Zaitsev (2002). the general solution of equation (1) is given in implicit form by % C1 − 2c wm+1 akλ2 (m + 1) 2b−aλ aλ −1/2 dw = z + C2 . the substitution ξ = z 2 wξξ + brings (1) to the generalized Emden–Fowler equation (2) 4(aλ−b) ac ξ 2b−aλ wm = 0.1. where C1 and C2 are arbitrary constants. Equations of the Form ∂ w = aw n ∂ w + f (x. F.5. w = C1 xt + C2 x + C3 t + C4 . 2 k(2b − aλ) The books by Polyanin and Zaitsev (1995. Polyanin and V. . .10 with f (w) = bw n .3. z = x + aC1 t2 + aC2 t. . ∂t2 = aeλx ∂x2 + beλx wn ∂2w ∂2w ∂w 2 2 3. C5 are arbitrary constants. 3◦ . D. f 2 (t) where C1 . 2 ∂t ∂x2 This is a special case of equation 3.4. Generalized separable solution: w = (x2 + C1 x + C2 )f (t) + (C3 x + C4 )f (t) dt . C4 are arbitrary constants. a(t + C3 )2 where C1 . w= 1◦ . For b ≠ 1 aλ. 8. . ' ¦& Reference: A. . © 2004 by Chapman & Hall/CRC . 2003) give more than 20 exact solutions to equation (2) for speciﬁc values of the parameter m. ∂2w ∂2w = aw . ∂x This is a special case of equation 3. Solutions: 3x2 + C1 x + C2 + C4 (x + C5 )(t + C3 )3 .

are also solutions of the equation (the plus or minus signs are chosen arbitrarily). C2 t + C4 ).2 with f (x) = beλx . . and C2 are arbitrary constants. where the function u = u(ξ) is determined by the ordinary differential equation 2u − 2ξuξ + 4ξ 2 uξξ = auuξξ + b. + beλx w5 . a > 0. 1 n ( ( ∂2w = aw4 ∂2w ∂2w = aw4 ∂2w ∂2w = awn ∂2w A2 (n + 2) k= a(2 − n) . 3. 5◦ . at2 t t 4 The ﬁrst solution is degenerate and the second one is a generalized separable solution (another arbitrary constant can be added. C2 . ∂t2 ∂x2 This is a special case of equation 3. t) = f (t)x2 + g(t)x + h(t). t) = k(x + 2 C1 ) n (At 2 + C 2 )− n . C1 t + C3 ). Suppose w(x. z = x + aC1 t2 + aC2 t. Then the functions −2 2 w1 = C1 w( C1 x + C2 . . . where C1 and C2 are arbitrary constants and the function U (z) is determined by the autonomous ordinary differential equation 2 2 (aU − a2 C2 )Uzz − 2aC1 Uz = 8aC1 − b. © 2004 by Chapman & Hall/CRC . t) = t2 u(ξ). t) is a solution of this equation. Self-similar solution: w(x. C2 C4 1 3x2 + C1 t3 + 2 x + C3 t3 + 2 − bt2 .1. ξ = xt−2 .4.3. . w= 3◦ . ( ( where C1 .2 with f (x) = bxn . 4. + bxn w5 .4. where C1 . Multiplicative separable solution: w(x.4. ∂t2 ∂x2 This is a special case of equation 3. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). Then the functions 5. t) is a solution of this equation. The second solution in Item 2◦ is a special case of the generalized separable solution w(x. w1 = (C2 /C1 )2/n w( C1 x + C3 . (1) where A. 2◦ . Solution: 2 w = U (z) + 4aC1 t2 + 4aC1 C2 t. 2◦ . ∂t2 ∂x2 1◦ . Suppose w(x. C1 . . EQUATIONS WITH POWER-LAW NONLINEARITIES 203 2. 4◦ . C4 are arbitrary constants. since the equation is invariant under translation in t). ∂t2 ∂x2 1◦ . and C3 are arbitrary constants. Solutions: 1 w = (C1 t + C2 )x + 2 bt2 + C3 t + C4 . ∂2w ∂2w = aw + b.4.

y = (x + A)eλnt . 1◦ . w(x. fxx − λf 1−n = 0. t+B √ w(x. and λ are arbitrary constants. ξ = x + k ln(At + B) + C.204 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3◦ . 1 ¦0 ) w(x. The general solutions of equations (2) and (3) can be written out in implicit form: C1 + C3 + 2aλ n+2 g n+2 2λ 2−n f 2−n −1/2 (2) (3) −1/2 where C1 . t) = 2 aA2 t2 + Bt + Ax + C. z = (x + B)(t + A)k . t) = A(x + aλt) + B + aλ2 .1. .6. Equations of the Form ∂ w = a ∂ w n ∂w ∂t2 ∂x ∂x 1. . ∂2w ∂t2 =a ∂ ∂x w ∂w ∂x . 2 x+A . . t) = f (t)x2 + g(t)x + h(t). + f (w) . 2 3. t) = (At + B)−2/n V (ξ). with C1 = 0. Expression (1) is a special case of a wider family of multiplicative separable solutions w = f (x)g(t). and λ are arbitrary constants.5 with n = 1. B. t) = −2 1 12 aA (At + B)4 + Ct + D + x(At + B). t) = (t + A)− 2k+2 n F (z). 2(n + 2) 4◦ . K. Generalized separable solution quadratic in x: w(x. C4 are arbitrary constants. B. w(x. D. 1 a where A. Solutions: 1 w(x. w(x. C. where the functions f = f (x) and g = g(t) are determined by solving the equations gtt − aλg n+1 = 0. This is a special case of equation 3. C. where A. Reference: S. ◦ 2 . . it follows that g(t) = (At + C)−2/n . Tomotika. A= ) aλn2 . t) = e−2λt U (y).1. © 2004 by Chapman & Hall/CRC ) df = C4 ) dg = C2 t.6. k. x. There are also solutions with the following forms: w(x. Tamada (1950). t) = w(x. In particular. t) = (At + B) Cx + D.

Solution: 2 w = U (z) + 4aC1 t2 + 4aC1 C2 t.1. 2◦ . g = g(t). C4 are arbitrary constants. 2 where C1 . . Suppose w(x. Self-similar solution: w(x. t2 h= 2 aC1 C3 1 1 + + C4 t2 + aC1 C2 t3 + aC 2 t8 . and. where the function u = u(ξ) is determined by the ordinary differential equation 2u − 2ξuξ + 4ξ 2 uξξ = a(uuξ )ξ + b. and h = h(t) are determined by the system of ordinary differential equations ftt = 6af 2 . ξ = xt−2 . . . hence. This is a special case of equation 8. the equation is integrable by quadrature. Traveling-wave solution in implicit form: ak 2 w2 − 2λ2 w = −b(kx + λt)2 + C1 (kx + λt) + C2 . htt = 2af h + ag 2 . gtt = 6af g. Solution: 2 w = U (z) + 4aC1 t2 + 4aC1 C2 t. ∂ ∂w ∂2w =a w 2 ∂t ∂x ∂x + b. the equation can be made homogeneous. C1 t + C3 ). By appropriate translations in both variables. and C3 are arbitrary constants. v) = au2 + b. 2. © 2004 by Chapman & Hall/CRC . where C1 and C2 are arbitrary constants and the function U (z) is determined by the ﬁrst-order ordinary differential equation 2 2 (U − aC2 )Uz − 2C1 U = 8C1 z + C3 . 3◦ . where C1 and C2 are arbitrary constants and the function U (z) is determined by the ﬁrst-order ordinary differential equation 2 2 (aU − a2 C2 )Uz − 2aC1 U = (8aC1 − b)z + C3 . Another arbitrary constant can be inserted in this solution through the shift in t.1. Then the functions 2 −2 w1 = C1 w( C1 x + C2 . 2 4t t 2 54 2 where C1 .2. k. the equation is integrable by quadrature. where C1 . and. C2 . z = x + aC1 t2 + aC2 t. the equation can be made homogeneous. 4◦ .3. z = x + aC1 t2 + aC2 t. .3 with F (u. 3◦ . 1◦ . and λ are arbitrary constants. A particular solution of this system is given by f= 1 . t) = t2 u(ξ). C2 . EQUATIONS WITH POWER-LAW NONLINEARITIES 205 where the functions f = f (t). hence. at2 g= C1 + C 2 t3 . are also solutions of the equation. t) is a solution of the equation in question. By appropriate translations in both variables.

can be inserted in the solution. √ ∂x w ∂x This is a special case of equation 3. k. t) = w(x. t) = w(x. 2 ∂x a ∂t ∂t 3 2 © 2004 by Chapman & Hall/CRC . since the system is translation invariant in t. B. + B)4 + Cx + D + t(Ax + B) . at2 t 4t2 t 2 54 2 9 where C1 . w(x. Solutions: w(x. t) = (aA2 t2 + Bt + C) sinh−2 (Ax + D). . w(x. ∂x w ∂x This is a special case of equation 3.6. 4. Another arbitrary constant. t) = a 2 1 2 2 A ax ∂2w + Bx + Aat + C . t) = f (t)x2 + g(t)x + h(t).1. . t) = (−aA2 t2 + Bt + C) cosh−2 (Ax + D). h = + + C4 t2 + aC1 C2 t3 + aC 2 t8 + bt2 3 ln |t| − 1 . . The substitution w = u2 leads to an equation of the form 3. 2◦ . B. w(x. t) = (At + B)eCx . w(x. A(t + λx) + B + aλ2 . Generalized separable solution quadratic in x: w(x. D.1. 2◦ . and D are arbitrary constants. htt = 2af h + ag 2 + b. ∂t2 =a ∂ 1◦ . C. w(x. . where the functions f = f (t).5 with n = −1. t) = (Ax + B)2 (Ct + D).6. C4 are arbitrary constants. t) = (aA2 t2 + Bt + C) cos−2 (Ax + D). t) = where A.1. C2 .5 with n = −1/2. and h = h(t) are determined by the system of ordinary differential equations ftt = 6af 2 . ∂2w =a ∂ 1 ∂w 1◦ . and λ are arbitrary constants. C5 . Traveling-wave solution in implicit form: λ2 w = ak 2 ln |w| + C1 (kx + λt) + C2 . g = 2 + C 2 t3 .206 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 5◦ . g = g(t). . where C1 . Multiplicative separable solutions: where A. ∂t2 w(x. and λ are arbitrary constants.2: ∂u ∂2u 1 ∂ = u . f= 3. 1 ∂w . t) = (at2 + At + B)(x + C)−2 . A particular solution of this system is given by 2 C1 aC1 C3 1 1 1 1 . gtt = 6af g. by substituting t + C5 for t. C. w(x.4. 2 2 −2 −1 1 12 A a (Ax 4 t+A 2 x+B .

Multiplicative separable solution: w = f (x)g(t). B. where C1 . or 3. . Then the functions w1 = (C2 /C1 )2/n w( C1 x + C3 . equation (4) can be solved for w to give an explicit expression of w = w(z). . This is a special case of equation 3. EQUATIONS WITH POWER-LAW NONLINEARITIES 207 ∂ ∂w ∂2w = a2 wn . 3◦ . 2◦ . C2 t + C4 ). .4. (1) (2) −1/2 and C1 . z=x λt. Degenerate solution: where A.6 with f (w) = a2 wn . see Kamke (1977) and Polyanin and Zaitsev (1995. If C ≠ 0.3. (3) . 4◦ . . Self-similar solution: 4 w = w(z). 2. C. t+B where the function w(ξ) is determined by the ﬁrst-order ordinary differential equation (C is an arbitrary constant): ξ 2 − a2 wn wξ = C. C4 are arbitrary constants. 1. 4 dg = C4 t. . . C4 and λ are arbitrary constants. 2 x+A . and s are arbitrary constants. n+1 (4) If n = − 1 . (5) w = w(ξ). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). . 5. by treating w in (5) as the independent variable. The functions f = f (x) and g = g(t) deﬁned by (1) and (2) can be represented in explicit form if C1 = 0 and C3 = 0. Suppose w(x. . t) = where b. and D are arbitrary constants.4. 1◦ . where w = w(z) is deﬁned implicitly by (A and B are arbitrary constants) λ2 w − a2 wn+1 = Az + B. © 2004 by Chapman & Hall/CRC 5◦ . where f = f (x) and g = g(t) are deﬁned implicitly by 4 4 4 w = (At + B)(Cx + D) n+1 . 2 ∂t ∂x ∂x This equation is encountered in wave and gas dynamics. 2003). 1 C1 + C3 + 2λ n+2 f n+2 2a2 λ n+2 g n+2 −1/2 f n df = C2 x. c. see formula (3). (6) The general solution of equation (6) is expressed in terms of Bessel functions. To the special case C1 = C3 = 0 there corresponds 4 w(x. t) is a solution of the equation in question. one obtains a Riccati equation for ξ = ξ(w): Cξw = ξ 2 − a2 wn .1. Traveling-wave solution: bx + c abt + s 2/n . ξ= To the special case C = 0 there corresponds the solution w = (ξ/a)2/n .

1. Polyanin (1996). the change of variable H(ψ) = ψu brings (7) to a ﬁrst-order equation. Solution (A. β. the transformation τ = x. References for equation 3. For n = −1. Adams (1981). the transformation τ = x. and E. and the function F = F (z) is determined by solving the generalized-homogeneous ordinary differential equation 2k(2k − 1)F + (nk + 1)(nk − 4k + 2)zFz + (nk + 1)2 z 2 Fzz = a2 (F n Fz )z . (t + β)nk+1 where α. 9◦ . and k are arbitrary constants. ζ = t. D. where the function ϕ = ϕ(y) is determined by solving the generalized-homogeneous ordinary differential equation 4µ2 ϕ + µ2 n(n − 4)yϕy + (µn)2 y 2 ϕyy = a2 (ϕn ϕy )y . equation (7) admits a ﬁrst integral that represents a separable equation. z= x+α . with the change of variable G(ψ) = (ψu )2 . In the general case. There are more complicated self-similar solutions of the form w = (t + β)2k F (z). N.3: ∂ ∂V ∂2V = a−2 eV 2 ∂τ ∂ζ ∂ζ 7 ¦6 . For n = −2. F. Lohner. Generalized self-similar solution (µ is an arbitrary constant): w = e−2µt ϕ(y). For n ≠ −1. V = ln w brings the original equation to an equation of the form 3. © 2004 by Chapman & Hall/CRC . y = xeµnt . J.6. Its order can be reduced. ζ = t. Ames. 8◦ . H. 5 u = x + b ln( t + A) + c. equation (7) can be reduced to a ﬁrst-order linear equation. V.5: W. 7◦ . For n = −4.4. R. Ibragimov (1994). 2 n n (7) Note two special cases where the equation obtained is integrable by quadrature. where the function ψ = ψ(u) is determined by the autonomous ordinary differential equation b(n + 4) 2(n + 2) ψ− ψu + b2 ψuu = a2 (ψ n ψu )u .2.208 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 6◦ . b. V = wn+1 brings the original equation to an equation of the similar form n ∂V ∂ ∂2V = a−2 V − n+1 2 ∂τ ∂ζ ∂ζ . and c are arbitrary constants): 5 w = ( t + A)−2/n ψ(u). Zaitsev and A. Its order can be reduced. F.

A2 . C1 t + C3 ). t) = U (z). 1◦ . and B2 are arbitrary constants. Other Equations 1.1. v) = bun + c. There are solutions of the following forms: w(x. ∂ ∂w ∂2w =a wn 2 ∂t ∂x ∂x + bwk . and C3 are arbitrary constants. where the functions ϕ = ϕ(x) and ψ = ψ(t) are determined by the autonomous ordinary differential equations (K is an arbitrary constant) a(ϕn ϕx )x + bϕn+1 + Kϕ = 0. A2 . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). if b(n + 1)/a = −k 2 < 0. Suppose w(x.3. 8 where C1 . ∂2w ∂t2 = aw ∂2w ∂x2 +b ∂w ∂x n + c.3 with F (u. 1◦ . 1◦ .2. B1 .7. V (ξ). t) = 2 t 1−k z = λx + βt ξ= k−n−1 xt 1−k traveling-wave solution. Suppose w(x. C2 . B1 . ∂ ∂w ∂2w =a wn 2 ∂t ∂x ∂x + bwn+1 + cw. are also solutions of the equation. Then the functions k−n−1 k−1 2 x + C2 . Multiplicative separable solutions for c = λ2 > 0: w = A1 eλt + A2 e−λt B1 cos(kx) + B2 sin(kx) w = A1 eλt + A2 e−λt B1 ekx + B2 e−kx where A1 . and B2 are arbitrary constants. © 2004 by Chapman & Hall/CRC . if b(n + 1)/a = k 2 > 0. Multiplicative separable solution: w = ϕ(x)ψ(t). This is a special case of equation 8. t) is a solution of the equation in question. Then the functions 2 −2 w1 = C1 w( C1 x + C2 . 1 n+1 1 n+1 1 n+1 1 n+1 if b(n + 1)/a = k 2 > 0. EQUATIONS WITH POWER-LAW NONLINEARITIES 209 6. 7. w1 = C1 w( C1 8 8 where C1 . 2◦ . w(x. C2 . Multiplicative separable solutions for c = −λ2 < 0: w = A1 cos(λt) + A2 sin(λt) B1 cos(kx) + B2 sin(kx) w = A1 cos(λt) + A2 sin(λt) B1 ekx + B2 e−kx where A1 . ψtt − cψ + Kψ n+1 = 0. 3. t) is a solution of this equation. 2◦ .1. C1 t + C3 ). if b(n + 1)/a = −k 2 < 0.1. self-similar solution. 3◦ . and C3 are arbitrary constants.

Then the functions m 2−n 2 2 w1 = C1 w C2 x. ∂2w = axn wm ∂2w . w(x. 2003) give more than 20 exact solutions to the Emden–Fowler equation (3) for speciﬁc values of the parameter m. In particular. C2 . 2. z = x + aC1 t2 + aC2 t. ◦ 3 . The transformation u(z. Self-similar solution: w(x.210 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . 2(m + 2) (4) . ∂t2 ∂x2 1◦ . 2−n kx m 2 (C1 t + C2 )− m . Suppose w(x. ξ = xt−2 . leads to an equation of the similar form ∂2u ∂2u = az 4−n−m um 2 . 1 9 (1) (2) (3) fxx − (λ/a)x−n f 1−m = 0. 2 ∂t ∂z In the special case n = 4 − m. and C3 are arbitrary constants. where the functions f = f (x) and g = g(t) are determined by solving the equations gtt − λg m+1 = 0.6.5. t). are also solutions of the equation. where C1 . 2◦ . C1 C2 t + C3 . The general solution of equation (2) can be written out in implicit form as 9 2λ m+2 g m+2 where C1 and C2 are arbitrary constants. Solution: 2 w = U (z) + 4aC1 t2 + 4aC1 C2 t. 4◦ . x z= 1 z (n−2)k−2 m F (y). t) is a solution of this equation. Expression (1) is a special case of a wider family of multiplicative separable solutions w = f (x)g(t). see also equation 3. equation (4) is greatly simpliﬁed to become ∂2u ∂2u = aum 2 ∂t2 ∂z and admits a traveling-wave solution u = u(kz + µt). t) = 1 w(x. with C1 = 0. it follows that C1 + g(t) = (At + C)−2/m . y = xtk . 5◦ . t) = t2 u(ξ).1. −1/2 dg = C2 t. where C1 and C2 are arbitrary constants and the function U (z) is determined by the autonomous ordinary differential equation 2 2 (aU − a2 C2 )Uzz + b(Uz )n − 2aC1 Uz = 8aC1 − c. t) = a(2 − n)(2 − n − m) 3◦ . Multiplicative separable solution: 2 m 2C1 (m + 2) k= . There is a self-similar solution of the form w=t where k is an arbitrary constant. where the function u = u(ξ) is determined by the ordinary differential equation 2u − 2ξuξ + 4ξ 2 uξξ = a(uuξ )ξ + b(uξ )n + c. © 2004 by Chapman & Hall/CRC 9 A= λm2 . The books by Polyanin and Zaitsev (1995. a > 0.

where the functions f = f (x) and g = g(t) are determined by solving the equations gtt − λg m+1 = 0. t) = U (z). (1) (2) (3) df = C2 t. EQUATIONS WITH POWER-LAW NONLINEARITIES 211 3.4. 3◦ . Multiplicative separable solution: w(x. are also solutions of the equation. t) is a solution of this equation.1. ∂ ∂w ∂2w =a xn w m . For n ≠ 1 and m ≠ −1. 2(m + 2) ϕ = um+1 . Suppose w(x. C3 . where C1 . 2◦ .5. −1/2 The general solution of equation (2) can be written out in implicit form as @ C1 + 2λ m+2 f m+2 where C1 and C2 are arbitrary constants. is also a solution of the equation. 2◦ . 1◦ . C2 t + C3 .1 with f (x) = aeλx . 4. 2 2C1 (m + 2) k= a(2 − n)(m − n + 2) 1 m 3◦ . brings (3) to the Emden–Fowler equation 1 λ(m + 1) n z 1−n ϕ m+1 . ∂2w ∂2w = aeλx wm . A= λm2 . ϕzz = © 2004 by Chapman & Hall/CRC @ .4. the transformation z = x1−n .2 with f (x) = axn .3. For other solutions. see equation 3. t) is a solution of the equation in question. λ C2 where C1 . with C1 = 0. a > 0. (4) 2 a(1 − n) The books by Polyanin and Zaitsev (1995. and C3 are arbitrary constants. it follows that @ f (t) = (At + C)−2/m . C1 C2 t + C3 . Generalized traveling-wave solution: 2 ln |t|. Then the function 1 Cm w1 = C1 w x + ln 12 . λ where the function U (z) is determined by the ordinary differential equation w(x. C2 . Suppose w(x. t) = 2−n kx m 2 (C1 t + C2 )− m . 2003) give more than 20 exact solutions to equation (4) for speciﬁc values of the parameter m. 2 ∂t ∂x ∂x This is a special case of equation 3. ∂t2 ∂x2 1◦ . z =x+ (aλ2 eλz U m − 4)Uzz + 2λUz = 0.5. Expression (1) is a special case of a wider family of multiplicative separable solutions w = f (x)g(t). Then the functions m 2−n 2 2 w1 = C1 w C2 x. In particular. a > 0. C2 . a(x f n m+1 fx )x − λf = 0. and λ are arbitrary constants.

b = 0. In the special case n = 2m + 3 with equation 3. m+1 3m + 4 .212 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 4◦ . Suppose m ≠ −1 and 2m − 2n − nm + 3 ≠ 0. ∂t2 ∂ξ ∂ξ (2) where the function F (ξ) is deﬁned parametrically by dx k . ξ= . c = 1. 1◦ . t) ax2 + bx + c.4.8: z= ax2 dx + bx + c ∂2u ∂2u 1 = ku4−2m 2 + k(ac − 4 b2 )u5−2m .1. 5◦ . The transformation w(x. m = 1. 1 ξ= dx (ax2 + bx + c)m (1) the original equation can be reduced to the divergence form 4−2m ∂v ∂ ∂ 2v = F (ξ)v 2m−3 . up to notation. 2 © 2004 by Chapman & Hall/CRC . a = −1.5: ∂ ∂u ∂ 2u =A um . m = 1 .4. t) = x m+1 u(ξ. 2◦ . 2 ∂t ∂ξ v 2 ∂ξ F (ξ) = (ax2 ∂ ∂2v =k ∂t2 ∂ξ ∂ ∂2v =k 2 ∂t ∂ξ cosh2 ξ ∂v . equation (5) is greatly simpliﬁed and coincides. leads to an equation of the form 3. There is a self-similar solution of the form w = (t + b) where b and k are arbitrary constants. ∂t2 ∂x2 This is a special case of equation 3. By the transformation w(x. 2 ∂t ∂ξ ∂ξ 1−n (n−2)k−2 m F (y).4. leads to an equation of the similar form 3m−3n−2nm+4 ∂ ∂u ∂2u =A ξ 2m−2n−nm+3 um . c = 1. ∂t2 ∂ξ ∂ξ ∂2w = k(ax2 + bx + c)m w4–2m ∂2w 2 5. t) = u(z. t) = v(ξ. The transformation √ w(x. v 2 ∂ξ ξ −3/2 ∂v . . 2 ∂t ∂z which has a traveling-wave solution u = u(z +λt) and a multiplicative separable solution u = f (t)g(z). (3) m 2 + bx + c)m + bx + c) (ax Note some special cases of equation (2) where F = F (ξ) of (3) can be represented in explicit form: ∂ cos2 ξ ∂v ∂2v =k . a = 1.6.4 with f (u) = ku−2m . ξ=x 2m−2n−nm+3 m+1 (5) where A = a 2m − 2n − nm + 3 . b = 0. cos ξ ∂ξ m = 1. b = 0. c = 1.5. t). a = −1. t)] 2m+3 . y = xtk .

C2 . 2◦ . ∂z∂y © 2004 by Chapman & Hall/CRC y = x + at . leads to the Liouville equation 3. t) = − 2 ln C1 eaλt β 2 w(x. The change of the independent variables z = x − at. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). + beβw . B. C2 . C.2: ∂2w 1 = − 4 a−2 b exp(βw). Equations of the Form ∂ w = a ∂ w + beβw + ceγw 2 ∂t ∂x2 1. t) = − ln C1 eaλt β 2 − ln (x + A)2 − a2 (t + B)2 + C . 2aλ √ −2bβ cosh(aλt + C2 ) .2. t cosh λ + a−1 x sinh λ). Suppose w(x. 1◦ . and C are arbitrary constants. ∂t2 ∂x2 This is a special case of equation 3. C3 . 4◦ . 2aλ where A. β w2 = w(x cosh λ + at sinh λ. and λ are arbitrary constants.1. t) = ln β bβ cos2 (Ax + Bt + C) where A.1.4. t) = ln β bβ sinh2 (Ax + Bt + C) 2(B 2 − a2 A2 ) 1 . Then the functions 2 ln |C1 |. 2aλ √ 2bβ cosh(λx + C2 ) . β √ 2bβ sinh(aλt + C2 ) . Functional separable solutions: w(x. t) = − 2 ln C1 eλx β 2 w(x.2. t) = 2(B 2 − a2 A2 ) 1 ln .2. w(x. B. t) is a solution of the equation in question. t) = w(x. Traveling-wave solutions: w(x. β bβ(Ax + Bt + C)2 2(a2 A2 − B 2 ) 1 . EQUATIONS WITH EXPONENTIAL NONLINEARITIES 213 3. t) = − ln C1 eλx β w(x. ln β bβ cosh2 (Ax + Bt + C) 2(B 2 − a2 A2 ) 1 . w1 = w( C1 x + C2 . 2aλ √ −2bβ sinh(λx + C2 ) . and λ are arbitrary constants. w(x.1 with f (w) = beβw .3. t) = 8a2 C 1 ln β bβ A A A A w(x. C1 t + C3 ) + A A ∂2w = a2 ∂2w where C1 . Equations with Exponential Nonlinearities 2 2 3.1.5. 3◦ . C1 .

k= b 1 ln 2β a leads to an equation of the form 3. R. V. A. ∂t2 ∂x2 1◦ . V. t) + k. C2 . where C1 and C2 are arbitrary constants. β ∂2w ∂2w where the functions ψ(t) and ϕ(x) are determined by the ﬁrst-order autonomous ordinary differential equations (ϕx )2 = −2aβϕ3 + C1 ϕ2 − C2 ϕ + C3 − bβ.1: √ ∂2u ∂2u = + 2 ab sinh(βu). © 2004 by Chapman & Hall/CRC . V. M. Zaitsev and A. F. one obtains separable equations. t) = − b 1 ln − + C1 exp a β a β x sinh C2 b t cosh C2 . the general solution of the original equation is expressed as w(x. Traveling-wave solution (generalizes the solution of Item 1◦ ): w(x. Bullough and P.1. K. t) = 2 bβ 1 f (z) + g(y) − ln k exp f (z) dz − 2 β β 8a k z = x − at. (ψt )2 = 2aβψ 3 + C1 ψ 2 + C2 ψ + C3 . Functional separable solution: = w(x. V. where f = f (z) and g = g(y) are arbitrary functions and k is an arbitrary constant. 2 2 ∂t ∂x 4. and C3 are arbitrary constants. + aeβw – be–2βw . Rodionov (1999). C2 . + aeβw + be2βw .214 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE Hence. R. Liouville (1853). V. ∂t2 ∂x2 1◦ . Solving these equations for the derivatives. K. O. and A. y = x + at. exp g(y) dy . Andreev. 3◦ . Traveling-wave solution for bβ > 0: 2. Kaptsov. Caudrey (1980). t) = − 3. J. = D ∂2w ∂2w w(x. t) = 1 ln[ϕ(x) + ψ(t)]. Pukhnachov. t) = − 2 2 aβ a2 β 2 + bβ(C1 − C2 ) 1 ln 2 + C3 exp(C1 x + C2 t) + exp(−C1 x − C2 t) . C ¦B References: A.3. aβ 1 + ln 2 2 β C2 − C1 2 2 a2 β 2 + bβ(C2 − C1 ) sin(C1 x + C2 t + C3 ) . D. ∂t2 ∂x2 The substitution w(x. where C1 . Polyanin (1996). 2 2 2 β C1 − C2 4C3 (C1 − C2 )2 where C1 . t) = u(x. 2 − C2 C2 1 ∂2w ∂2w = + aeβw – be–βw . 2◦ . Infeld (1992). Z. Traveling-wave solution: w(x. Grundland and E. C ¦B References: J. and C3 are arbitrary constants. Zhdanov (1994).

and σ are arbitrary constants. Solutions: β w(x. © 2004 by Chapman & Hall/CRC . This is a special case of equation 3.1: ∂ 2U ∂2U = + aλeλU . Fordy and J. ∂2w ∂t2 = ∂2w ∂x2 + aeβx eλw .4. ∂2w ∂t2 = ∂2w ∂x2 + aeβt eλw . Gibbons (1980). t) = − t − λ 2 1 2 ln C1 + C2 x C2 + 2 λa t .8 with f (w) = ceλw . Solutions: β w(x.6 with f (w) = aeλw .3. ∂ξ∂η 3◦ . 2◦ .1: ∂ 2U ∂2U = + aλeλU . The substitution λU = λw + βt leads to an equation of the form 3. λ 8σ 2 C1 G where C1 . The equation can be integrated with the inverse scattering method. P. F. V. t) = − x − λ β w(x. This is a special case of equation 3. A. 2◦ . The transformation t = (a2 bβ 3 )−1/6 (ξ + η). x = (a2 bβ 3 )−1/6 (ξ − η). t. w) 2 ∂t ∂x2 1. C2 . ∂t2 ∂x2 2. 1◦ . λ 2 λa σt ln C1 e−σt + C2 eσx − e . Degasperis (1982). F ¦E References: A. and σ are arbitrary constants. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 215 2◦ .7 with f (w) = aeλw . The substitution λU = λw + βx leads to an equation of the form 3. t) = − x − λ 2 1 2 ln C1 + C2 t C2 − 2 λa x .1. Calogero and A. C2 .2. t) = − t − λ β w(x.2. w= 1 b 1 U+ ln β 3β a leads to an equation of the form 3.2. This is a special case of equation 3.3: ∂2U = eU − e−2U .1.1. Equations of the Form ∂ w = a ∂ w + f (x. ∂2w ∂t2 = ∂2w ∂x2 + ceax+bt eλw .2.5. λ 2 λa σx ln C1 e−σx + C2 eσt + e . 2 2 3.4.2.1. 1◦ .1. Mikhailov (1979). A. 2 ∂t ∂x2 3. λ 8σ 2 C1 G where C1 .1.4.

∂t2 ∂x2 Functional separable solutions: 5. ∂2w ∂2w = + βeλw + (αx + γ)e2λw . I. t) = − P ¦I αk 1 ln Cekx + (x λ 2β t)ekx + λβ . A. Vyazmin. Zhurov. Kazenin (1998). H H w(x. and D. t) = − β2λ 1 ln C exp (x λ 2α t) − α2 1 (αx + γ) − 3 . Kazenin (1998).216 4. V. Kazenin (1998). t) = − 1 ln C exp λ k2γ + β 2λ αβλ γ k2γ − β 2λ x+ t + 2 ekt − . Functional separable solutions for k 2 γ − β 2 λ ≠ 0: = H ∂2w ∂2w w(x. t) = − 1 ln C exp λ k2γ − β 2 λ αβλ γ k2γ + β 2 λ t+ x − 2 ekx − . A. t) = − P ¦I αk λβ 1 ln Cekt + (t x)ekt − 2 . V. H w(x. Generalized traveling-wave solutions for k 2 γ − β 2 λ = 0: H w(x. 2◦ . Polyanin. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w = + βeλw + (αt + γ)e2λw . D. k ∂2w ∂2w = + βeλw + (αekx + γ)e2λw . ln C exp − (t x) − (αt + γ) + 3 λ 2α β β λ where C is an arbitrary constant. A. t) = − where C is an arbitrary constant. 2kγ 2kγ k γ − β2λ β where C is an arbitrary constant. © 2004 by Chapman & Hall/CRC . V. and D. Vyazmin. ∂t2 ∂x2 Functional separable solutions: H w(x. A. and D. ∂t2 ∂x2 1◦ . 2kγ 2kγ k γ + β2λ β where C is an arbitrary constant. A. t) = − β2λ α2 1 1 . Zhurov. α 1 ln Cekt + (t λ 2β x)ekt − λβ . A. I. A. A. I. Polyanin. 6. Zhurov. Generalized traveling-wave solutions for k 2 γ + β 2 λ = 0: w(x. k2 Reference: A. 8. D. Functional separable solutions for k 2 γ + β 2 λ ≠ 0: H w(x. Vyazmin. Polyanin. 2◦ . ∂2w ∂2w = + kβeλw + (αekt + λβ 2 )e2λw . β β λ where C is an arbitrary constant. + βeλw + (αekt + γ)e2λw . P ¦I Reference: A. ∂t2 ∂x2 1◦ . A. D. λ 2β k Reference: A. ∂t2 ∂x2 Generalized traveling-wave solutions: 7.

t) = − 1 ln C exp λ β2λ βγλ α 4k 2 α − β 2 λ x− t + 2 e−kt − ekt . are also solutions of the equation. C2 . w. C 1 t + C 3 + Q 1 ln C1 . This is a special case of equation 3. D. Vyazmin. and D. Zhurov. t) = F (z) − 1 ln |t|. ∂t2 ∂x2 Functional separable solutions: = Q ∂2w ∂2w w(x.4. A. 2 ∂t ∂x2 Functional separable solutions: Q w(x. t) = − where C is an arbitrary constant. A. I. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 217 9. 10. A. Suppose w(x. k ∂2w ∂2w = + βekt eλw + (αe2kt + γ)e2λw . Traveling-wave solution: w=− exp(Ax + Aµt + B) − b 1 ln .2. V. and C3 are arbitrary constants.3. t. ∂2w ∂t2 =a ∂2w ∂x2 + beλw ∂w ∂x .3. and B are arbitrary constants. ∂t2 ∂x2 Functional separable solutions: Q w(x. Kazenin (1998). 2◦ .2. There is an exact solution of the form w(x. α 1 ln Cekx + (x λ 2β t)ekx − λβ . ∂w 2 ∂t ∂x2 ∂x 1. Polyanin. Equations of the Form ∂ w = f (x) ∂ w + g x. t) = − 1 ln C exp λ β2λ βγλ α 4k 2 α + β 2 λ t+ x − 2 e−kx − ekx . λ A(a − µ2 ) where µ. 2 2 3. t) is a solution of the equation in question. + βekx eλw + (αe2kx + γ)e2λw . 4kα 4kα 4k α + β 2 λ β where C is an arbitrary constant. ∂2w ∂2w = – kβeλw – (αekx + λβ 2 )e2λw . λ z= x .2. λ where C1 . 3◦ .3 with f (w) = beλw . 1◦ . Then the functions w1 = w C1 x + C2 . 11. S ¦R Reference: A. 4kα 4kα 4k α − β 2 λ β where C is an arbitrary constant. t © 2004 by Chapman & Hall/CRC . A.

1◦ . t © 2004 by Chapman & Hall/CRC . λ where C1 and C2 are arbitrary constants. are also solutions of the equation. are also solutions of the equation. a > 0. xn 1◦ . λ where C1 and C2 are arbitrary constants.2. There is an exact solution of the form w(x. Then the functions w1 = w C12−n x. a(2 − n)2 4 . 3◦ . r2 = 4k where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 2 A A= . where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation c λw n+1 wr + e = 0. C1 t + C2 + 2◦ . t) is a solution of the equation in question. Functional separable solution for n ≠ 2 (generalizes the solution of Item 2 ◦ ): w = w(r). ∂2w ∂t2 = a ∂ xn ∂w 2 ln |t|. r2 = k x2 − a(t + C)2 .5 with b = an. T 2 2 ln C1 . t) = F (z) − 2 ln |t|. Suppose w(x. r 2−n 4◦ . Suppose w(x. a > 0. wrr + r ak 4◦ . Functional separable solution for n ≠ 2 and λ ≠ 0: w=− 2cλ(2 − n) x2−n 1 1 ln − (t + C)2 λ n a(2 − n)2 4 1 x2−n − (t + C)2 . t) is a solution of the equation in question. 2◦ . There is an exact solution of the form w(x.218 2. wrr + wr + ck −1 eλw = 0. λ z = x|t| n−2 . t) = F (z) − 3. 2 ∂t ∂x ∂x This is a special case of equation 3.3. 3◦ . Then the functions w1 = w C1 x. 2 + ceλw . C1 t + C2 + T 2 ln C1 . the equation describes the propagation of nonlinear waves with axial and central symmetry. Functional separable solution (generalizes the solution of Item 2 ◦ ): w = w(r). ∂x ∂x For n = 1 and n = 2. λ z= x . respectively. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w ∂2w = axn + ceλw . Functional separable solution for n ≠ 0 and λ ≠ 0: w=− cλ 2 1 ln x − a(t + C)2 λ 2an .

3. (2) wzz = 2 (1 − A) In the special case A = 1 . Suppose w(x. w(z) = ln λ 2kcλ cos2 (pz + q) where p and q are arbitrary constants.5 with b = 0: 4. B= . λ cλξ cos2 (p ln |ξ| + q) −2ap2 (2 − n)2 1 . ∂t2 ∂x2 This is an equation of the propagation of nonlinear waves in an inhomogeneous medium. which corresponds to b = a(n − 1). C is an arbitrary constant. ∂x 1◦ . w1 = w C1 x. exact solutions of equation (1) are expressed as 2 w(ξ) = w(ξ) = w(ξ) = where p and q are arbitrary constants. are also solutions of the equation.2. 5. ξ = 1 a(2 − n)2 (t + C)2 − x2−n . w(z) = ln λ 2kcλ(z + q)2 ap2 (2 − n)2 1 .3. ∂t2 ∂x2 2◦ . ln λ cλξ cosh2 (p ln |ξ| + q) 1 For A ≠ 1. Then the functions 4 − 2n 2−n 2 ln C1 . a > 0. ln λ 2kcλ cosh2 (pz + q) −ap2 (2 − n)2 1 . a > 0. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 219 ∂2w ∂2w = a(x + β)n + ceλw . 2 ∂t ∂z + ceλw . 2a(2 − n) a(2 − n)2 For A ≠ 1. 4 Here. the substitution ξ = kz 1−A (k = 1) brings (1) to the generalized Emden–Fowler equation 2A−1 kB z 1−A eλw . solutions of equation (2) are given 2 by −a(2 − n)2 1 . and the function w = w(ξ) is determined by the ordinary differential equation ξwξξ + Awξ = Beλw . C1 t + C2 + λ where C1 and C2 are arbitrary constants. 2a(2 − n)2 1 ln .2. (1) where c a(4 − 3n) + 2b . λ λBξ U ∂2w = axn ∂2w + bxn–1 ∂w For A = 1. an exact solution of equation (1) is given by A= w(ξ) = 1−A 1 ln . ∂2w ∂ 2w = az n 2 + ceλw . which corresponds to b = 1 an. t) is a solution of this equation. λ cλξ(ln |ξ| + q)2 2ap2 (2 − n)2 1 ln . The substitution z = x + β leads to a special case of equation 3. w(z) = © 2004 by Chapman & Hall/CRC U . Functional separable solution for n ≠ 2: w = w(ξ).

Functional separable solution for λ ≠ 0: e−λx 1 − (t + C1 )2 . ln λ cλ cosh2 (py + q) 2p2 (A2 − aB 2 ) 1 . w(y) = ln λ cλ cos2 (py + q) where p and q are arbitrary constants. µ z =x+ 2 ln |t|. λ µ where C1 and C2 are arbitrary constants. ∂x ∂x This is a special case of equation 3.9 with b = aλ.2. µ 2 2kC2 1 − ln cµ cosh2 (C r + C ) if ckµ > 0. w(y) = 6. .3. 2◦ . and the function w = w(y) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wyy + (b − a)Bwy + ceλw = 0. C1 t + C2 + ln |C1 |. 2 3 2 µ 2kC2 where C2 and C3 are arbitrary constants. + ceµw . w(y) = − ln λ B(b − a) Solutions of equation (3) with b = a: 2(A2 − aB 2 ) 1 . ∂t2 aeλx ∂2w = ∂ ∂w ∂2w = axn ∂2w + bxn–1 eλw ∂w V 1◦ .5 with f (w) = beλw .220 3◦ . λ V © 2004 by Chapman & Hall/CRC . B. and C are arbitrary constants. ∂t2 ∂x2 ∂x This is a special case of equation 3. There is an exact solution of the form w(x. w(y) = ln λ cλ(y + q)2 2p2 (aB 2 − A2 ) 1 . are also solutions of the equation. Its general solution is expressed as − 1 ln − cµ (r + C3 )2 if ckµ < 0.4. Solution for n = 2: HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE w = w(y). t) = F (z) − 2 ln |t|. Then the functions 2 2 w1 = w x − ln |C1 |. a > 0. 2 3 µ 2 2kC2 w= 1 − ln − cµ sinh2 (C2 r + C3 ) if ckµ < 0.3. 7. aλ2 4 where C1 and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the autonomous ordinary differential equation wrr + ck −1 eµw = 0. (3) √ Solution of equation (3) with A = B a: cλ 1 y + C1 . Suppose w(x. w = w(r). µ 2k 1 − ln − cµ sin2 (C r + C ) if ckµ < 0. a > 0. y = At + B ln |x| + C. r2 = 4k 3◦ . where A. t) is a solution of the equation in question.

9 with b = 0. ∂2w = aeλx ∂2w + beλx+µw ∂w 11. This is a special case of equation 3. a > 0. ∂t2 ∂x2 This is an equation of the propagation of nonlinear waves in an inhomogeneous medium.10 with f (w) = beµw . k = 1.11 with f (w) = beµw . ∂x This is a special case of equation 3. if b = aλ. . W w1 = w x − 2 2 ln |C1 |.2. λ µ where C1 and C2 are arbitrary constants.4. where C is an arbitrary constant and the function w = w(z) is determined by the ordinary differential equation c µw 2(aλ − b) 1 wz + e = 0. are also solutions of the equation. a > 0. Then the functions 9. ∂t2 ∂x2 ∂2w ∂t2 = aeλx ∂2w ∂x2 + beλx+µw ∂w ∂x + ceβw . ∂x 1◦ . © 2004 by Chapman & Hall/CRC . t) is a solution of this equation. if b = 1 aλ. and g(w) = ceβw . z = 4ke−λx − akλ2 (t + C)2 1/2 . ∂2w ∂2w = aeλx + ceµw . C1 t + C2 + ln |C1 |. a > 0. EQUATIONS WITH EXPONENTIAL NONLINEARITIES 221 8. This is a special case of equation 3. ∂2w ∂t2 = aeλx ∂2w ∂x2 + beλx ∂w + ceµw .3.4.2. a > 0.3. (1) wzz + aλ z akλ2 A solution of equation (1) has the form w(z) = 2kλ(aλ − 2b) 1 ln . 2 2aA2 kλ2 1 ln µ cµ cosh2 (Az + B) −2aA2 kλ2 1 w(z) = ln µ cµ sinh2 (Az + B) −2aA2 kλ2 1 w(z) = ln µ cµ cos2 (Az + B) w(z) = 8ABakλ2 1 ln µ cµ(Az 2 + B)2 where A and B are arbitrary constants. if b = aλ.3. µ cµz 2 Note some other exact solutions of equation (1): w(z) = w(z) = −2akλ2 1 ln µ cµ(z + B)2 if b = aλ. Suppose w(x. if b = aλ. 2◦ .3. 10. Functional separable solution: W w = w(z).

Other Equations . ∂t2 ∂x2 1◦ . D. Solutions: w(x. Suppose w(x. x + C2 2(k − 1) ln(t + C1 ) + f (ζ). U (ξ) = zwz . Then the functions β−λ β w1 = w( C1 x + C2 . C2 . z= (aeλw − z 2 )wzz − zwz = 0. ` ¦Y Reference: V. are also solutions of the equation. t) = ln . while the others are representable as the sum of functions with different arguments. t) = C 2 sinh2 (Ax + B) 1 . © 2004 by Chapman & Hall/CRC X w(x. Zaitsev and A. ln λ aA2 cosh2 (Ct + D) 1 4BCβ 2 (x + A)2 (Aeβx + Be−βx )2 1 . λ w1 = w(C1 C2 x + C3 . Then the functions 1. λ aA2 cos2 (Ct + D) . t) = H(ρ) − t. ln λ aA2 cosh2 (Ct + D) C 2 cos2 (Ax + B) 1 . t) = ln . D. t) is a solution of this equation. The ﬁrst solution is degenerate. B. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). . . C. t) = F (η) − ln |t|. η = x + k ln |t|. t) = w(x. w(x. t) = 1 C 2 cosh2 (Ax + B) . ) − 2 ln |C2 |. t) = ln λ a cosh2 (Bt + C) λ aA2 (t + C)2 2 w(x. x+A . w(x.4. Self-similar solution: 4◦ . F. and the function f = f (ζ) is determined by the ordinary differential equation 2(k − 1) = aeλf fζζ . C2 . 1 (x + A)2 B 1 cosh2 (Ax + B) 1 . w(x. There are exact solutions of the following forms: 2 w(x. ln λ a(Beβt + Ce−βt )2 λ 4aABβ 2 (t + C)2 where A. ln λ aA2 sinh2 (Ct + D) 1 C 2 cosh2 (Ax + B) ln . C1 t + C3 ) + 2 ln |C1 |. t) = Axt + Bx + Ct + D. ∂t2 ∂x2 1◦ . t) is a solution of this equation. Suppose w(x. . Solution: ∂2w = aeλw ∂2w where C1 . ρ = xet . λ 2 w(x. and C3 are arbitrary constants. .222 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3. w= 2. 2◦ . t) = w(x.2. t) = w(x. k 2 ζ 2 fζζ + k(k + 1)ζfζ − λ 5◦ . a > 0. Polyanin (1996). t+B where the function w(z) is determined by the ordinary differential equation w = w(z). ζ = . X X 3◦ . and β are arbitrary constants. + beβw . λ where k is an arbitrary constant. C1 t + C4 . whose order can be reduced with the transformation ξ = z −2 eλw . λ (t + C1 )k where C1 . and k are arbitrary constants. C4 are arbitrary constants. ∂2w = aeλw ∂2w where C1 .

ξ= x+A . Additive separable solutions: 1 λ 2 w(x. D. ξ = x|t| β . ln(Ax2 + Bx + C) + ln λ λ aA sinh2 (pt + q) 1 −p2 1 . and q are arbitrary constants. 2 2 k1 − ak2 eλu where C1 and C2 are arbitrary constants. t) = λ 1 w(x. t+B © 2004 by Chapman & Hall/CRC . . ∂ ∂w = aeλw . λ 1 p2 2 ln(Ax + Bx + C) + ln . 4◦ . √ 2 ln | A a t + C|. p. and the function u(z) is determined by the autonomous ordinary differential equation 2 2 (k1 − ak2 eλu )uzz = beβu . β where the function U (ξ) is determined by the ordinary differential equation w = U (ξ) − (λ − β)2 2 2 (λ − β)(λ − 2β) + ξUξ + ξ Uξξ = aeλU Uξξ + beβU .3. C4 are arbitrary constants. t) = ln(Ax2 + Bx + C) + ln λ λ aA cosh2 (pt + q) where A. t) = λ 1 w(x. Suppose w(x. t) = w(x. C1 t + C4 . Then the functions λ w1 = w(C1 C2 x + C3 . EQUATIONS WITH EXPONENTIAL NONLINEARITIES 223 2◦ . . 2◦ . a ∂2w where C1 . Solution: z. λ 2 2 2 ln(aA x + Bx + C) − ln(aAt + D). t) is a solution of this equation. where k1 and k2 are arbitrary constants. ∂t2 ∂x ∂x 1◦ . Self-similar solution: w = u(ξ). Traveling-wave solution: w = w(z). Traveling-wave solution: w = u(z). . t) = ln |Ax + B| + Ct + D. C. ) − 2 ln |C2 |. a > 0.2. Expressions (1) to (6) exhaust all solutions that can be representable in the form of the sum of functions with different arguments. are also solutions of the equation. where w = w(z) is deﬁned implicitly by (A and B are arbitrary constants) λµ2 w − aeλw = Az + B. λ aA cos2 (pt + q) a (1) (2) (3) (4) (5) (6) ln |Ax + B| − 1 p2 1 . λ−β 2 ln |t|. B. 3◦ . F (u) = 2b eβu du + C2 . z = k2 x + k1 t. t) = λ w(x. . w(x. z=x a µt. β β2 β2 3. Its solution can be written out in implicit form as a du √ = C1 F (u) 3◦ .

t) = H(ζ) − t. x + C2 2(k − 1) ln(t + C1 ) + f (ζ). and the function u = u(ξ) is determined by the ordinary differential equation (ξ 2 uξ )ξ = (aeλu uξ )ξ . c ¦b 5◦ . ∂x ∂x 1◦ . d d ∂2w = ∂ ∂w where C1 . Lohner. z = k2 x + k1 t. Solution: References: W. where k1 and k2 are arbitrary constants. Polyanin and V. and C3 are arbitrary constants. w(x. η = xet . The substitution Θ(u) = (uz )2 leads to the ﬁrst-order linear equation 2 2 2 (k1 − ak2 eλu )Θu − 2ak2 λeλu Θ = 2beβu. A and B are arbitrary constants. C1 t + C3 ) + 2 ln |C1 |. For other solutions. t) is a solution of this equation. Then the functions 4. and the function f = f (ζ) is determined by the ordinary differential equation 2(k − 1) = a(eλf fζ )ζ . Adams (1981). λ−β 2 ln |t|. 2◦ . see equation 3. Zaitsev (2002). There are exact solutions of the following forms: w= 2 ln |t|. Suppose w(x. one obtains a Riccati equation for ξ = ξ(u). F. R.224 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE Here. ζ = . C2 . ∂t2 aeλw β−λ β w1 = w( C1 x + C2 . treating u in (7) as the independent variable.4. A. which admits the ﬁrst integral ξ 2 − aeλu uξ = C. λ where k is an arbitrary constant. λ 2 w(x. ξ = x|t| β . and the function u(z) is determined by the autonomous ordinary differential equation 2 2 k1 uzz − ak2 (eλu uz )z = beβu . k 2 ζ 2 fζζ + k(k + 1)ζfζ − λ 6◦ . which is considered in the book by Polyanin and Zaitsev (2003).6 with f (w) = aeλw . J. (7) To the special case C = 0 there corresponds a solution of the form (2). Traveling-wave solution: w = u(z). β β2 β2 © 2004 by Chapman & Hall/CRC . λ (t + C1 )k where C1 . η = x + k ln |t|. D. + beβw .4. Cξu = ξ 2 − aeλu . t) = F (η) − 7◦ . β where the function U (ξ) is determined by the ordinary differential equation 3◦ . F. Ames. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). For C ≠ 0. C2 . Solution: w = U (ξ) − (λ − β)2 2 2 (λ − β)(λ − 2β) + ξUξ + ξ Uξξ = (aeλU Uξ )ξ + beβU . and k are arbitrary constants. and E.

w= 6. K. It arises in some areas of physics. 3◦ . Item 2 ◦ . Traveling-wave solutions: bλ(kx + µt + θ0 ) 2 ln tan . OTHER EQUATIONS INVOLVING ARBITRARY PARAMETERS 225 5.1 with f (w) = b sinh(λw). V. t) = e ∂2w =a ∂2w 1◦ .3. and C are arbitrary constants. Rodionov (1999).3. µ.1. Denote: k = 2b Traveling-wave solutions: 1 − k sin z 1 arccosh . ∂t2 ∂x2 The substitution βu = λx + µt + βw leads to an equation of the form 3.4.3. √ bλ 2 1 ln x + C1 x + C2 .2. see equation 3. λ 2 1−z where the functions f = f (t) and g = g(x) are determined by the ﬁrst-order autonomous ordinary differential equations 2 ft = Af 4 + Bf 2 + C. For other exact solutions of this equation. M. 2◦ . arctanh exp λ bλ(µ2 − ak 2 ) e w(x. + b sinh(λw).1 with f (w) = b sinh(λw). R. ∂ ∂w ∂2w =a eλw 2 ∂t ∂x ∂x Functional separable solution: + b – ce–2λw . a > 0. V. Functional separable solution: 1 1+z 4 arctanh z = ln . Andreev.1. w(x. z = 2bβ(1 − k 2 ) (x sinh C1 w= β sin z − k e t cosh C1 + C2 ) if |k| < 1. Zhdanov (1994). It is assumed that bλ(µ2 − ak 2 ) > 0 in both formulas. g ¦f 2 References: A. Other Equations Involving Arbitrary Parameters 3. V. w(x. t) = arctanh f (t)g(x) .3. This is a special case of equation 3. B. .1: ∂2u ∂2u = aeβu 2 .4. cλ t − λ 2a where C1 and C2 are arbitrary constants. Grundland and E. where A. and θ0 are arbitrary constants. Pukhnachov. e w= 2bβ(k 2 − 1) (x sinh C1 e a . V. ∂t2 ∂x ∂2w = aeλx+µt+βw ∂2w 3. t cosh C1 + C2 ) if |k| > 1. t) = where k. a gx = Cg 4 + (B − bλ)g 2 + A. Double sinh-Gordon equation. A.4. 2. Z. Infeld (1992). ∂2w ∂t2 = ∂2w ∂x2 + a sinh(βw) + b sinh(2βw).1. . Kaptsov. and A. Equations with Hyperbolic Nonlinearities 1. λ 2 bλ(µ2 − ak 2 ) bλ(kx + µt + θ0 ) 4 . O. © 2004 by Chapman & Hall/CRC e ξ 2 k+1 arctanh tanh . ξ= β k−1 2 where C1 and C2 are arbitrary constants. ∂t2 ∂x2 Sinh-Gordon equation.

4. 3. ∂x ∂x This is a special case of equation 3.1.10 with f (x) = cxk and g(t) = stn . © 2004 by Chapman & Hall/CRC .6 with f (w) = b sinhk (λw).9 with f (t) = 0.4.1.3.1 with f (w) = bw ln w + kw. ∂x2 This is a special case of equation 3. + k sinh(λw). for k = 1.3. whose general solutions can be represented in implicit form.1.6 with f (w) = a sinh(λw).2.1. 5.1 with f (w) = k sinh(λw) and n = b/a. Multiplicative separable solution: w(x.1. ∂t2 ∂x2 This is a special case of equation 3. ∂t2 ∂x2 This is a special case of equation 3. 4. this equation is reduced to a simpler equation of 3.4.1. ∂2w =a ∂2w 4. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w = + beβt sinhk (λw). for k = 1.1 with f (w) = bw k ln w.2. 6.7 with f (w) = b sinhk (λw).1.1. ∂x ∂x This is a special case of equation 3.226 3. Equations with Logarithmic Nonlinearities 1. t) = ϕ(t)ψ(x). Hence.4. + beβx sinhk (λw). + bwk ln w.1. ∂t2 xn xn .4.1.4. ∂t2 = ∂w ∂ a sinh(λw) .4. aψxx + bψ ln ψ = 0. ∂t2 ∂x2 This is a special case of equation 3. For k = 1.3. 7.4.4. a > 0. ∂2w ∂2w ∂ a cosh(λw) ∂w ∂2w = a ∂ ∂w ∂2w = ∂2w 3. ∂t2 ∂x2 This is a special case of equation 3. a > 0. + bw ln w + cxk + tn )w.4. + bw ln w + kw. ∂t2 + a(x2 – t2 ) lnk (λw).4. where the functions ϕ(t) and ψ(x) are determined by the autonomous ordinary differential equations ϕtt − bϕ ln ϕ − kϕ = 0. ∂t2 ∂x2 This is a special case of equation 3. see also equation 3.4.2 with f (w) = a lnk (λw).6 with f (w) = a cosh(λw).1. Hence. this equation is reduced to a simpler equation of 3. = ∂2w ∂2w ∂2w =a ∂2w h ∂2w =a ∂2w 2. = ∂t2 ∂x ∂x This is a special case of equation 3.

5 with f (w) = b lnk (λw). τ = 1 β exp 2 1 2 βt cosh 1 2 βx leads to a simpler equation of the form 3. n = 0.6 with f (w) = a lnk (λw).2 with β = 0 and f (w) = b lnk (λw). OTHER EQUATIONS INVOLVING ARBITRARY PARAMETERS 227 5. The plus or minus signs in the ﬁrst expression are chosen in any sequence.4.3. This is a special case of equation 3. ∂t2 ∂x2 The transformation w = U (z. + b lnk (λw). waves in ferromagnetic materials. 8. ∂2w ∂t2 =a ∂ ∂x lnk (λw) ∂w ∂x . ∂t2 ∂x2 The transformation w = U (z. Suppose w = ϕ(x. 10. .2. and others). ∂2w ∂2w = + beβt w ln w. C2 . 2 ∂τ ∂z 2 7.3. 2. ∂2w ∂2w = + beβx w ln w.4.2.2.1 with f (w) = cw k ln w and b = an. z= 1 2 exp 1 2 βx cosh 1 2 βt .3. 9. It arises in differential geometry and various areas of physics (superconductivity. τ ). laser pulses in two-phase media. . z = 1 β exp 2 1 2 βt sinh 1 2 βx .3. ∂2w ∂t2 ∂2w = a ∂ xn ∂x ∂2w xn ∂w ∂x + cwk ln w. ∂t2 ∂x2 Sine-Gordon equation. ∂τ 2 ∂z 2 6.4. + b sin(λw). . w1 = λ √ sinh σ w2 = ϕ x cosh σ + t a sinh σ.1: ∂2U ∂2U = + bU ln U . a where C1 . 3. ∂t2 ∂x2 This is a special case of equation 3. t) is a solution of the sine-Gordon equation. ∂2w ∂t2 = axn ∂2w ∂x2 + bxn–1 lnk (λw) ∂w = axn . ∂x This is a special case of equation 3. C2 t).1: ∂2U ∂2U = + bU ln U .3. 1◦ . © 2004 by Chapman & Hall/CRC i i i i i i ∂2w =a ∂2w .3. 1.4. are also solutions of the equation. x √ + t cosh σ . Then the functions 2πn ϕ(C1 x.4.3. τ ). and σ are arbitrary constants. This is a special case of equation 3. dislocations in crystals. Sine›Gordon Equation and Other Equations with Trigonometric Nonlinearities 1.3. τ= 1 2 exp 1 2 βx sinh 1 2 βt leads to a simpler equation of the form 3. .

and V. J. B = k 2 γ 2 > 0. µ. where k. P. ai = where µi and Ci are arbitrary constants. Functional separable solution: w(x. G. t) = µ sinh(kx + A1 ) 4 arctan √ . and C > 0. and B1 are arbitrary constants.2. (3) where k. and C = 0. (2) ∂2 ∂2 − 2 (ln F ) . λ µ ekγ(x+B1 ) + e−kγ(x+B1 ) µ2 = ak 2 γ 2 + bλ > 0. w(x. Zakharov (1984). t) = µ sin(kx + A1 ) 4 . Mij = ai + aj 2 1 − µi . Note some exact solutions that follow from (1) and (2). = −aCg 4 + (aB + bλ)g 2 − aA. and λ = 1) w(x. Traveling-wave solutions: bλ(µ2 − ak 2 ) bλ(kx + µt + θ0 ) π 4 w(x. For A = k 2 > 0. and B1 are arbitrary constants. we have w(x. t) = arccos 1 − 2 F = det Mij .1. b = −1. 4◦ . L. S. B. B = −k 2 < 0. Zakharov (1984). and γ are arbitrary constants. where k. 2 ∂x ∂t zi + zj 2 cosh . Manakov. B. λ k a cosh(µt + B1 ) µ2 = ak 2 + bλ > 0. V. L. S. Formula (3) corresponds to the two-soliton solution of Perring–Skyrme (1962). Novikov. and C are arbitrary constants. r ¦q Reference: R. Novikov. and C > 0. arctan √ λ k a cosh(µt + B1 ) µ2 = bλ − ak 2 > 0. w(x. Manakov. Bullough and P. E. Caudrey (1980). t) = 4 arctan f (x)g(t) . E. B1 .3. For A = 0. λ (1) where the functions f = f (x) and g = g(t) are determined by the ﬁrst-order autonomous separable ordinary differential equations fx gt 2 2 where A. 1 + µi x − µi t + C i . S. 3. B = k 2 > 0. 3. r ¦q Reference: R.228 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 2◦ . S. Lamb (1980). zi = p 1− µ2 i . P. and V. For A = 0. t) = − + arctan exp λ λ bλ(ak 2 − µ2 ) p w(x. V. Steuerwald (1936). and θ0 are arbitrary constants. An N -soliton solution is given by (a = 1. Pitaevskii. 3. Pitaevskii. K. The ﬁrst expression corresponds to a single-soliton solution. L. if bλ(µ2 − ak 2 ) < 0. t) = γ eµ(t+A1 ) + ak 2 e−µ(t+A1 ) 4 arctan . 3◦ . = Af 4 + Bf 2 + C. B. A1 . © 2004 by Chapman & Hall/CRC p p 4 arctan exp λ bλ(kx + µt + θ0 ) if bλ(µ2 − ak 2 ) > 0. A1 . A1 . t) = where k.

3. K. c = − 1 . ∆ = ln 5 + 2 . For a = 1. G. c = 1 . Segur (1973). B. C1 .5: ∂zy w = − 4 a−2 sin w. 2. t) = A + 4 4 arctan B1 eθ + C1 + arctan B2 eθ + C2 . b = 1. It arises in nonlinear optics (propagation of ultrashort pulses in a resonance degenerate medium) and low temperature physics (propagation of spin waves in anisotropic spin liquids). and V. L. For other exact solutions of the original equation. Pitaevskii.4. λ λ if c2 < 4b2 . Manakov. k =µ+ δ is any. 7◦ . b = −1. A.3. Belokolos (1995) obtained a general formula for the solution of the sine-Gordon equation with arbitrary initial and boundary conditions. t) = arctan λ 2b − c 4 bλ(ak 2 − µ2 ) √ √ λ c2 − 4b2 (kx + µt + θ0 ) c2 − 4b2 4 tan w(x. 2. A. Kaup. w(x. S. Manakov. D. µ. B1 . R. Bullough and P.2.3.1. B2 . θ0 is an arbitrary constant. V. λ = 1: 2 √ 5 w(x.1. Ablowitz. k = µ + 3 µ−1 . Pitaevskii. Here. b. V. ∂t2 ∂x2 Double sine-Gordon equation. 2. t) = 2π − δ + 4 arctan √ eθ − √ 15 15 t ¦s ∆ = ln √ 3+2 . ∂2w =a ∂2w 1◦ . b = −1. For a = 1. A.4. © 2004 by Chapman & Hall/CRC u t ¦s References for equation 3. Takhtajan. The transformation z = x − at. Traveling-wave solutions: w(x. J. 15 −1 16 µ . J. J. J. Caudrey (1980). 4 δ is any. 2. M. M. Krichever (1980). k. Calogero and A.3. Note some special cases of interest that arise in applications. y = x + at 1 leads to an equation of the form 3. The sine-Gordon equation is integrated by the inverse scattering method. Caudrey (1980). and θ0 are arbitrary constants. Zakharov. Zakharov (1984). C2 . Clarkson (1991). and H. B. Ablowitz and P. see the book by Novikov. λ = 1: 2 1 4 . and L. t) = 4 arctan eθ−∆ + 4 arctan eθ+∆ . and λ of the original equation. Faddeev (1974). if c2 > 4b2 . and k are related by algebraic constraints with the parameters a. µ. J. F. λ = 1: 2 w(x. Ablowitz and H.1: R. J. t) = 2π + 4 arctan eθ−∆ − 4 arctan eθ+∆ . E. . P. M. and Zakharov (1984). M. c. Segur (1981). k = µ + 4 µ−1 . 6◦ . J.1 with f (w) = b sin(λw). L. Whitham (1974). Traveling-wave solutions: √ √ λ 4b2 − c2 (kx + µt + θ0 ) 4b2 − c2 4 tanh w(x.5. t) = δ − 2π + 4 arctan √ eθ + √ 15 15 2. Item 3◦ . For a = 1. θ = µt kx + θ0 . Weiss (1984). w(x. see equation 3. t) = arctan λ c − 2b 4 bλ(ak 2 − µ2 ) 2◦ . Bullough and P. K. E. c = 1 . 15 −1 16 µ . Steuerwald (1936). For a = 1. It is assumed that bλ(ak 2 − µ2 ) > 0 in both formulas. S. where the parameters A. D. I. k =µ+ References: R. Novikov.1. 1 + b sin(λw) + c sin 2 λw . C. Degasperis (1982). λ = 1: 2 1 4 . b = 1.3. Newell. OTHER EQUATIONS INVOLVING ARBITRARY PARAMETERS 229 5◦ . c = − 1 .

the equation is reduced to a simpler equation of 3. 6. for k = 1. ∂t2 ∂x2 This is a special case of equation 3. Ames (1989).3.4. ∂t2 .7 with f (w) = b sink (λw). the equation reduced to a simpler equation of 3.4. the equation is reduced to a simpler equation of 3. Equations of the Form ∂ w + a ∂w = ∂ f (w) ∂w ∂t2 ∂t ∂x ∂x ∂2w ∂2w = ∂ a cosn (λw) ∂w ∂2w = ∂ a sinn (λw) ∂w 1. a ∂ ∂w = n xn + k sin(λw). C.4. ∂τ 2 ∂τ ∂z References: C.4. ∂t2 ∂x2 This is a special case of equation 3.1. 5. ∂t2 2 3.1.2.3.4. Therefore. the equation is reduced to a simpler equation of 3. for k = 1. © 2004 by Chapman & Hall/CRC w ¦v .4. 7.6 with f (w) = b sink (λw).1 with f (w) = k sin(λw) and b = an. ∂t2 ∂x2 This is a special case of equation 3.230 3. + beβt cosk (λw).1.3. Therefore. 8.4. . for k = 1.3. ∂w ∂ ∂w + =a w–2 . Rogers and W.1.3. ∂x ∂x This is a special case of equation 3.3.4. 10. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w =a + b cos(λw). 9. dz = aw −2 wx dt + (w + wt )dx.7 with f (w) = b cosk (λw). leads to the linear telegraph equation ∂2u ∂ 2 u ∂u + =a 2.4. + beβt sink (λw). Therefore.3.6 with f (w) = a cosn (λw).3.3. Ruggeri (1985). ∂t2 ∂x2 This is a special case of equation 3.6 with f (w) = a sinn (λw).3.3. ∂2w where the subscripts denote the corresponding partial derivatives. Therefore. ∂t2 ∂x2 π leads to an equation of the form 3. ∂t2 ∂t ∂x ∂x The transformation τ = t + ln |w|. ∂x ∂x This is a special case of equation 3.3. ∂t2 x ∂x ∂x This is a special case of equation 3.3. for k = 1.1. ∂t2 ∂x 4.6 with f (w) = b cosk (λw).1: The substitution w = u + 2λ ∂2u ∂ 2u = a 2 − b sin(λu). Rogers and T. ∂2w = ∂2w ∂2w ∂2w = + beβx sink (λw). F. a > 0. ∂2w = ∂2w ∂2w = ∂2w + beβx cosk (λw).4. u = 1/w (dz = zt dt + zx dx). a > 0.1.

t) = f (t)x2 + g(t)x + h(t) − b. g(t). C2 . . gtt + kgt = 6af g. 2◦ . D. utt + kut = n2 This equation is easy to integrate for n = −2 and n = −1. y ¦x 4◦ . OTHER EQUATIONS INVOLVING ARBITRARY PARAMETERS 231 2.3. C4 are arbitrary constants. . htt + kht = 2af h + ag 2 . 2at + C1 e−kt + C2 − b. and the function u = u(t) is determined by the ordinary differential equation 2a(n + 2) n+1 u . t) = (x + C2 )1/(1+n) (C1 e−kt + C2 ) − b. Ibragimov (1994).3. Polyanin and V. and C3 are arbitrary constants. ∂2w ∂t2 +a ∂w ∂t = ∂ ∂x beλw ∂w ∂x . ∂w ∂w ∂ ∂2w +k a(w + b)n . 2◦ . where the functions f (t). C2 . = 2 ∂t ∂t ∂x ∂x w(x. Generalized separable solution for n = 1: w(x. Solution for n = −1: References: N. w(x. and λ are arbitrary constants. © 2004 by Chapman & Hall/CRC . H. Additive separable solution: w(x. is also a solution of the equation. λ where C1 . Zaitsev (2002). its exact solutions are given in the handbook by Polyanin and Zaitsev (2003). 1◦ . λ where C1 . t) = 5◦ . C2 . Solution for n ≠ −1: where C1 and C2 are arbitrary constants. F. . 1◦ . Solution: w(x. t + C 3 − 2 ln |C1 |. t) = (x + C)2/n u(t) − b. t) = 1 ln(C1 x + C2 ) + C3 e−at + C4 . and C3 are arbitrary constants. Suppose w(x. A. k(x + C3 )2 where C1 . 3◦ . . where C is an arbitrary constant. For n = −3/2 and −3. t) is a solution of this equation. and h(t) are determined by the system of ordinary differential equations ftt + kft = 6af 2 . Then the function w1 = w C1 x + C2 . Traveling-wave solution in implicit form: w+b 0 bun − λ2 dw = x + λt + C2 . kλu + C1 where C1 . 3.

. C2 . 1 − a2 x . Equations of the Form ∂ w + f (w) ∂w = ∂ g(w) ∂w ∂t2 ∂t ∂x ∂x Equations of this form admit traveling-wave solutions w = w(kx + λt). a ≠ 0. are also solutions of the equation. Suppose w(x. C1 t + C3 ). and the function u = u(t) is determined by the ordinary differential equation utt + aut = 2bC1 eλu . w(x. Traveling-wave solution in implicit form: beλw − λ2 dw = x + λt + C2 . C2 .232 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3◦ . −1: w= if n ≠ −1. 1). aλw + C1 where C1 . t) is a solution of this equation. 1. t) = 4◦ . ∂2w + wn ∂w = ∂2w where C1 . Polyanin and V. if n = −1. Example 1. Then the functions n n w1 = C1 w( C1 x + C2 . ¦ References: N. Ibragimov (1994). C2 . a stationary solution dependent on x alone. ξ= a(x − at) . Example 2. Additive separable solution: 1 ln(λC1 x2 + C2 x + C3 ) + u(t). w = −2C1 tan(C1 ξ + C2 ). t where the function ϕ = ϕ(ξ) is determined by the ordinary differential equation w = t−1/n ϕ(ξ). where w is the electric ﬁeld strength. F. C2 . 2 3. 1). ∂t2 ∂t ∂x2 1◦ . λ where C1 . where C1 . C2 . 2◦ . and C3 are arbitrary constants. ξ= (1 − ξ 2 )ϕξξ + ϕn − 1 n+1 2(n + 1) ξϕξ + ϕn+1 + ϕ = 0. such equations are encountered in the theory of electric ﬁeld and nonlinear Ohm laws. and a are arbitrary constants (a ≠ 0. Traveling-wave solution for n ≠ 0. to k = 0 there corresponds a homogeneous solution dependent on t alone and to λ = 0. For g(w) = const. and λ are arbitrary constants. Traveling-wave solutions in implicit form: a(x − at) (n + 1)dw =− + C2 wn+1 + C1 1 − a2 a(x − at) dw =− + C2 ln |w| + C1 1 − a2 where C1 . n ξ+C n+1 −1/n . D. Traveling-wave solutions for n = 1: w = 2C1 tanh(C1 ξ + C2 ). n n n2 3◦ . ξ= a(x − at) . Self-similar solution: © 2004 by Chapman & Hall/CRC .3. 1 − a2 where C and a are arbitrary constants (a ≠ 0. 1. and C3 are arbitrary constants. A. and a are arbitrary constants.5. Zaitsev (2002). H.

∂w ∂2w ∂2w + awn =b . D. t) = u(z)t−1/n . C2 . Generalized separable solution for n = 1: w = ϕ(x) t + C1 + C2 dx . Passing to the new independent variables τ = at and z = aβ −1/2 x. Zaitsev (2002). Traveling-wave solutions: C1 1 exp 2 (ax − t) + C2 C1 a −1 1 (ax − t) + C1 .1: ∂w ∂ 2 w ∂2w + wn . F. 2◦ . are also solutions of the equation. Self-similar solution: w(x. Ibragimov (1994. C2 . ∂2w ∂2w . ∂t2 ∂t ∂x2 1◦ . which has a particular solution ϕ = 6(x + C)−2 . where the function u = u(z) is determined by the ordinary differential equation n2 (z 2 − b)uzz + nz(2n + 2 − naun )uz + u(1 + n − naun ) = 0. 2 where the function u = u(z) is determined by the ordinary differential equation 4bn2 uk − (2n − k)2 z 2 uzz + 4bkn2 uk−1 uz Reference: N. C2 . ∂t2 ∂t ∂x2 1◦ . C1 t + C3 ) + ln C1 .3. z = xt(k−2n)/(2n) . t) is a solution of this equation. C1 t + C3 ). t) = − ln where C1 . P. Ibragimov (1994). t) = u(z)t−1/n. 2. + (2n − k)(k − 4 − 4n + 2naun )zuz = 4u(1 + n − anun ). Emech and V. t) = − ln 2 a −1 w(x. H. Then the functions 4. B. Then the functions ∂t2 + awn wk 2n−k 2n 2 w1 = C1 w( C1 x + C2 .3. N.3. z = xt−1 . are also solutions of the equation. w(x. where C1 . Taranov (1972). H. and C3 are arbitrary constants. and C3 are arbitrary constants. Suppose w(x. and a are arbitrary constants (a ≠ 1). = ∂τ 2 ∂τ ∂z 2 3.5. ∂t ∂x ∂x 1◦ . t) is a solution of this equation. we arrive at an equation of the form 3. © 2004 by Chapman & Hall/CRC + ew ∂w = −1 . the function ϕ = ϕ(x) is determined by the autonomous ordinary differential equation ϕ xx = ϕ2 . ϕ2 (x) Here. OTHER EQUATIONS INVOLVING ARBITRARY PARAMETERS 233 4◦ . Polyanin and V. w1 = w( C1 x + C2 . Self-similar solution: w(x. A. ∂2w ∂w =b ∂ ∂w where C1 . 2◦ . ¦ ¦ References: Y. 2◦ . 1995). . . Suppose w(x.

and u = λw. t + C2 ). are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). B. we arrive at an equation of the form 3. w(x. z = xt(λ−2µ)/(2µ) . t) = u(z) − (µ − 2λ)2 z 2 − 4bλ2 eµu uzz − 4bµλ2 eµu (uz )2 + (µ − 2λ)(µ − 4λ + 2aλeλu )zuz + 4λ(1 − aeλu ) = 0. z = xt−1 .4: ∂u ∂ 2 u ∂ 2u + eu .4. N. Ibragimov (1994. w) 2 ∂t ∂x2 1. ¦ . t) is a solution of this equation. C1 t + C3 ) + 2 ln C1 . ¦ 2◦ . ∂t2 ∂t ∂x ∂x 1◦ .234 3◦ . + 2a |t2 − x2 | 1 − C2 x−t 1 x+t x + C(x t) + (x t) ln w(x. ∂t2 ∂x2 Nonlinear Klein–Gordon equation. 1 ln t. w1 = w( x + C1 . Passing to the new variables τ = at. C2 . t) is a solution of the equation in question. ∂t2 ◦ 1 . 1◦ . H. Then the functions where C1 . and C3 are arbitrary constants.1. w2 = w x cosh β + tα1/2 sinh β. t. t) = − ln 2 4 x−t where C and a are arbitrary constants. ∂w ∂ ∂w + aeλw =b eµw . Taranov (1973). H.3. C2 . Equations Involving Arbitrary Functions 2 2 3. λ where the function u = u(z) is determined by the ordinary differential equation w(x. Equations of the Form ∂ w = a ∂ w + f (x. = 2 ∂τ ∂τ ∂z 2 6. Solution: Reference: N.5. t cosh β + xα−1/2 sinh β . ∂2w ∂2w =α + f (w). Solution: 5. 2◦ . 1995). P. Then the functions 2λ−µ 2λ w1 = w( C1 x + C2 . Suppose w(x. Ibragimov (1994). ∂2w where C1 . z = aβ −1/2 x. 3. are also solutions of the equation. Emech and V. ∂2w + aeλw ∂w ∂t =b ∂2w ∂x2 . t) = u(z) − λ(z 2 − b)uzz + λz(2 − aeλu )uz + 1 − aeλu = 0. Solutions: HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE a x+t t + Cx . and β are arbitrary constants. Suppose w = w(x.4. λ where the function u = u(z) is determined by the ordinary differential equation w(x. © 2004 by Chapman & Hall/CRC . 1 ln t. t) = − ln References: Y.

.2. are also solutions of the equation (the plus or minus signs in w 1 are chosen arbitrarily). Infeld (1992). Zaitsev (2002).1. whose general solution can be represented in implicit form as 3 .2. 4 −1/2 where C1 and C2 are arbitrary constants. w(z) = arcsinh sinh k sin f (w) = −b2 2 cosh w cosh k λ2 − 1 bz + c tan w √ . 2◦ . see equations 3. Functional separable solution: w = w(ξ). and λ are arbitrary constants. 3. D. where C1 and C2 are arbitrary constants. α 4◦ .1. λ. (1) λ2 − αk 2 where C1 . f (w) = b sinh(λw). A. λ2 = 1 + b2 cos−2 k. In these cases. k. Then the functions 2. the following relationships between the wave speed. Suppose w = w(x. where β is an arbitrary constant.4. M. 3. τ = xt. © 2004 by Chapman & Hall/CRC . Zhdanov (1994). . w2 = w(x cosh β + t sinh β. w(z) = arcsin sin k sin f (w) = −b2 2w cos cos k λ2 − 1 where k and c are arbitrary constants.3. ∂t2 ∂x2 1◦ .1.1. R. b. t) is a solution of this equation. C2 . respectively. Functional separable solution: w = w(ξ). 3◦ . f (w) = bw ln w. = w1 = w( x. f (w) = beβw . Here.3. + (x2 – t2 )f (w). F (w) = f (w) dw. and the amplitude. ¦ References: A.1.1.1. For solutions of the original equation with some other f = f (w). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 235 2◦ . Grundland and E. see Example 11 in Subsection S. Traveling-wave solution in implicit form: 2 f (w) dw dw = kx + λt + C2 .3. ξ= 1 2 ∂2w ∂2w x2 − t 2 . t). Nesterov (1978) indicated several cases where solution (1) can be written out in explicit form (α = µ = 1): bz + c tanh w √ .3.3. correspond to periodic solutions in z with period 2π: C1 + λ2 = 1 + b2 cosh−2 k. and 3. Polyanin and V. w = w(τ ). x sinh β + t cosh β).1. and f (w) = b sin(λw).3. where the function w = w(ξ) is determined by the ordinary differential equation ξwξξ + wξ + ξf (w) = 0. F. Z. Self-similar solution: ◦ C1 + 2F (w) −1/2 dw = C2 τ.5. the function w = w(τ ) is determined by the autonomous ordinary differential equation wτ τ = f (w). 1 ξ = 1 α(t + C1 )2 − 4 (x + C2 )2 .1. 3. and the function w = w(ξ) is determined by the ordinary differential equation 1 ξwξξ + wξ − f (w) = 0. For exact solutions of the nonlinear Klein–Gordon equation with f (w) = bw m .

Polyanin and V. τ ). 5◦ . + (t2 – x2 )n f (w). Here. ∂τ 2 ∂z 2 © 2004 by Chapman & Hall/CRC . F (w) = f (w) dw. . 5. F. 2 ∂τ ∂z 2 ¦ Reference: A. The transformation w = U (z. D.16 with f (y) = y n and g(z) = z n .4. 2◦ . 3. n = 2. 2 τ = xt leads to a simpler equation of the form 3. Here.1. 3.1. ∂t2 ∂x2 Solution: w = w(ξ). and the function w = w(r) is determined by the autonomous ordinary differential equation 2 2 (C2 − 4C1 )wrr = f (w).4. w). z = 1 (x2 + t2 ). where the function w(ξ) is determined by the ordinary differential equation (a − b2 )wξξ + f (ξ. w). w). + (x2 – t2 )f (xt. Functional separable solution: w = w(z). C2 . Self-similar solution: w = w(τ ). the function w = w(z) is determined by the autonomous ordinary differential equation wzz + f (w) = 0. w) = 0. 6◦ . = ∂2w ∂2w =a + f (x + bt. whose general solution can be represented in implicit form as C1 − 2F (w) −1/2 dw = C2 z. = ∂2w ∂2w τ = xt. Functional separable solution (generalizes the solution of Items 3 ◦ and 4◦ ): w = w(r).236 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 4◦ . . 1 z = 2 (x2 + t2 ). and C3 are arbitrary constants. the function w = w(τ ) is determined by the ordinary differential equation wτ τ = f (τ . ∂t2 ∂x2 1◦ . ∂t2 ∂x2 This is a special case of equation 3. r = C1 x2 + C2 xt + C1 t2 + C3 . . 4. where C1 and C2 are arbitrary constants. w). The transformation leads to the simpler equation 1 z = 2 (x2 + t2 ). τ = xt ∂2w ∂2w = + f (τ . Zaitsev (2002). ∂2w ∂2w ξ = x + bt. where C1 .1: ∂2U ∂2U = + f (U ).

4. τ ). ∂t2 ∂x2 The transformation 7. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 237 6. ∂t2 ∂x2 1◦ . ∂2w ∂2w = + eβx f (w). 2◦ . the transformation ∂2w = ∂2w ξ = ax + bt.1. 9.14 with f (z) = e−az . w = U (z. and C is an arbitrary constant. leads to an equation of the form 3. this equation admits a traveling-wave solution U = U (kz + λτ ) and a solution of the form U = U (z 2 − τ 2 ).1: ∂2U ∂2U = + 4β −2 f (U ). see equation 3.4.1. There is a solution of the form w = w(ax + bt).4.3. + eax+bt f (w). where the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations ϕtt − b ln ϕ + f (t) + C ϕ = 0. τ ). © 2004 by Chapman & Hall/CRC . 2 ∂τ ∂z 2 For arbitrary f = f (U ). τ = exp 1 2 βx sinh 1 2 βt leads to a simpler equation of the form 3. see equation 3.1. and for b = −a.4. ∂t2 ∂x2 The transformation w = U (z. z = exp 1 2 βx cosh 1 2 βt .4. Zaitsev (2002).13 with f (z) = eaz . this equation admits a traveling-wave solution U = U (kz + λτ ) and a solution of the form U = U (z 2 − τ 2 ). ¦ Reference: A. z = exp 1 2 βt sinh 1 2 βx .4. For b ≠ a. t) = ϕ(t)ψ(x). D.1: ∂2U ∂2U = + 4β −2 f (U ). ∂τ 2 ∂z 2 For arbitrary f = f (U ). Polyanin and V. ∂2w ∂2w =a + bw ln w + f (t)w. ∂t2 ∂x2 Multiplicative separable solution: w(x. ∂τ 2 ∂ξ 2 a −b 3◦ .1. aψxx + b ln ψ − C ψ = 0. τ = exp 1 2 βt cosh 1 2 βx leads to a simpler equation of the form 3.1. ∂2w = ∂2w + eβt f (w). For b = a. 8.6: τ = bx + at 1 ∂ 2w ∂ 2 w = + 2 2 eξ f (w). F.

∂t2 ∂x2 Multiplicative separable solution: w(x. + f (t)w ln w + bf (t)x + g(t) w. Polyanin and V. D. a τ= 1 1 (t + x) + 2 2 t−x f (σ) dσ a where a is an arbitrary constant. aψxx + b ln ψ + f (x) − C ψ = 0. t) = ϕ(t)ψ(x). t) = e−bt ϕ(x). ∂t2 ∂x2 The transformation 14.1. ∂τ 2 ∂z 2 e ¦d Reference: A. τ ). and C is an arbitrary constant. 12. 11. aϕxx + f (x)ϕ ln ϕ + g(x) − b2 ϕ = 0. leads to an equation of the form 3.1: ∂2U ∂2U = + g(U ).1: ∂2U ∂2U = + g(U ). ∂t2 ∂x2 Multiplicative separable solution: w(x. 13.4. 2 τ= 1 2 t+x a f (λ) dλ + 1 (t − x) 2 where a is an arbitrary constant.238 10.4. where the function ϕ(t) is determined by the ordinary differential equation ϕtt = f (t)ϕ ln ϕ + g(t) + ab2 ϕ. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂2w =a + bw ln w + f (x) + g(t) w. Zaitsev (2002). t) = e−bx ϕ(t). ∂2w ∂2w =a + f (x)w ln w + bf (x)t + g(x) w.1. w = U (z. leads to an equation of the form 3. ∂2w =a ∂2w where the function ϕ(x) is determined by the ordinary differential equation ∂t2 ∂x2 The transformation w = U (ξ. ∂t2 ∂x2 Multiplicative separable solution: w(x. z= 1 1 (t + x) − 2 2 t−x f (σ) dσ. F. ∂2w = ∂2w + f (t + x)g(w). 2 ∂τ ∂z 2 © 2004 by Chapman & Hall/CRC . z= 1 2 t+x a f (λ) dλ − 1 (t − x). τ ). ∂2w = ∂2w + f (t – x)g(w). where the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations ϕtt − b ln ϕ + g(t) + C ϕ = 0.

Equations of the Form ∂ w = a ∂ w + f x. respectively. where the function w(ξ) is determined by the ordinary differential equation wξξ + 2. ∂2w ∂2w = + f (t + x)g(t – x)eβw .1. ∂2w ∂t2 =a ∂2w ∂x2 + f (x) ∂w 1 a+b w = f (w). b τ= 1 2 t+x a f (λ) dλ + 1 2 t−x g(σ) dσ. τ ). Functional separable solution: w = w(ξ).1. ∂2w ∂t2 ∂x2 The transformation w = U (z. aξ ξ ak ∂x Multiplicative separable solution: + bw ln w + g(x) + h(t) w. leads to an equation of the form 3. 1 2 t+x a f (λ) dλ − 1 2 t−x g(σ) dσ. and C is an arbitrary constant. 2 2 3. b where a and b are arbitrary constants. a ∂2w =a ∂2w + b ∂w To m = 1 and m = 2 there correspond nonlinear waves with axial and central symmetry. m= b . z= 1 2 t+x a f (λ) dλ − 1 2 t−x g(σ) dσ.3.1: ∂2U ∂2U = + h(U ). This equation can be rewritten in the form ∂t2 ∂x2 a ∂ ∂w ∂2w = m xm ∂t2 x ∂x ∂x + f (w). z= = ∂2w + f (t + x)g(t – x)h(w). ∂τ 2 ∂z 2 16. t. D.1: ∂2U ∂ 2U = + eβU .3. + f (w). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 239 15.4. ∂t2 ∂x2 The transformation w = U (z. w. τ ). ϕtt − b ln ϕ + h(t) + C ϕ = 0. © 2004 by Chapman & Hall/CRC . ∂w 2 ∂t ∂x2 ∂x 1.4.2. This is a special case of equation 3. ξ= ak(t + C)2 − kx2 . b τ= 1 2 t+x a f (λ) dλ + 1 2 t−x g(σ) dσ. w(x. x ∂x 1◦ . 2 ∂τ ∂z 2 g ¦f Reference: A. b where a and b are arbitrary constants.4 with n = 0. where the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations aψxx + f (x)ψx + b ln ψ + g(x) − C ψ = 0. leads to an equation of the form 3.4. Polyanin and V. F. t) = ϕ(t)ψ(x). 2◦ . Zaitsev (2002).4.2.

ξ = x + λt. Generalized separable solution quadratic in x: w(x. and λ are arbitrary constants. t) = ϕ(t)x2 + ψ(t)x + χ(t). t) + C2 cos(kt) + C3 sin(kt) 2◦ . 4b 2b (4) 3◦ . ψ(t). Particular solution of equation (2): Θ=− 1 c (ξ + C3 )2 + (a − λ2 ). 4. and χ(t) are determined by the system of ordinary differential equations ϕtt = 4bϕ2 + cϕ. where λ is an arbitrary constant. Another particular solution to equation (5) is given by ϕ = − 4 c/b. (3) if c = −k < 0. are also solutions of the equation. ψtt = (4bϕ + c)ψ. © 2004 by Chapman & Hall/CRC . Suppose w(x. ∂t2 ∂x2 ∂x 1◦ . and C3 are arbitrary constants. where the functions ϕ(t). f (τ ) sinh k(t − τ ) dτ if c = k 2 > 0. 2 f (τ ) sin k(t − τ ) dτ where C1 and C2 are arbitrary constants. t) = U (t) + Θ(ξ).240 3. F (w) + C1 F (w) = f (w) dw. C2 . where C1 . t) + C2 cosh(kt) + C3 sinh(kt) if c = k 2 > 0. (a − λ )Θξξ + b Θξ The solution of equation (1) is given by U (t) = C1 cosh(kt) + C2 sinh(kt) + 1 U (t) = C1 cos(kt) + C2 sin(kt) + k 1 k t 0 t 0 2 2 (1) (2) + cΘ = 0. χtt = cχ + bψ 2 + 2aϕ + f (t). (5) (6) (7) Equation (5) has the trivial particular solution ϕ(t) ≡ 0. which leads to ﬁrst-order linear equation. t) is a solution of this equation. ∂2w ∂w 2 ∂2w =a +b + cw + f (t). to which there corresponds a solution 1 of (4) linear in x. 2 2 ∂t ∂x ∂x Traveling-wave solution in implicit form: (λ2 − a) dw = x + λt + C2 . To the stationary solution there correspond λ = 0. C2 . 2 Equation (2) can be solved with the change of variable z(Θ) = Θξ . Then the functions w1 = w( x + C1 . Solutions: h h if c = −k 2 < 0. and the functions U (t) and Θ(ξ) are determined by the ordinary differential equations Utt − cU − f (t) = 0. where C1 . h w(x. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w ∂2w =a + f (w) . w2 = w( x + C1 .

t) = ϕ(t) + ψ(t) exp(λx). the general solution to (6) is expressed as ψ(t) = C3 ϕ(t) + C4 ϕ(t) ¯ ¯ where C3 and C4 are arbitrary constants. t) is a solution of this equation.3. Zaitsev (2002). w(x. Equation (2) is linear in ψ and. + cw ∂w ∂x + kw2 + f (t)w + g(t). the general solution to (6) is expressed in terms of exponentials or sine and cosine. leads to the simpler equation 2 + cz. F. t + C1 ) + C2 cos(kt) + C3 sin(kt) 2◦ . The substitution w = z(x. and the function ϕ(x) is determined by the ordinary differential equation 2 aϕxx + b ϕx + cϕ + f (x) = 0. with ϕ = const. where C1 and C2 are arbitrary constants. respectively. Hence. Note that equation (6) has a particular solution ψ = ϕ(t).4. w1 = w(x. Reference: A. © 2004 by Chapman & Hall/CRC . where C1 . (2) In the special case f (t) = const. Suppose w(x. ∂t2 ∂x2 ∂x 1◦ . where ϕ(t) is any nontrivial particular ¯ ¯ solution to (5). Then the functions 5. w(x. are also solutions of the equation. t) = ϕ(x) + ψ(t). which are linear in ψ and χ. Additive separable solution: Here. Polyanin and V. C2 . t) + U (t). The functions ψ = ψ(t) and χ = χ(t) can be found by successively integrating equations (6) and (7). ψ(t) = ∂2w ∂t2 =a ∂2w ∂x2 +b ∂w 2 ∂x Generalized separable solution: 6. w2 = w(x. hence. due to its autonomity. (1) 2 ψtt = (cλ + 2k)ϕ + f (t) + aλ ψ. C1 cosh(kt) + C2 sinh(kt) if c = k 2 > 0. where λ is a root of the quadratic equation bλ2 + cλ + k = 0. ϕ2 (t) ¯ where the function U (t) is deﬁned by formula (3). 4◦ . ∂2w ∂w 2 ∂2w =a +b + cw + f (x). and C3 are arbitrary constants. where C1 and C2 are arbitrary constants. D. equation (1) has exact solutions of the form ϕ = const and. can be integrated by quadrature. C1 cos(kt) + C2 sin(kt) if c = −k 2 < 0. and the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕtt = kϕ2 + f (t)ϕ + g(t). i i if c = −k 2 < 0. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 241 The general solution of the autonomous equation (5) can be represented in implicit form: i 3 8 3 bϕ + cϕ2 + C1 −1/2 dϕ = C2 t. t + C1 ) + C2 cosh(kt) + C3 sinh(kt) if c = k 2 > 0. g(t) = const. ∂2z ∂z ∂2z =a 2 +b ∂t2 ∂x ∂x k ¦j dt .

which are linear in ψ and χ. ∂2w =a ∂2w HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE + f (t) ∂w 2 + g(t)w + h(t). The solution of the ﬁrst equation for ϕ(t) is expressed as ϕ(t) = C1 cosh(kt) + C2 sinh(kt) + ϕ(t) = C1 cos(kt) + C2 sin(kt) + where C1 and C2 are arbitrary constants. Here. t) = ϕ(t) + ψ(x). and h are not speciﬁed) ϕtt = 4f ϕ2 + gϕ. Here. the functions ψ = ψ(t) and χ = χ(t) can be obtained by successively solving equations (3) and (4). and χ(t) are determined by the system of ordinary differential equations of the second order with variable coefﬁcients (the arguments of f . g. B. ∂2w ∂w ∂2w =a + f (x) 2 2 ∂t ∂x ∂x Additive separable solution: 2 + bw + g(x) + h(t). t) = 1 At2 + Bt + C + 2 t 0 (t − τ )h(τ ) dτ + ϕ(x). If the functions f and g are proportional to each other. If a solution ϕ = ϕ(t) of the nonlinear equation (2) has been found. w(x. respectively. 8. where ϕ(t) is any nontrivial particular ¯ ¯ solution to (2). w(x. to which there corresponds a solution of (1) linear in x. and C are arbitrary constants. the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations aψxx + f (x)(ψx )2 + bψ + g(x) = 0. ψtt = (4f ϕ + g)ψ. (2) (3) (4) χtt = gχ + f ψ 2 + h + 2aϕ. A. t) = ϕ(t)x2 + ψ(t)x + χ(t).242 7. 9. ψ(t). Note that equation (3) has a particular solution ψ = ϕ(t). Equation (2) has the trivial particular solution ϕ(t) ≡ 0. ϕ2 (t) ¯ where C1 and C2 are arbitrary constants. ϕtt − bϕ − h(t) = 0. ∂t2 ∂x2 ∂x Generalized separable solution quadratic in x: w(x. Hence. 1 k h(τ ) sin k(t − τ ) dτ . 1 then a particular solution to equation (2) is given by ϕ = − 4 g/f = const. (1) where the functions ϕ(t). ∂2w ∂t2 ∂x2 Additive separable solution: =a ∂2w + f (x) ∂w ∂x 2 + g(x) + h(t). the general solution to (3) is expressed as ψ(t) = C1 ϕ(t) + C2 ϕ(t) ¯ ¯ dt . © 2004 by Chapman & Hall/CRC 1 k t 0 t 0 h(τ ) sinh k(t − τ ) dτ if b = k 2 > 0. if b = −k 2 < 0. and the function ϕ(x) is determined by the ordinary differential equation aϕxx + f (x) ϕx )2 + g(x) − A = 0.

(9) where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations of the second order with variable coefﬁcients ϕtt = bf ϕ2 + ψ 2 + gϕ + h. (α. 2. Generalized separable solutions: 10. g = αf .4. the equation is autonomous and. 2 particular solutions to equation (2) are expressed as (4) where k1 and k2 are roots of the quadratic equation bk + αk + β = 0. equation (3) can be rewritten in the form ψtt = (2bkn + α)f − ab ψ. If the functions f . 3◦ . and h are not speciﬁed) ϕtt = bf ϕ2 + gϕ + h. ψtt = (2bf ϕ + g − ab)ψ. ϕ = k2 . g. B = −2. t) = ϕ(t) + ψ(t) exp x −b . t) = ϕ(t) + ψ(t) A exp x −b + B exp −x −b . h = βf ϕ = k1 . Polyanin (1996). g. b > 0. (7) (8) We express ϕ from (8) in terms of ψ and then substitute into (7) to obtain a nonlinear fourth-order equation for ψ. ψtt = 2bf ϕ + g − ab ψ. hence. and h = const was considered). n ¦m (10) (11) References: V. b < 0. the case of f = a. g. n = 1. t) = ϕ(t) + ψ(t) cos x b + c . its order can be reduced. Generalized separable solution (generalizes the solutions of Item 1 ◦ ): √ √ w(x. In this case. g = const. Generalized separable solution (c is an arbitrary constant): √ w(x. the general solution of equation (5) is the sum of exponentials (or sine and cosine).3. A. l b < 0. Galaktionov (1995. F. the function ψ = ψ(t) can be obtained by solving equation (3) linear in ψ. t) = ϕ(t) + ψ(t) sinh x −b . with f . and h are proportional to each other. t) = ϕ(t) + ψ(t) cosh x −b . (1) where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations of the second order with variable coefﬁcients (the arguments of f . B = 2. w(x. D. ψtt = 2bf ϕ + g − ab ψ. © 2004 by Chapman & Hall/CRC . (5) Kamke (1977) and Polyanin and Zaitsev (2003) present many exact solutions of the linear equation (5) for various f = f (t). √ 1 1 A = 2. Zaitsev and A. √ w(x. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 243 ∂t2 ∂x2 ∂x ◦ 1 . In the special case f = const. 2◦ . w(x. Note two special cases of solutions (6) that can be expressed in terms of hyperbolic functions: √ 1 1 A = 2. ∂2w =a ∂2w + f (t) ∂w 2 + bf (t)w2 + g(t)w + h(t). (2) (3) If a solution ϕ = ϕ(t) to equation (2) has been found. h = const. β = const). V. (6) where the functions ϕ(t) and ψ(t) are determined by the following system of second-order ordinary differential equations with variable coefﬁcients ϕtt = bf ϕ2 + 4ABψ 2 + gϕ + h.

14. 13. Here. if b = −k 2 < 0. t) = ϕ(t) + ψ(x). B. ∂w ∂x + bw + g(t). w(x. The solution of the ﬁrst equation is expressed as ϕ(t) = C1 cosh(kt) + C2 sinh(kt) + ϕ(t) = C1 cos(kt) + C2 sin(kt) + where C1 and C2 are arbitrary constants. if b = 0. 1 k g(τ ) sin k(t − τ ) dτ . ∂2w ∂t2 =a ∂2w ∂x2 + f x. aψxx + f x. t) = 1 At2 + Bt + C + 2 t 0 (t − τ )g(τ ) dτ + ϕ(x). t) = ϕ(x) + ψ(t). ∂2w ∂t2 =a ∂2w ∂x2 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE + f (t) ∂w 2 ∂x ∂x There is a generalized separable solution quadratic in x: + g1 (t)x + g0 (t) ∂w + h(t)w + p2 (t)x2 + p1 (t)x + p0 (t). and the function ϕ(x) is determined by the ordinary differential equation aϕxx + f x. w(x. ∂2w ∂t2 =a ∂2w ∂x2 + f x. and the function ϕ(x) is determined by the ordinary differential equation k aϕxx + f (x) ϕx + g(x)ϕx + bϕ + h2 (x) = 0. ϕtt − bϕ − g(t) = 0. A. Additive separable solution: w(x.244 11. if b = −k 2 < 0. 12. © 2004 by Chapman & Hall/CRC Additive separable solution: Here. ϕx − A = 0. t) = ϕ(t)x2 + ψ(t)x + χ(t). ∂2w ∂t2 =a ∂2w ∂x2 + f (x) ∂w ∂x k + g(x) ∂w ∂x + bw + h1 (t) + h2 (x). where C1 and C2 are arbitrary constants. ψx + bψ = 0. and C are arbitrary constants. the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations 1 k t 0 t 0 g(τ ) sinh k(t − τ ) dτ if b = k 2 > 0. Additive separable solution: Here. t C cosh(kt) + C sinh(kt) + 1 1 sinh k(t − τ ) h1 (τ ) dτ 2 k 0 t 1 ψ(t) = C1 cos(kt) + C2 sin(kt) + sin k(t − τ ) h1 (τ ) dτ k 0 t C +C t+ 1 (t − τ )h1 (τ ) dτ 2 0 if b = k 2 > 0. w(x. ∂w ∂x + g(t).

t. Solution for n = 2: w = w(y). F (w) = f (w) dw.1. and the function w = w(y) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wyy − aBwy + f (w) = 0. ∂t2 ∂z 2 2.3. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 245 ∂2w 1 ∂w ∂2w =a + wf t. t) = eλx ϕ(t). the general solution of equation (1) is written out in implicit form as o C1 + 8 F (w) k −1/2 dw = r + C2 . λ) ϕ. The substitution y = x + β leads to a special case of equation 3. ∂x2 ∂x dx √ leads to an equation of the form 3. ∂w 2 ∂t ∂x2 ∂x 1. 15. (2) kn2 The book by Polyanin and Zaitsev (2003) presents a number of exact solutions to equation (2) for various f = f (w).1. where A and B are arbitrary constants. 4 a(2 − n)2 ∂2w = a(x + β)n ∂2w where k and the expression in square brackets must have like signs. wξξ = Special case.1. where λ is an arbitrary constant. ∂t2 1◦ . ∂x2 This equation describes the propagation of nonlinear waves in an inhomogeneous medium. r2 = k (x + β)2−n 1 (t + C)2 − .4. + f (w). where C1 and C2 are arbitrary constants. For n = 0 see equation 3.4 with b = 0.4. © 2004 by Chapman & Hall/CRC (3) . 3◦ . a > 0. and w(r) is determined by the ordinary differential equation wrr + n 4 2(1 − n) 1 w = f (w). 2−n r r k (1) The substitution ξ = r 2−n leads to the generalized Emden–Fowler equation 4(2 − n)2 4(1−n) ξ n f (w). 2 2 3.1: ax2 + b ∂2w ∂2w = + f (w). y = At + B ln |x + β|. . ∂2w ∂t2 = (ax2 + b) ∂2w The substitution z = ∂w + ax + f (w).4.4.3. 2◦ . Equations of the Form ∂ w = f (x) ∂ w + g x.4. and the function ϕ(t) is determined by the second-order linear ordinary differential equation ϕtt = aλ2 + f (t. For n = 1. Functional separable solution for n ≠ 2: w = w(r). 2 2 ∂t ∂x w ∂x Multiplicative separable solution: w(x.3. w.

F. √ Solution of equation (2) with A = B a in implicit form: aB where C is an arbitrary constant. Functional separable solution for n ≠ 2: w = w(r). f (w) Reference: V. kn2 (1) The book by Polyanin and Zaitsev (2003) presents a number of exact solutions to equation (1) for various f = f (w). r ¦q (2) dw = −z + C. Zaitsev and A. ∂t2 ∂x2 w = w(ξ). where A and B are arbitrary constants. D. 2◦ . 1 ξ = 4 a(2 − n)2 (t + C)2 − x2−n . 1◦ . and the function w = w(z) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wzz + aBwz + f (w) = 0.246 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE √ Solution of equation (3) with A = B a in implicit form: p p aB where C is an arbitrary constant. 2a(2 − n) B= 1 . ∂ ∂w ∂2w = a(x + β)n + f (w). Polyanin (1996). 2 ∂t ∂x ∂x This equation describes the propagation of nonlinear waves in an inhomogeneous medium. and the function w(r) is determined by the ordinary differential equation wrr + n 4 2 1 w = f (w). + f (w). C is an arbitrary constant. Zaitsev and A. ∂2w = axn ∂2w + bxn–1 ∂w a > 0. ∂x 1◦ . F. f (w) Reference: V. Functional separable solution for n ≠ 2: 4. and the function w = w(ξ) is determined by the ordinary differential equation ξwξξ + Awξ − Bf (w) = 0. Solution for n = 2: w = w(z). a > 0. z = At + B ln |x + β|. (1) where A= a(4 − 3n) + 2b . (t + C)2 − 4 a(2 − n)2 where k and the expression in square brackets must have like signs. r ¦q dw = y + C. D. Polyanin (1996). 2−n r r k The substitution ξ = r n−2 leads to the generalized Emden–Fowler equation wξξ = 4(2 − n)2 − 4 ξ n f (w). 3. a(2 − n)2 © 2004 by Chapman & Hall/CRC . r2 = k (x + β)2−n 1 . Here.

(2) wzz − 2 (1 − A) In the special case A = 1 . are also solutions of the equation.3. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 1 247 For A ≠ 1. and the function w(y) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wyy + (b − a)Bwy + f (w) = 0. and the function w(z) is determined by the ordinary differential equation 2 a(1 − n) + f (w) wz = 0. © 2004 by Chapman & Hall/CRC s F (w) = f (w) dw. B 2 (a − b)2 whose exact solutions for various f = f (w) can be found in Polyanin and Zaitsev (2003). z = a(2 − n)2 (t + C)2 − 4x2−n 1/2 where C is an arbitrary constant. anw − 2F (w) + C1 a(2 − n) where C1 and C2 are arbitrary constants. B. Suppose w(x. t) is a solution of this equation. y = At + B ln |x| + C. 2◦ . where C1 and C2 are arbitrary constants. uuw − u = 5. the substitution ξ = kz 1−A (k = 1) brings (1) to the generalized Emden–Fowler equation 2A−1 kB z 1−A f (w) = 0. w = w(y). (3) s F (w) = f (w) dw. the integration of which yields the general solution in implicit form: 1 dw = ln z + C2 . Solution for n = 2: C1 + s 2 F (w) A2 − aB 2 −1/2 dw = y + C2 . 2 2 ∂t ∂x ∂x 1◦ . the substitution u(w) = B(a − b) y A2 − aB 2 f (w). Functional separable solution for n ≠ 2: w = w(z). (b − a)B f (w) Solutions of equation (3) with b = a: s 2◦ . where A. . √ Solutions of equation (3) with A = B a in implicit form: dw = −y + C1 . the general solution of equation (2) 2 is expressed as s C1 + 8kBF (w) −1/2 dw = z + C2 . F (w) = f (w) dw. . which corresponds to b = a(n − 1). The books by Polyanin and Zaitsev (1995. s where C1 and C2 are arbitrary constants. Then the functions 2−n 2 w1 = w C1 x.4. √ aB 2 − A2 w brings (3) to the Abel equation For A ≠ B a and b ≠ a. ∂2w ∂w ∂2w = axn + xn–1 f (w) . (1) wzz + a(2 − n)z The substitution u(w) = zwz brings (1) to a separable ﬁrst-order equation. C1 t + C2 . and C are arbitrary constants. 2003) present a number of exact solutions to equation (2) for some f = f (w).

Functional separable solution for n ≠ 2: 6. Zaitsev and A. a > 0.3. where A. ∂2w ∂w ∂2w = axn + xn–1 f (w) + g(w). 2 2 ∂t ∂x ∂x 1◦ . ∂t2 ∂x2 This is a special case of equation 3. (n − 2)2 (n − 2)2 w = w(y).9 with b = 0. Reference: V. and C are arbitrary constants. and the function w(ξ) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wξξ + B f (w) − a wξ + g(w) = 0. F.9 with b = aλ. 2 where the function w = w(ξ) is determined by the ordinary differential equation aξ n−1 − 4◦ . and the function w(y) is determined by the autonomous ordinary differential equation (aB 2 − A2 )wyy + B f (w) − a wy = 0. y = At + B ln |x| + C. and the function w(z) is determined by the ordinary differential equation 1 1 2 a(1 − n) + f (w) wz − g(w) = 0. wzz + a(2 − n) z ak(2 − n)2 2◦ . k = 1. t w = w(z). Polyanin (1996). where A. z = ka(2 − n)2 (t + C)2 − 4kx2−n ]1/2 . ξ = At + B ln |x| + C. D. Solution for n = 2: 2(n − 4) 4 ξ wξξ + ξ n−2 f (w) + wξ = 0. √ whose solution with A ≠ B a is given by aB 2 − A2 B v ¦u t dw = −y. where C is an arbitrary constant. the change of variable u(w) = wξ brings (1) to an Abel equation. ∂2w ∂t2 = ∂ aeλx ∂w + f (w). and C are arbitrary constants.248 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3◦ . a > 0. ∂x ∂x This is a special case of equation 3. √ Solution of equation (1) with A = B a: B t (1) f (w) − a dw = −ξ + C1 .3. ∂2w ∂2w = aeλx + f (w). B. © 2004 by Chapman & Hall/CRC . 7. B. Solution for n = 2: w = w(ξ). 8. g(w) In the general case.4. F (w) − aw + C1 F (w) = f (w) dw.4. ξ = x|t| n−2 . Self-similar solution for n ≠ 2: w = w(ξ). whose exact solutions for various f = f (w) and g = g(w) can be found in Polyanin and Zaitsev (2003).

(1) wzz + aλ z akλ2 For b = aλ. 2 k(2b − aλ) The books by Polyanin and Zaitsev (1995. where C1 and C2 are arbitrary constants. © 2004 by Chapman & Hall/CRC w z = 4ke−λx − akλ2 (t + C)2 . see equation 3. 2 2 ∂t ∂x ∂x Functional separable solution: 9. a > 0. 2◦ . where C is an arbitrary constant. the substitution ξ = z 2 wξξ + 2b−aλ aλ brings (1) to the generalized Emden–Fowler equation (2) 4(aλ−b) a ξ 2b−aλ f (w) = 0. For λ = 0. F (w) = f (w) dw.2. z = 4e−λx − aλ2 (t + C)2 1/2 w ∂2w = aeλx ∂2w + eλx f (w) ∂w . and the function w(z) is determined by the ordinary differential equation 1 2 1− f (w) wz = 0.3. where C1 and C2 are arbitrary constants. Suppose w(x. the integration of which yields the general solution in implicit form: 1 dw = ln z + C2 . 1/2 where C is an arbitrary constant and the function w = w(z) is determined by the ordinary differential equation 1 2(aλ − b) 1 wz + f (w) = 0. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 249 ∂2w ∂w ∂2w = aeλx + beλx + f (w).4. k = 1. ξ = t2 eλx . 2F (w) − aλw + C1 aλ where C1 and C2 are arbitrary constants. are also solutions of the equation. 10. w = w(z). t) is a solution of this equation. Then the functions ∂t2 ∂x2 w1 = w x + 2C1 . eλC1 t + C2 . For b ≠ 1 aλ. Generalized self-similar solution: w = w(ξ). 2003) present a number of exact solutions to equation (2) for some f = f (w). (1) wzz + z aλ The substitution u(w) = zwz brings (1) to a separable ﬁrst-order equation.4. ∂x 1◦ . . Zaitsev and A. y ¦x Reference: V. . 3◦ . F (w) = f (w) dw. Functional separable solution for λ ≠ 0: w = w(z). F. where the function w = w(z) is determined by the ordinary differential equation (aλ2 ξ 2 − 4ξ)wξξ + λξf (w) + aλ2 ξ − 2 wξ = 0. Polyanin (1996).3. D. the solution of equation (1) is expressed as w C1 − 2 F (w) akλ2 −1/2 dw = z + C2 .

∂t2 ∂x2 1◦ . ψtt = 2[2f (t) + a]ϕψ + g(t)ψ + h1 (t). 12. | ¦{ Reference: V. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂2w ∂w ∂2w = aeλx + eλx f (w) + g(w). 2 w1 = C1 w x. z ∂2w = aw4 ∂2w where C1 and C2 are arbitrary constants. Suppose w(x.4. ∂x Multiplicative separable solution: where the functions ϕ(x) and ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) f (x)ϕxx + g(x)ϕx + aϕ ln ϕ + C + h(x) ϕ = 0. ψ = ψ(t). χ = χ(t) are determined by the system of ordinary differential equations ϕtt = 2[2f (t) + a]ϕ2 + g(t)ϕ + h2 (t). z = 4ke−λx − akλ2 (t + C)2 1/2 . and the function w(z) is determined by the ordinary differential equation 1 1 2 1− f (w) wz + g(w) = 0. Galaktionov (1995). A. Suppose u = u(x) is any nontrivial solution of the second-order linear ordinary differential equation auxx + f (x)u = 0. k = 1. w(x. ψtt − aψ ln ψ + C − p(t) ψ = 0. (1) The transformation ξ= dx . C1 t + C2 . t. + f (x)w5 . Functional separable solution for λ ≠ 0: z w = w(z). wzz + z aλ akλ2 2◦ . 2 2 3. where C is an arbitrary constant.4. u2 z= w u © 2004 by Chapman & Hall/CRC . ∂t2 ∂x2 ∂x Generalized separable solution quadratic in x: w(x. t) = ϕ(t)x2 + ψ(t)x + χ(t). ∂w 2 ∂x ∂t ∂x2 1. 2◦ . are also solutions of the equation. For λ = 0. t) is a solution of this equation.250 11. χtt = 2aϕχ + f (t)ψ 2 + g(t)χ + h0 (t). ∂2w ∂w 2 ∂2w = aw + f (t) + g(t)w + h2 (t)x2 + h1 (t)x + h0 (t). t) = ϕ(x)ψ(t). where the functions ϕ = ϕ(t). w. there is a traveling-wave solution: w = w(αx + βt). Then the functions 2. ∂2w ∂t2 = f (x) ∂2w ∂x2 + g(x) ∂w + aw ln w + h(x) + p(t) w. Equations of the Form ∂ w = f (w) ∂ w + g x. 2 2 ∂t ∂x ∂x 1◦ . the case of f = const and h1 = h2 = 0 was treated.

∂t2 ∂x ∂x 1◦ . C1 and C2 are arbitrary constants. the solutions of equation (1) are expressed in terms of Bessel functions. Zaitsev and A. ∂2w where C1 and C2 are arbitrary constants. C. (2) Given a particular solution u = u(x) of the linear equation (1). © 2004 by Chapman & Hall/CRC } . 2◦ . ∂ ∂w =a w–4/3 + f (x)w–1/3 . −2 3 w1 = C1 w x. 4A2 a The ﬁrst solution is degenerate and the second one is a special case of a multiplicative separable solution z(ξ. and λ are arbitrary constants): z(ξ. Multiplicative separable solution: w(x. and a self-similar solution of the form z = tk ϕ(ζ). C1 t + C2 . t) = f (ξ)g(t). D. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 251 brings the original equation to a simpler equation of the form 3. F.g.6. z(ξ. this equation has the following solutions (A. D. There are also a traveling-wave solution. a > 0. the auxiliary equation (1) employed to determine the transformation (2) has the following solution: u(x) = 1/2 C1 exp(λx) + C2 exp(−λx) if ab > 0. Polyanin (1996). z = wu3 (2) ξ= u2 brings the original equation to a simpler equation of the form 3.5 with n = −4/3: ∂ ∂z ∂2z =a z −4/3 . z = z(αξ +βt). Ag 2 = ¦~ } 1/2 dx u2 2 . and the function g = g(x) is determined by the Yermakov’s equation agxx + f (x)g − 3λ2 g −3 = 0. Reference: V. where C is an arbitrary constant. t) = ( 2λt + C)−1/2 g(x).1. 3 The transformation dx .5 with n = 4: ∂2z ∂2z = az 4 2 . Suppose u = u(x) is any nontrivial particular solution of the second-order linear ordinary differential equation (1) auxx − 1 f (x)u = 0. where λ = 1 b/a . ∂t2 ∂ξ ∂ξ 3◦ . B. the general solution of the nonlinear equation (2) is expressed as (e. are also solutions of the equation. F. see Polyanin and Zaitsev. ¦~ Reference: V. 3 For f (x) = bxm or f (x) = beβx.1. Zaitsev and A. 3λ + (Aξ + B)2 .. Polyanin (1996). For f = b = const. C1 cos(λx) + C2 sin(λx) if ab < 0. D. t) = λ−1/4 (t + C)−1/2 3◦ .4. where k is an arbitrary constant.3. ζ = ξt−2k−1 . t) is a solution of this equation.5. ∂t2 ∂ξ For example. t) = Aξt + Bξ + Ct + D. Suppose w(x. Then the functions 3. 2003): 3λ2 2 u + u2 B + A a where A and B are arbitrary constants (A ≠ 0).

ψtt − g(t)ψ + Cψ n+1 = 0. Adams (1981). f (w) t = z. C2 . 5. 1◦ . 2◦ . t) is a solution of the equation in question. J. Traveling-wave solution: f (w) dw. F. z=x λt. w = ϕ(x) + ψ(t). C1 t + C3 ). By the change of variable U = eλϕ the ﬁrst equation is reduced to the linear equation aUxx + λf (x)U + λC = 0. are also solutions of the equation.252 4. Multiplicative separable solution: where the functions ϕ = ϕ(x) and ψ = ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) a(ϕn ϕx )x + f (x)ϕn+1 + Cϕ = 0. and C3 are arbitrary constants. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w ∂2w =a wn 2 ∂t ∂x ∂x + f (x)wn+1 + g(t)w. where C1 . . Ames. and E. w = ϕ(x)ψ(t). 2 ∂t ∂x ∂x This equation is encountered in wave and gas dynamics. g(u) = 1 . ∂2w ∂t2 =a ∂ eλw ∂w + f (x)eλw + g(t). 6. © 2004 by Chapman & Hall/CRC w = w(z). ∂x ∂x Additive separable solution: where the functions ϕ = ϕ(x) and ψ = ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) a(eλϕ ϕx )x + f (x)eλϕ + C = 0. 2 ∂τ ∂z ∂z where the function g = g(u) is deﬁned parametrically as u= 3◦ . Then the functions w1 = w( C1 x + C2 . R. The transformation x = τ. ∂ ∂w ∂2w = f (w) . Lohner. ψtt − g(t) + Ceλψ = 0. Reference: W. u= f (w) dw where w = w(z) is deﬁned implicitly by (A and B are arbitrary constants) λ2 w − ¦ f (w) dw = Az + B. Suppose w(x. leads to an equation of the similar form ∂ ∂w ∂2u = g(u) .

For C ≠ 0. F. . w = w(ξ). 8◦ . t= 1 C1 v 2 + C3 v + 2 F (w) = f (w) dw. Adams (1981). one obtains a Riccati equation for ξ = ξ(w): Cξw = ξ 2 − f (w). Ames. C4 and λ are arbitrary constants. ¦ 4◦ . D. the function H = H(w) is determined by the ordinary differential equation Lf [H] − λ2 H = 0. J. Self-similar solution: (1) To the special case C = 0 there corresponds the solution (in implicit form): References: W. dw f (w) dw (3) © 2004 by Chapman & Hall/CRC . 6◦ . 1 1 H (w) + C4 . t+b where a and b are arbitrary constants.4. Here and henceforth. 7◦ . Polyanin (1996). Zaitsev and A.3. . C1 . Solution in parametric form: x = [C1 F (w) + C2 ]v 2 + C3 F (w) + C4 + 2 t= 1 C1 v 3 + C3 v + 2v 3 f (w) [C1 F (w) + C2 ] dw dw. (2) The handbooks by Polyanin and Zaitsev (1995. [C1 F (w) + C2 ] dw + C5 . ξ 2 = f (w). treating w in (1) as the independent variable. Lohner. C5 are arbitrary constants. t = (2C1 w + C3 )v + C2 w + C5 . V. . . F. 2003) present a large number of solutions to equation (2) for various f = f (w). . equation (1) is reduced to the second-order linear equation yww = C −2 f (w)y. and E. and the function w(ξ) is determined by the ordinary differential equation ξ 2 wξ )ξ = f (w)wξ ξ . Solution in parametric form: x = (C1 eλv + C2 e−λv )H(w) + C3 . . ◦ 5 . and the differential operator Lf is expressed as Lf [ϕ] ≡ 1 dϕ d . . By the change of variable ξ = −Cyw /y. Solution in parametric form: x = [C1 F (w) + C2 ]v + C3 F (w) + C4 . ξ= which admits the ﬁrst integral ξ 2 − f (w) wξ = C. . R. Solution in parametric form: x = C1 v 2 + C2 v + f (w)(2C1 w + C3 ) dw + C4 . t = (C1 eλv − C2 e−λv ) λ f (w) w where C1 . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 253 x+a . [C1 F (w) + C2 ] dw + C5 .

Solution in parametric form: x = [2C1 F (w) + C3 ]v + C2 F (w) + C5 . . . and the function H = H(w) is determined by the ordinary differential equation Hww − λ2 f (w)H = 0. . 14◦ . Solution in parametric form: x= 1 [C2 sin(λv) − C1 cos(λv)]Zw (w) + C3 . [2C1 F (w) + C3 ] dw + C4 . . C4 and λ are arbitrary constants. 15◦ . (4) (5) x = x(w. 10◦ . λ t = [C1 sin(λv) + C2 cos(λv)]Z(w) + C4 . t = C 1 v 2 + C2 v + 11◦ . C4 and λ are arbitrary constants. and the function Z = Z(w) is determined by the ordinary differential equation Zww + λ2 f (w)Z = 0. ∂x ∂t ∂w ∂v = . λ t = (C1 eλv + C2 e−λv )H(w) + C4 . Solution in parametric form: x= 1 C1 v 3 + C3 v + 2v 3 f (w)(C1 w + C2 ) dw + C5 . . = . . v). Solution in parametric form: x= 1 (C1 eλv − C2 e−λv )Hw (w) + C3 . . Solution in parametric form: x= 1 C1 v 2 + C3 v + f (w)(C1 w + C2 ) dw + C5 . t = (C1 w + C2 )v 2 + C3 w + C4 + 2 13◦ . ∂t ∂x t = t(w. . . . The original equation can be represented as the system of equations f (w) The hodograph transformation ∂w ∂v = . where C1 . t = [C2 sin(λv) − C1 cos(λv)] λ f (w) w where C1 . C4 and λ are arbitrary constants. 1 1 Z (w) + C4 . v). brings (4) to the linear system ∂x ∂x ∂t ∂t = . 12◦ . F (w) = f (w) dw. . . the function Z = Z(w) is determined by the ordinary differential equation Lf [Z] + λ2 Z = 0. Solution in parametric form: x = [C1 sin(λv) + C2 cos(λv)]Z(w) + C3 . f (w)(C1 w + C2 ) dw dw. 2 t = (C1 w + C2 )v + C3 w + C4 .254 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 9◦ . where w and v are treated as the independent variables and x and t as the dependent ones. and the differential operator Lf is deﬁned by (3). where C1 . (6) f (w) ∂v ∂w ∂v ∂w © 2004 by Chapman & Hall/CRC .

from system (6) we obtain another linear equation for t = t(w. Polyanin (2001). Lohner. v) is deﬁned by (9) and. ψk−1 (w) = Ak F (w) + Bk + 2k(2k + 1) where the Ak and Bk are arbitrary constants (k = n. ¦ Reference: V. gives a solution of the original nonlinear equation in parametric form. v0 ) dη. Zaitsev and A. gives a solution of the original nonlinear equation in parametric form.4. . . References for equation 3. one can ﬁrst construct an exact solution to the linear equation (8) for t = t(w. v): ∂2t ∂2t − f (w) 2 = 0. Likewise. F (w) = f (w) dw. f (w) ψk (w) dw dw. D. v) is deﬁned by (9) and. The dependence t = t(w. F. v) and then determine x = x(w. 16◦ .6: W. Polyanin (2001). 17◦ . D. J. v). and E. one ﬁnds an exact solution of the linear equation (7) for x = x(w. First. . Ibragimov (1994). F (w) = f (w) dw. Further this solution is substituted into the linear system (6). . ϕk−1 (w) = Ak F (w) + Bk + 2k(2k − 1) where the Ak and Bk are arbitrary constants (k = n. Zaitsev and A. we obtain a linear equation for x = x(w. 1). together with (11). . .3. The thus obtained expressions of (5) will give an exact solution of the original equation in parametric form. F. V. F. . R. together with (10). v) from (6). Solutions of equation (7) with odd powers of v: n x= k=0 ψk (w)v 2k+1 . 1).4. v) in the form w v ∂x 1 ∂x (w. N. 2 ∂w ∂v (8) (7) The procedure for constructing exact solutions of the original nonlinear equation consists of the following two stages. Zaitsev (2002). The dependence t = t(w. Solutions of equation (7) with even powers of v: n x= k=0 ϕk (w)v 2k . A. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 255 Eliminating t. D. f (w) ϕk (w) dw dw. ξ) dξ + (η. .4. Polyanin and V. which is then solved to obtain t = t(w. H. v): 1 ∂x ∂2x ∂ − = 0. F. ∂w f (w) ∂w ∂v 2 Likewise. (9) t= f (w) ∂w w0 ∂v v0 where w0 and v0 are any numbers. © 2004 by Chapman & Hall/CRC . Adams (1981). Ames. (11) where the functions ψk = ψk (w) are determined by the recurrence relations ψn (w) = An F (w) + Bn . (10) where the functions ϕk = ϕk (w) are determined by the recurrence relations ϕn (w) = An F (w) + Bn .

z = x + λt. 1 2 bx + C1 x + C2 . where the function F = F (ξ) is deﬁned parametrically by F (ξ) = h(w) [f (w) − λ2 ]. The substitution ξ = − g(w) dw brings (1) to the canonical form uuξ − u = F (ξ).256 7. f (w) − λ2 (2) (1) A large number of exact solutions to the Abel equation (2) for various F = F (ξ) can be found in Polyanin and Zaitsev (2003). 2 ∂t ∂x ∂x f (w) Functional separable solution in implicit form: f (w) dw = at − where C1 and C2 are arbitrary constants. z = x + λt. ∂2w ∂t2 = f (w) ∂2w ∂x2 + g(w) ∂w ∂x + h(w). (2) (1) this equation is reduced to the Abel equation A large number of exact solutions to the Abel equation (2) for various F = F (ξ) can be found in Polyanin and Zaitsev (2003). By the change variable u(w) = wz this equation is reduced to the Abel equation [f (w) − λ2 ]uuw + g(w)u + h(w) = 0. ∂ ∂w = f (w) ∂t2 ∂x ∂x Traveling-wave solution: ∂2w + g(w) ∂w ∂x + h(w). w = w(z). w = w(z). 9. g(w) ξ=− g(w) dw. The substitution ξ = − g(w) dw brings (1) to the canonical form f (w) − λ2 uuξ − u = F (ξ). where the function F = F (ξ) is deﬁned parametrically as F (ξ) = h(w) . With the change of variable u(w) = [f (w) − λ2 ]wz . uuw + g(w)u + h(w)[f (w) − λ2 ] = 0. © 2004 by Chapman & Hall/CRC . 8. and the function w(z) is determined by the autonomous ordinary differential equation {[f (w) − λ2 ]wz }z + g(w)wz + h(w) = 0. g(w) ξ=− g(w) dw . HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂ ∂w f (w) ∂2w = f (w) – a2 3 + b. where λ is an arbitrary constant. 2 Traveling-wave solution: where λ is an arbitrary constant. and the function w(z) is determined by the autonomous ordinary differential equation [f (w) − λ2 ]wzz + g(w)wz + h(w) = 0.

where λ is an arbitrary constant. C2 .4. t) = w(ξ). EQUATIONS INVOLVING ARBITRARY FUNCTIONS 257 . where C1 and C2 are arbitrary constants. b. w. Suppose w(x. Suppose w(x. ξ = t+c where k. t) is a solution of this equation. In the general case. t) = ϕ(x) + ψ(t). 1◦ .5. λm2 . Multiplicative separable solution: hxx − λ f (x) −1 1−m (2) 2λ m+2 g m+2 where C1 and C2 are arbitrary constants. t) is a solution of the equation in question. are also solutions of the equation. © 2004 by Chapman & Hall/CRC a= . In particular. t) = ϕ(t)x2 + ψ(t)x + χ(t). A qualitative analysis of the structure of solutions to the original equation was undertaken in Glassey. t) = ϕ(x)ψ(t). The general solution of equation (1) is written out in implicit form: ∂2w = f (x)wm ∂2w 2◦ . the general solution of equation (2) is expressed as x (x − ξ) dξ + Ax + B. −1/2 dg = C2 t. w) ∂ w + g x. and x0 are arbitrary constants. where C1 . 4 . the equation has exact solutions of the form w(x. ∂t2 ∂x2 ∂x Equations of this form are encountered in the theory of liquid crystals and other applications. C1 t + C2 . 3◦ .3.4. 2 2 3. . 10. and C3 are arbitrary constants. and Zheng (1997) and Melikyan (1998). ∂w 2 ∂t ∂x2 ∂x 1. w(x. z = kx + λt (traveling-wave solution). Then the functions w1 = w( C1 x + C2 . The book by Polyanin and Zaitsev (2003. B. Equations of the Form ∂ w = f (x. Then the functions m 2 w1 = C1 w x. 2(m + 2) For m = 1. are also solutions of the equation. if C1 = 0. (1) h = 0. w(x. Sections 2. Hunter. The structure of other exact solutions for some speciﬁc f (w): f (w) = Aw + B. f (w) = Awk . where the functions g = g(t) and h = h(x) are determined by the ordinary differential equations gtt − λg m+1 = 0. h(x) = λ f (ξ) x0 where A. w(x. βw ◦ ∂2w = f (w) ∂2w + afw (w) ∂w 2 w(x. ∂t2 ∂x2 1◦ . it follows that C1 + g(t) = (at + C)−2/m . w(x. C1 t + C3 ). λ.3 and 2. 2◦ . t. t) = g(t)h(x). t) = w(z).7) presents a large number of exact solutions to the generalized Emden–Fowler equation (2) for various f = f (x). and c are arbitrary constants. f (w) = Ae . x+b (self-similar solution).

if C1 = 0. z= 3◦ . t) = leads to an equation of the similar form z= ∂2u = z 4−m f ∂t2 2. Suppose w(x. t). ∂t2 ∂ξ ∂ξ where the function F = F(ξ) is deﬁned parametrically by F = f (x) ψ(x) 3m+4 m+1 . t). C1 t + C2 . x 1 x u(z. are also solutions of the equation. The general solution of equation (1) is written out in implicit form: (1) (2) 2λ m+2 g m+2 where C1 and C2 are arbitrary constants.3 and 2. it follows that C1 + g(t) = (at + C)−2/m . Then the functions m 2 w1 = C1 w x. Sections 2.7) presents a large number of exact solutions to equation (2) for various F = F (z). f (x) The book by Polyanin and Zaitsev (2003. [f (x)hm hx ]x − λh = 0. and λ is an arbitrary constant.258 3◦ . 2(m + 2) (3) dx. F = λ(m + 1)f (x). t) = g(t)h(x). where C1 and C2 are arbitrary constants. f (x) . dx . Multiplicative separable solution: w(x. where the functions g = g(t) and h = h(x) are determined by the ordinary differential equations gtt − λg m+1 = 0. t) is a solution of this equation. t) = ψ(x) 1 m+1 u(ξ. ∂t2 ∂x ∂x 1◦ . ξ= ψ(x) m+2 m+1 © 2004 by Chapman & Hall/CRC a= λm2 . ∂2w 1 m ∂2u u . ψ(x) = dx . z ∂z 2 ∂ ∂w = f (x)wm . The transformation z= −1/2 dg = C2 t. f (x) dx. 2◦ . Φ = hm+1 f (x) brings (2) to the generalized Emden–Fowler equation Φzz − F (z)Φ m+1 = 0. 1 ξ= ψ(x) m+2 m+1 leads to an equation of the similar form ∂ ∂u ∂2u = F(ξ)um . The transformation HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 1 w(x. The transformation w(x. ψ(x) = dx . In particular. where the function F = F (z) is deﬁned parametrically by dx .

4. t) = (C1 t + C2 )x + a t (t − τ )(C1 τ + C2 )2 f (τ ) dτ + C3 t + C4 . Suppose w(x. t). −1 w1 = C1 w C1 x. t) = one arrives at the simpler equation 1 w(x. C4 and a are arbitrary constants.3. are also solutions of the equation.4. is also a solution of the equation. 2◦ . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 259 w ∂2w ∂2w = w4 f . Then the functions 3. leads to an equation of the form 3. t) = (C1 t + C2 )(C3 x + C4 )1/2 . C1 t + C2 . ∂t2 ∂x ∂x 1◦ . w. t) = ϕ(t)x2 + ψ(t)x + χ(t). t . ψtt = 6f (t)ϕψ. Suppose w(x. where the functions ϕ = ϕ(t).4. . The transformation √ w(x. t) ax2 + bx + c. ∂w 2 ∂t ∂x2 ∂x 1. ψ = ψ(t). t. ∂t2 x ∂x2 1◦ . 2◦ . With the transformation u(z. © 2004 by Chapman & Hall/CRC z= . t) is a solution of this equation. Equations of the Form ∂ w = f (t. Degenerate solutions: w(x. . w(x. t) is a solution of this equation. where C1 and C2 are arbitrary constants. . dx ax2 + bx + c 4. . w) ∂ w + g x. and χ = χ(t) are determined by the system of ordinary differential equations ϕtt = 6f (t)ϕ2 . ∂2w ∂t2 = w4 f w √ ax2 + bx + c ∂2w ∂x2 .6. ∂ ∂w = f (t) w . x z= 1 x ∂2u ∂2u = u4 f (u) 2 . 2 2 3.4. χtt = 2f (t)ϕχ + f (t)ψ 2 . t) = u(z. ∂t2 ∂z which has a traveling-wave solution u = u(z + λt) and self-similar solutions of the form u = u(z/t).8: ∂2u ∂2u 1 = u4 f (u) 2 + (ac − 4 b2 )u5 f (u). Then the function −2 w1 = C1 w C1 x + C2 . ∂2w where C1 and C2 are arbitrary constants. 3◦ . Generalized separable solution quadratic in x: w(x. where C1 . 2 ∂t ∂z which has a traveling-wave solution u = u(z + λt).

is also a solution of the equation. t) is a solution of this equation. = [a(t)w + b(t)] + c(t) ∂t2 ∂x2 ∂x There is a generalized separable solution quadratic in x: 3. t) = Φ(t)Ψ(x). 3. Suppose w(x. in the general case.4. t) = ϕ(t)x2 + ψ(t)x + χ(t). . A. χtt = 2f (t)ϕχ + f (t)ψ 2 + g(t)χ + h0 (t). The last equation is autonomous and has a particular solution Ψ = integrable by quadrature. where C1 . and C3 are arbitrary constants. t − 2◦ . the case of a = 1. Galaktionov (1995). (ΨΨx)x = CΨ. . and c = const was considered. λ where C1 and C2 are arbitrary constants. t) = 1 ln(C1 x + C2 ) + C3 λ F (t) dt + C4 . ¦ Reference: V. C2 . Then the function 1. 3◦ . w(x. t) = © 2004 by Chapman & Hall/CRC .7. ψ = ψ(t). it is ∂w ∂ w + g(t)w + h2 (t)x2 + h1 (t)x + h0 (t). F (t) = exp − f (t) dt . 2. Additive separable solution: w(x.260 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 4◦ . ∂2w ∂w 2 ∂2w + d(t)w + e(t)x2 + f (t)x + g(t). b = e = f = 0. and the function u = u(t) is determined by the ordinary differential equation utt + f (t)ut = 2C1 g(t)eλu . Additive separable solution: 1 ln(λC1 x2 + C2 x + C3 ) + u(t). w(x. . ∂t2 ∂t ∂x ∂x 1◦ . t) = ϕ(t)x2 + ψ(t)x + χ(t). C4 are arbitrary constants. ∂2w 1 2 6 Cx . . = f (t) ∂t2 ∂x ∂x Generalized separable solution quadratic in x: w(x. ψtt = 6f (t)ϕψ + g(t)ψ + h1 (t). and χ = χ(t) are determined by the system of ordinary differential equations ϕtt = 6f (t)ϕ2 + g(t)ϕ + h2 (t). 2 ln |C1 |. w1 = w C1 x + C2 . Multiplicative separable solution: w(x. where the functions Φ = Φ(t) and Ψ = Ψ(x) are determined by the ordinary differential equations (C is an arbitrary constant) Φtt = Cf (t)Φ2 . where the functions ϕ = ϕ(t). Other Equations Linear in the Highest Derivatives ∂w ∂ ∂w ∂2w + f (t) = g(t) eλw . λ where C1 .

w(x. see Baikov. where C1 and C2 are arbitrary constants. ψtt − g t.4. Suppose w(x. For exact solutions of this equation for some speciﬁc f (w) and g(w). ∂w ∂t + bw. =a + f x. Here. ∂2w ∂t2 =a ∂2w ∂x2 + f x. ∂2w ∂t2 ∂2w ∂x2 1 k t 0 t 0 ∂2w =a ∂2w + f x. t) + C1 cos(kt) + C2 sin(kt) if b = −k 2 < 0. k. ϕx = C. ψt − bψ = C. 5. ∂w ∂t . and Ibragimov (1989) and Ibragimov (1994). where the functions ϕ(x) and ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) aϕxx + f x. t) + C1 cosh(kt) + C2 sinh(kt) if b = k 2 > 0. Then the functions 3. wx ) = f (wx ). t) = ϕ(x) + ψ(t). are also solutions of the equation. t) is a solution of this equation. Gazizov. t) = ϕ(t) + ψ(x). w(x. λF (w) + C1 where C1 . ψx + bψ = 0. Additive separable solution: w(x. ∂w ∂x + g t. k 2 g(w) − λ2 dw = kx + λt + C2 . The solution of the ﬁrst equation is expressed as ϕ(t) = C1 cosh(kt) + C2 sinh(kt) + 1 ϕ(t) = C1 cos(kt) + C2 sin(kt) + k where C1 and C2 are arbitrary constants. ∂w 2◦ . ∂x 1◦ . g(τ ) sin k(t − τ ) dτ Special case. 2◦ . C2 .3. 4. + bw + g(t). ψtt − g t. and λ are arbitrary constants. Additive separable solution: g(τ ) sinh k(t − τ ) dτ if b = k 2 > 0. t) = ϕ(t) + ψ(z). = 2 ∂t ∂t ∂x ∂x 1◦ . there are more complicated solutions of the form w(x. For f (x. F (w) = f (w) dw. where the functions ϕ(x) and ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) aϕxx + f x. ψt = C. the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations ϕtt − bϕ − g(t) = 0. t) = ϕ(x) + ψ(t). if b = −k 2 < 0. where z = x + λt. Additive separable solution: © 2004 by Chapman & Hall/CRC . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 261 ∂w ∂w ∂ ∂2w + f (w) g(w) . ϕx + bϕ = C. ∂w ∂x + g t. Traveling-wave solution in implicit form: 2. aψxx + f x. ∂t2 ∂x2 w1 = w(x. w2 = w(x.

C5 are arbitrary constants. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE + wg t. and D are arbitrary constants. 2 2 ∂t ∂x w ∂x Multiplicative separable solution: w(x. ψtt − ψg t. 1◦ . ϕx /ϕ + bϕ ln ϕ + Cϕ = 0. t) = Axt + Bx + Cx + D. 4◦ . where the function ψ = ψ(z) is determined by the ordinary differential equation [f (zψz + ψ) − z 2 ](zψzz + 2ψz ) = 0. is also a solution of the equation. . C. ∂t2 ∂x ∂x2 For f (z) = −z. . Solution of the more general form w(x. 3◦ . where A. C1 t + C3 ) + C4 t + C5 . 7. B. ϕ= 1 2A ξf (ξ) dξ + C2 . 1 ∂w w ∂t + bw ln w. 2◦ . ∂w ∂ 2 w ∂2w =f .262 6. Then the function where C1 . and λ are arbitrary constants. The general solution of this equation in parametric form: z= f (τ ). t) = At2 + Bt + ϕ(z). and the function ϕ = ϕ(z) is determined by the ordinary differential equation 2A = f (ϕz ) − λ2 ϕzz . −1 w1 = C1 w(C1 x + C2 . 4A 5◦ . B. ψt /ψ − bψ ln ψ + Cψ = 0. Additive separable solution: w(x. and the function ϕ = ϕ(x) is determined by the ordinary differential equation 2A = f (ϕx )ϕxx . t) is a solution of the equation in question. . this equation is encountered in aerodynamics (theory of transonic gas ﬂows). ϕ= 1 2A ξf (ξ) dξ − λ2 2 ξ + C2 . t) = ϕ(x)ψ(t). where A and B are arbitrary constants. where A. Degenerate solution: w(x. ∂2w 1 ∂w ∂2w =a + wf x. where the functions ϕ(x) and ψ(t) are determined by the ordinary differential equations (C is an arbitrary constant) aϕxx + ϕf x. Suppose w(x. 1 ψ= √ 2 f (τ ) τ fτ (τ ) √ dτ + C. we have f (zψz + ψ) − z 2 = 0. 2A w = xψ(z). f (τ ) © 2004 by Chapman & Hall/CRC . z = x + λt. Its general solution can be represented in parametric form (C1 and C2 are arbitrary constants): x= 1 2A f (ξ) dξ + C1 . Self-similar solution: z = x/t. Equating the expression in square brackets to zero. . Its general solution can be represented in parametric form (C1 and C2 are arbitrary constants): z= 1 2A f (ξ) dξ − λ2 ξ + C1 . t) = At2 + Bt + ϕ(x).

h = h(t). There is the solution in multiplicative separable form: w = ϕ(x)ψ(t). 1◦ . Zaitsev (2002).4. H. and C3 are arbitrary constants. 3◦ . leads to the linear equation ∂2u ∂2u = f (τ ) 2 . Special case 1. EQUATIONS INVOLVING ARBITRARY FUNCTIONS 263 6◦ . 2 ∂τ ∂z ∂w leads to an equation of the form 3. A particular solution of the system of the ﬁrst three equations is given by f= h= 1 . 3a(t + C1 )2 g= C2 + C3 (t + C1 )3 . Generalized separable solution quadratic in x: w = (C1 t + C2 )x2 + −2 1 aC1 (C1 t 3 + C 2 )4 + C 3 t + C 4 x −4 1 + 63 a2 C1 (C1 t + C2 )7 + 1 aC1 C3 t4 6 + 1 a(C1 C4 3 + C2 C3 )t3 + aC2 C4 t2 + C5 t + C6 . C2 . D.4. Below are exact solutions of the equation for some speciﬁc f = f (U ). t) = ∂x ∂ ∂v ∂2v = f (v) . τ ) = tz + xτ − w(x. The substitution v(x. . ∂t τ= ∂w . ∂t2 ∂x ∂x ¦ Reference: N. htt = 6af h + 4ag 2 . p = p(t) are determined by the system of ordinary differential equations ftt = 18af 2 . (t + C1 )2 2 2 2aC3 aC2 C4 + C5 (t + C1 )2 + + 2aC2 C3 (t + C1 )3 + (t + C1 )8 . ptt = 2agh. where the functions f = f (t). g = g(t). ¦ Reference: A. Let f (U ) = aU k .4. C5 are arbitrary constants. . Multiplicative separable solution: w = ϕ(x)ψ(t). © 2004 by Chapman & Hall/CRC . F. . The function p = p(t) is determined from the last equation by integrating the right-hand side twice. Ibragimov (1994). . C6 are arbitrary constants. Below are exact solutions representable in explicit form: ψ(t) = A1 t−2/k if C2 = 0. ∂x where u is the new dependent variable. Polyanin and V. Generalized separable solution cubic in x: w = f (t)x3 + g(t)x2 + h(t)x + p(t).3. where C1 .6: 7◦ . t + C1 (t + C1 )2 27 where C1 . . ◦ 8 . 1◦ . and z and τ are the new independent variables. 2aC1 k+2 ψ + C2 . Let f (U ) = aU . k+2 2 (ϕ )k+2 = C1 ϕ2 + C3 . gtt = 18af g. The general solutions to these equations can be written out in implicit form. . ϕ(x) = A2 x(k+2)/k if C3 = 0. 2◦ . . The coefﬁcients A1 and A2 are determined by substituting these expressions into the above equations. k+2 x where the functions ϕ(x) and ψ(t) are determined by the autonomous ordinary differential equations (ψt )2 = and C1 . t). The Legendre transformation u(z. Special case 2. . z= ∂w .

∂t Dx = ∂ . . τ where C3 and C4 are arbitrary constants (C3 can be set equal to zero). N. . H. . t) is a solution of this equation. ¦ ln + C1 C2 x + C4 dx + ln λ x + C5 . C2 .7: N. Conservation laws for a = 1: Dt w t + Dx − Dt w t w x Dt 1 k+1 wx = 0. Zaitsev (2002). Ibragimov (1994). A. z= 1 λ f (τ ) dτ − aλ ln |τ | + C4 . a ≠ 0. a + =f w1 (x. Solution: w(x. k+1 1 2 1 w k+2 − wt + Dx − k+2 x 2 = 0. Self-similar solution: HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE w = tσ u(ζ). . A. F. The general solution of equation (1) is given by aC2 λ 1 ln cos2 (βt + C3 ) λ 2β 2 aC2 λ 1 sinh2 (βt + C3 ) ϕ(t) = − ln λ 2β 2 ϕ(t) = − ϕ(t) = − aC2 λ 1 ln − cosh2 (βt + C3 ) λ 2β 2 1 C x2 2 2 if if if where C3 and β are arbitrary constants. aC2 λ > 0. 1 k+1 wt wx k+1 = 0. © 2004 by Chapman & Hall/CRC . t) = U (z) + C1 e−t/a + C2 . . 2◦ . ϕtt = aC2 exp(λϕ). aC2 λ < 0. The general solution of equation (2) is expressed as ψ(x) = where C4 and C5 are arbitrary constants.7. 3◦ . Then the function 8. where C1 and C2 are arbitrary constants. Integrating yields its solution in parametric form: U= 1 λ f (τ ) dτ − aλτ + C3 . and λ are arbitrary constants. ζ = tβ x. ∂x References: V. Let f (U ) = a exp(λU ). the functions ϕ(t) and ψ(x) are determined by the ordinary differential equations (1 ) (2 ) aC2 λ > 0. Suppose w(x. Ibragimov (1994). σ = −(kβ + 2β + 2)/k. 1 1 2 w + w k+2 2 t (k + 1)(k + 2) x + Dx − = 0. D. is also a solution of the equation. where C1 . H.264 2◦ . Nurgalieva (1985). x + C2 ) + C3 e−t/a + C4 . G. where β is an arbitrary constant. exp(λψx )ψxx = C2 (x + C1 ). Dt twt − w + Dx − where Dt = ¦ 1 tw k+1 k+1 x ∂ . z = x + λt. λ References for equation 3. t) = w(t + C1 . Special case 3. Vinokurov and I. ∂t2 ∂t ∂x ∂x2 1◦ . Polyanin and V. C4 are arbitrary constants. ∂2w ∂w ∂w ∂2w where C1 .4. and the function u(ζ) is determined by the ordinary differential equation σ(σ − 1)u + β(2σ + β − 1)ζuζ + β 2 ζ 2 uζζ = a(uζ )k uζζ . Here. and the function U (z) is determined by the autonomous ordinary differential equation λUz = f (Uz ) − aλ2 Uzz . Generalized separable solution: w = (x + C1 )ϕ(t) + ψ(x).

x = w + awt . R. ¯¯ ¯¯ 2 w = x + awt /wx . where the function ϕ(z) is determined by the autonomous ordinary differential equation λϕz + C1 = f (ϕz ) − aλ2 ϕzz . transformation (1) leads to the linear telegraph equation a ¦ ¯ ∂2w ¯ ∂2w ∂w ¯ + =b . Dx = . u ∂ ∂ . EQUATIONS INVOLVING ARBITRARY FUNCTIONS 265 3◦ . there is an additional conservation law (k ≠ 0): 3b 1 3 3 2 2 akwt +wt (kw +2x−kxwx )+ ekwx +Dx et/a akxwt −b w + x−xwx +3awt ekwx aet/a 2 k 2 k References: S. where the prime stands for the differentiation. Ibragimov (1994). For f (wx ) = b(wx )−2 . Additive separable solution: w(x. For f (wx ) = bekwx . Φ = 0 (n + 1)(n + 2) x Ψ = bwx (ln |wx | − 1). where =F 2 ¯2 ¯ ∂t ∂t ∂x ∂x ¯ ¯ Transformation (1) has an inverse. Integrating yields its solution in parametric form: 1 1 ξf (ξ) dξ + C4 . there is an additional conservation law: Dt aet/a 3n + 4 2 awt + wt (n + 2)w − nxwx + (3n + 4)Ψ 2 + Dt et/a dΨ n 2 axwt + nxwx − a(3n + 4)wt − (n + 2)w − nxΨ + Φ 2 dwx if n ≠ −1. 5◦ . Φ = 2bw Dt ¦ Special case 3. Special case 1. H. where C1 . 1988). C2 . t) = C1 t + C2 + C3 e−t/a + ϕ(z). For f (wx ) = bwx (n ≠ 0. where Ψ= b w n+2 . N. ¯¯ ¯ ¯¯ ¯ ¯¯ ¯¯ a F (u) = 1 f u2 1 . = 0. Svirshchevskii (1986.3. ¯ wx = 1/wx . Conservation laws: Dt awt + w + Dx −Ψ (wx ) = 0. x = f (ξ) dξ + C5 . ¯¯ wt = −wt /wx ¯¯ (1) wx = 1/wx . t) = C1 t + C2 + C3 e−t/a + ϕ(x). n Special case 2. 1988).4. Dt aet/a wt + Dx −et/a Ψ (wx ) = 0. ϕ= C1 C1 where C4 and C5 are arbitrary constants (C4 can be set equal to zero). and C3 are arbitrary constants. © 2004 by Chapman & Hall/CRC . = 0. the function ϕ(x) is determined by the autonomous ordinary differential equation f (ϕx )ϕxx = C1 . ¯ leads to an equation of the similar form ¯ ¯ ∂w ∂2w ¯ ∂2w ∂w ¯ + . it is given by ¯ t = t + a ln |wx |. Ψ(u) = (u − ζ)f (ζ) dζ + C1 u + C2 . Ibragimov (1994). N. 1 Dt aet/a wt wx + Dx et/a Ψ(wx ) − wx Ψ (wx ) − 2 a(wt )2 = 0. R. The solutions of Items 3◦ and 4◦ are special cases of the more general solution w(x. z = x + λt. u wt = −wt /wx . if n = −1. The contact transformation ¯ t = t + a ln |wx |. 2 ¯ ¯ ∂t ∂t ∂ x2 ¯ References: S. −2). w = x + awt /wx . Svirshchevskii (1986. x = w + awt . 4◦ . ◦ 6 . ¯¯ (2) The formulas of (2) can be used if the Jacobian function J = (wx + awxt ) − awxx (wt + awtt¯) is ¯¯ ¯¯¯ ¯¯¯ ¯¯ ¯¯ nonzero. Dt = ∂t ∂x 0 and C1 and C2 are arbitrary constants. H.

2 3. C1 w = x n−1 U (ξ). 1◦ . C4 are arbitrary constants. βξUξξ + n−1 ξ For exact solutions of this equation. Equations Involving Arbitrary Parameters of the Form ∂ 2 w = f (w) ∂x∂y 1. τ ) 2 . w. where β is an arbitrary constant. ξ = yxβ . y) = 2(1 + n) w(x. Traveling-wave solution in implicit form (generalizes the ﬁrst solution of Item 2 ◦ ): C2 + 4◦ . is also a solution of the equation. C 1 t + C 3 + C4 . .266 9.1. Solutions: a(1 − n)2 w(x. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂w ∂ 2 w ∂2w =f . ∂x∂y This is a special case of equation 3. 2◦ . τ ) = tz + xτ − w(x. and z and τ are the new independent variables.5. . z= ∂w . 2 ∂τ ∂z Exact solutions of this equation for some speciﬁc f (z. is also a solution of the equation.5. Then the function −1 w1 = C1 w C1 x + C2 . Equations of the Form ∂ w = F x. ∂t2 ∂t ∂x ∂x2 1◦ . see the book by Polyanin and Zaitsev (2003). . and the function U (ξ) is determined by the modiﬁed Emden–Fowler equation nβ − 1 U = aU n . C4 are arbitrary constants. © 2004 by Chapman & Hall/CRC . ∂w . 3 . τ ) can be found in Polyanin (2002). y) = a(1 − n)2 ◦ 1 1−n 1 y + C2 C1 x + C1 2 1−n . where C1 . where C1 .1 with f (w) = aw n . y) is a solution of the equation in question. . leads to the linear equation ∂2u ∂2u = f (z. . Suppose w(x. 1 1−n 1 1−n xy + C1 x + C2 y + C1 C2 −1/2 . . C2 y + C4 . . The Legendre transformation u(z. . y. ∂x where u is the new dependent variable. Then the function w1 = (C1 C2 ) n−1 w C1 x + C3 .5. Suppose w(x. Self-similar solution: 2a n+1 w n+1 β−1 dw = C1 x + y + C3 . t) is a solution of this equation.3. ∂2w = awn . 2◦ . ∂w ∂x∂y ∂x ∂y 3. ∂t τ= ∂w . t). .

z = z(x.3. 1◦ . Caudrey (1980).5. 3. ∂y ∂z ∂z . Solutions (for a = λ = 1): w = ln f (x)g(y) cosh−2 C1 + C2 w = ln f (x)g(y) sinh−2 C1 + C2 w = ln f (x)g(y) cos−2 C1 + C2 1 2C2 1 g(y) dy + 2C2 1 g(y) dy + 2C2 g(y) dy − f (x) dx f (x) dx f (x) dx . ∂x ∂y 5◦ . . ∂w ∂x . S. Then the function w1 = w C1 x + C2 . Suppose w(x. v = v(x.3. . ∂w ∂y 267 2. V. ∂x ∂x λ 2 ∂w a 1 ∂u =− − exp λ(w − u) . ∂2w = aeλw . .1 with f (w) = ae λw . R. where f (x) and g(y) are arbitrary functions. K. . The original equation can also be linearized with either of the differential substitutions w= 2 ∂v 1 ln 2 λ v ∂x 2 1 w = ln λ cos2 z ∂v . 2◦ . ∂y ∂y k 2 4◦ . N. ¦ References: J. . . Bullough and P.3. H. This is a special case of equation 3. This is a special case of equation 3. Solutions: w = ln 1 − 2 ∂ 2 (ln ζk ) . Khabirov (1990). The Liouville equation is related to the linear equation ∂xy u = 0 by the B¨ cklund transformation a 1 ∂u ∂w 2k = + exp λ(w + u) . ∂2w ∂x∂y = ew – e–2w . C 3 y + C 4 + 1 ln(C1 C3 ). y.1 with f (w) = ew − e−2w .5. General solution: w= 2 1 f (x) + g(y) − ln k λ λ exp f (x) dx + aλ 2k exp g(y) dy . C4 are arbitrary constants. y). J.5. where f = f (x) and g = g(y) are arbitrary functions and k is an arbitrary constant. is also a solution of the equation. ∂x∂y (1) © 2004 by Chapman & Hall/CRC . EQUATIONS OF THE FORM ∂2 w ∂x∂y = F x. ∂x∂y Liouville equation. Liouville (1853). w. λ where C1 . 3◦ . Ibragimov (1994). and C1 and C2 are arbitrary constants. 1◦ . y). y) is a solution of the equation in question.

Caudrey (1980). one obtains an equation of the form 3. k. This is a special case of equation 3. Grauel (1985). ∂x∂y Sine-Gordon equation. Zakharov (1984). S.3. a (Ax + By + C) AB a (Ax + By + C) − AB if aAB > 0. L. 2 Ck k = 1. A. Saﬁn and R.2. if aAB < 0. and C are arbitrary constants. one obtains an equation of the form 3. ¦ Reference: R. V. k1 k2 A2 k 2 6 3 y − exp 2kx + y . 5. ∂2w ∂2w = + a sinh w. Manakov. and k2 are arbitrary constants. 2. ∂t2 ∂z 2 ¦ ∂2w References: S. Novikov. A1 .5. k1 . B. A. E. and V. S. P. Shan’ko (1999. J. ζ3 = 1 + A(k 2 x − 3y) exp kx + 2◦ . ∂2w = a sin w. O. 2◦ . A2 . Kaptsov and Yu. V. Bullough and P. C1 − C2 cosh(v1 + v2 ) vk = 1 a Ck x − y . ∂t2 ∂z 2 3◦ . 1◦ . y) = 4 arctanh exp where A.1.3.268 where HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3 y . On passing to the new independent variables z = x − y and t = x + y. other exact solutions are also given there). ∂x∂y u ¦ 2 2 (k1 − k2 )2 (k1 − k1 k2 + k2 ) exp (k1 + k2 )x + 2 + k k + k2) (k1 + k2 )2 (k1 1 2 2 Reference: S. The substitution u = ew leads to the Tzitz´ ica equation: e 1 ∂ 2 (ln u) =u− 2. y) = 4 arctan C1 + C2 sinh(v1 − v2 ) . Solution: w(x. = a sinh w. where C1 and C2 are arbitrary constants. k 12 k √ 3 3 ζ4 = sin kx − y + 3 kx + y .1 with f (w) = a sin w. k 3 ζ2 = 1 + A1 exp k1 x + y k1 ζ1 = 1 + A exp kx + + A1 A2 + A2 exp k2 x + 3 y k2 3 3 + y .1. Sharipov (1993). © 2004 by Chapman & Hall/CRC . On passing to the new independent variables z = x − y and t = x + y. K. V. ∂x∂y Sinh-Gordon equation.1: 4. k k and A. B. Traveling-wave solution: 4 arctan exp w(x. Pitaevskii.4: ∂2w ∂2w = + ew − e−2w .

The B¨ cklund transformation a w+u ∂u ∂w = + 2k sin . K. On passing to the new independent variables z = x − y and t = x + y. V. Manakov. Novikov. Lamb (1974). the formulas of (1) allow us to successively generate other solutions of the sine-Gordon equation. The ﬁrst three of them read as follows: 2 Dx wy + Dy 2a cos w = 0. ∂x ∂z = a sin w. The equation in question is related to the equation ∂ 2z =z ∂x∂y by the transformation z= ¦ a2 − ∂z ∂y 2 ∂w . N.5: R. C. I. E.3. McLaughlin (1973). J. Caudrey (1980). L. Pitaevskii.1. The sine-Gordon equation has inﬁnitely many conservation laws. one obtains an equation of the form 3. ∂2w ∂x∂y = a sin w + b sin 1 w 2 . © 2004 by Chapman & Hall/CRC . W. Bullough (1977). S.2: ∂2w ∂2w 1 = + a sin w + b sin 2 w . H. Steuerwald (1936). ¦ (1) ∂ ∂y (analogous laws can be obtained by swapping the independent References: A. 4◦ . K. R.3. where the function U = U (ξ) is determined by the second-order ordinary differential equation ξUξξ + Uξ = a sin U . Self-similar solution: w = U (ξ). Y. M. S. ξ = xy. 2 3 2 4 6 2 2 Dx 3wy − 12wy wyy + 16wy wyyy + 24wyyy + Dy a(2wy − 24wyy ) cos w = 0. Dodd and R. 6. ∂x ∂x 2 ∂w 2a w−u ∂u =− − sin ∂y ∂y k 2 brings the original equation to the identical equation ∂2u = a sin u. J. P. R. 4 2 2 Dx wy − 4wyy + Dy 4awy cos w = 0. Zakharov (1984). Chu. and V. B. ∂x∂y Given a single exact solution. ∂ where Dx = ∂x and Dy = variables x y). Krichever (1980). K. Ibragimov (1994).5. Scott. L. Calogero and A. ∂t2 ∂z 2 ¦ Reference: F. Degasperis (1982) . Bullough and P. 5◦ .∂2 w =F ∂y 269 3◦ . ∂y References for equation 3. 6◦ . and D. F.

General solution in implicit form: a exp − eβw dw = ϕ(x) + ψ(y). ∂2w ∂x∂y =a ∂w ∂w ∂x ∂y .5. ∂w ∂w ∂2w = awn . ∂2w ∂x∂y = aeβw ∂w ∂w ∂x ∂y .2. see Zhiber and Sokolov (2001).3. ∂2w ∂x∂y =a ∂w ∂w ∂x ∂y .7 with f (w) = aeβw . This equation arises in some problems of chemical engineering and chromatography. B. ∂x∂y ∂x ∂y This is a special case of equation 3. y) = f (x)g(y). 5. 1 ln f (x) + g(y) . w(x. The substitution u = eaw leads to the linear equation ∂u ∂u ∂2u +b +c = 0.3. ∂2w ∂x∂y +a ∂w ∂w ∂x ∂y +b ∂w ∂x +c ∂w ∂y = 0. w ∂2w ∂x∂y = 1– ∂w ∂x 2 (the original equation is reduced to a 1– ∂w ∂y 2 . Thomas (1944). General solution: where f (x) and g(y) are arbitrary functions. w ∂2w ∂x∂y = ∂w ∂w ∂x ∂y . C.5. 1 2 4a This is a special case of equation 3. ∂x∂y ∂x ∂y ¦ Reference: H. 3.270 HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE 3. y) = − where f (x) and g(y) are arbitrary functions. a General solution: w(x. 6.5. © 2004 by Chapman & Hall/CRC . This is a special case of equation 3. Whitham (1972). β where ϕ(x) and ψ(y) are arbitrary functions. Other Equations Involving Arbitrary Parameters 1. For this and some other integrable nonlinear hyperbolic equations.5. G. 7.7 with f (w) = aw n .3. y) = linear one). 2. 4.8 with f (x.

1◦ . η= g(y) dy The transformation © 2004 by Chapman & Hall/CRC . we obtain an equation of the form 3. C2 . ∂ξ∂η 3. w. where the function w(ξ) is determined by the second-order ordinary differential equation ξw ξξ +wξ = f (w). y. ∂x∂y ξ= f (x) dx. Equations Involving Arbitrary Functions 1. 4◦ . where C1 . ∂2w = f (x)g(y)h(w).3.5. y) is a solution of this equation.5. ξ = xy. and F (w) = ∂x ∂y 1 2 2 wx = 0. ∂t2 ∂z 2 5◦ . 2 Dx −F (w) + Dy where Dx = ∂ ∂ .1. Traveling-wave solution: w = w(z). C 1 y + C 3 . z = ax + by. Dy = . EQUATIONS OF THE FORM ∂2 w = F x. ∂ξ∂η f (x) dx.5. 2◦ . and the function w(z) is determined by the autonomous ordinary differential equation abwzz = f (w).2: ∂2w = eβw . On passing to the new independent variables z = x − y and t = x + y. 3◦ . η= g(y) dy The transformation leads to an equation of the form 3.4. ∂2w ∂x∂y = f (w).1: ∂2w ∂2w = + f (w).3. is also a solution of the equation. Suppose w(x. ∂2w = f (x)g(y)eβw . f (w) dw. and C3 are arbitrary constants. Then the function −1 w1 = w C1 x + C2 . Self-similar solution: w = w(ξ).1.5.1: ∂2w = h(w).3. Conservation laws: 2 Dx 1 wy + Dy −F (w) = 0. where a and b are arbitrary constants. 2. ∂w 271 3. ∂x∂y ξ= leads to an equation of the form 3.

∂x∂y ∂y Integrating the original equation with respect to y. ∂x a b where ψ(x) is an arbitrary function. F (w) = exp − f (w) dw ∂x∂y ∂x ∂y ◦ 1 . τ ) dτ + g(x. y)u. where G(w) = g(w) dw. 2◦ . Functional separable solution in implicit form: dw = ϕ(y) + G(w) Here. ∂2w =a ∂w ∂w + f (x. ∂x v= ∂w . 8. y) − ah(x. ∂y © 2004 by Chapman & Hall/CRC . + g(x. ∂x∂y 2◦ . The substitution u= leads to the constant coefﬁcient linear equation ∂2u = 0.272 4. Integrating the original equation with respect to y. y). y) and v = v(x. ∂2w = f (x. and a and b are arbitrary constants. ∂x∂y ∂x ∂y Goursat equation. HYPERBOLIC EQUATIONS WITH ONE SPACE VARIABLE ∂w ∂2w = f (x)g(w) . The equation obtained can be treated as an ordinary differential equation for w = w(x) with parameter y. ∂2w = f (w) ∂w ∂w . 5. ∂w ∂w ∂2w = 2 f (x. F (w) dw. w) ∂w f (x) dx. y) ∂w + g(x. General solution in implicit form: exp − f (w) dw dw = ϕ(x) + ψ(y). ∂x∂y ∂x ∂y ∂x ∂y The substitution u = e−aw leads to the linear equation 6. y) ∂w + h(x. s) ds + ψ(x). ϕ(y) is an arbitrary function. y) by the differential relations u= ∂w . we arrive at a ﬁrst-order partial differential equation: ∂w = f (x) g(w) dw + ψ(x). we arrive at a ﬁrst-order partial differential equation: w y ∂w = f (x. ∂x where ψ(x) is an arbitrary function. where ϕ(x) and ψ(y) are arbitrary functions. y). y) + g(x. ∂x∂y ∂y 1◦ . Introduce functions u = u(x. ∂x∂y ∂x ∂y 7. y) . ∂u ∂u ∂2u = f (x.

2◦ . y). f x. Integrating the original equation with respect to y. λ) dλ = b w g(x. ∂w 273 On differentiating these relations with respect to y and x. ϕ(y) is an arbitrary function. and c are arbitrary constants. © 2004 by Chapman & Hall/CRC . The equation obtained can be treated as an ordinary differential equation for w = w(x) with parameter y. y). where G(w) = g(w) dw. I. 2 ∂x Reference: E. ∂v = u f (x. The equation obtained can be treated as a ﬁrst-order ordinary differential equation in x for which the constant of integration will be dependent on y. one arrives at a ﬁrst-order partial differential equation: wx f (x. w) ∂w + h(x. ∂x a b where ψ(x) is an arbitrary function and a and b are arbitrary constants. 11. ψ(x) is an arbitrary function. one arrives at a ﬁrst-order partial differential equation: 2 w y ∂w =2 f (x. y). Functional separable solution in implicit form: 9. and eliminating w using the original equation. . one arrives at the system ∂u =v ∂y f (x. y). respectively. we have ∂w ∂x 2 = 2f (x) g(w) dw + ψ(x). y): ∂u ∂2u = g(x. ∂x ∂x∂y ∂y Integrating the original equation with respect to y. Here.5. y) = 1 ∂ ln f (x. y). s) ds + ψ(x). w) + g(x. y. ∂x ∂x∂y ∂y Integrating the original equation with respect to y. b. where ψ(x) is an arbitrary function. ∂x ∂x∂y ∂y 1◦ . ∂x∂y ∂y ¦ where g(x. τ ) dτ + c y h(x. y)u. s) ds + ψ(x). ∂w ∂2w = g(x. 10.3. a where wx is the partial derivative of w with respect to x. ∂x Eliminating v yields a linear equation for u = u(x. y) + f (x. and a. The equation obtained can be treated as an ordinary differential equation for w = w(x) with parameter y. ∂w ∂ 2 w = f (x)g(w) ∂w dw √ = ϕ(y) G(w) 2f (x) dx. ∂w ∂w ∂ 2 w = f (x. τ ) dτ + 2 g(x. EQUATIONS OF THE FORM ∂2 w = F x. w. Ganzha (2000).

Equations with Two Space Variables Involving Power-Law Nonlinearities 2 4. ζ). z1 = x|t| n−2 . y. w(x. t) = |t| 1−p F (z1 . where C1 and C2 are arbitrary constants. b(2 − m)2 4 1 x2−n − (t + C)2 . t) is a solution of the equation in question. 2 a(2 − n) b(2 − m)2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation wrr + A w + ck −1 wp = 0. z2 ). 2 ξ2 = 4 y 2−m x2−n + . y. are also solutions of the equation.1. 2◦ . y. Solution for n ≠ 2 and m ≠ 2 (generalizes the solution of Item 2◦ ): w = w(r). Suppose w(x. t) = W (y. y. Equations of the Form ∂ w = ∂ f (x) ∂w + ∂ g(y) ∂w +aw p ∂t2 ∂x ∂x ∂y ∂y 1.1. Solution for n ≠ 2. t). and p ≠ 1: 1 w= 2c(p − 1) 1+p 2 2 + + 1−p 2−n 2−m 1 p−1 1 1−p y 2−m 1 x2−n + − (t + C)2 2 a(2 − n) b(2 − m)2 4 . w(x.4. r2 = 4k y 2−m 1 x2−n + − (t + C)2 . 3◦ . 2 © 2004 by Chapman & Hall/CRC .2 with f (w) = cw p .Chapter 4 Hyperbolic Equations with Two or Three Space Variables 4. C1 2 t + C2 . η). a(2 − n)2 b(2 − m)2 1 y 2−m − (t + C)2 . r r A= 2(4 − n − m) . Then the functions w1 = C1 w C12−n x. ∂t2 axn by m p−1 p−1 p−1 ∂2w = ∂ ∂w + ∂ ∂w 1◦ . y. a(2 − n)2 4 2 η2 = 4 ζ2 = 4 w(x.1. C12−m y. t) = V (x. (2 − n)(2 − m) 4◦ . ∂x ∂x ∂y ∂y This is a special case of equation 4. + cwp . z2 = y|t| m−2 . There are “two-dimensional” solutions of the following forms: w(x. m ≠ 2. t) = U (ξ.1.

C2 . a(2 − n)2 4 2 2 z1 = x|t| n−2 . Then the functions ¡ w1 = C1 w x + p−1 1−p 1−p ln C1 . This is a special case of equation 4. 2 a(2 − n) bλ2 e−λy 1 − (t + C)2 . 1◦ . y. z2 ). y. 3. C1 2 t + C2 .4. y + p−1 p−1 1−p ln C1 . © 2004 by Chapman & Hall/CRC . are also solutions of the equation. 1◦ . 2◦ . r 2−n 4◦ . 2 a(2 − n) bλ2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 2 A A= .276 2. λ ≠ 0. Suppose w(x. r2 = 4k e−βx e−λy 1 + − (t + C2 )2 . Solution for p ≠ 1. t) is a solution of the equation in question. t) = U (ξ. 3◦ . t).4. and k are arbitrary constants. bλ2 4 η2 = 4 ζ2 = 4 ¡ w(x. ∂2w ∂t2 = ∂ ∂x aeβx ∂w ∂x + ∂ ∂y beλy 1 x2−n − (t + C)2 .3 with f (w) = cw p . 2 ξ2 = 4 ¡ e−λy x2−n + . z2 = y + ln |t|. w(x. wrr + wr + ck −1 wp = 0. λ where C1 and C2 are arbitrary constants. y. λ + cwp . and λ ≠ 0: c(p − 1)2 (r + C1 )2 w= − 2k(1 + p) ¡ 1 1−p . w(x. ζ). There are “two-dimensional” solutions of the following forms: w(x.1. β λ where C1 and C2 are arbitrary constants. t) = |t| 1−p F (z1 . and p ≠ 1: 1 w= 2c(p − 1) 1+p 2 + 1−p 2−n 1 p−1 1 1−p e−λy 1 x2−n + − (t + C)2 a(2 − n)2 bλ2 4 . t) = W (y.4 with f (w) = cw p . ∂w ∂y This is a special case of equation 4. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂2w = axn 2 ∂t ∂x ∂x + ∂ ∂w beλy ∂y ∂y + cwp . aβ 2 bλ2 4 where C1 . y + ln C1 . t) is a solution of the equation in question. C1 2 t + C2 . Solution for n ≠ 2 and λ ≠ 0 (generalizes the solution of Item 2◦ ): w = w(r). Solution for n ≠ 2.1. y. η). Suppose w(x. r2 = 4k e−λy 1 x2−n + − (t + C)2 . 2◦ . t) = V (x. y. Then the functions ¡ w1 = C1 w C12−n x. are also solutions of the equation. y. β ≠ 0.

z2 = y + ln |t|. y. 4◦ . y. Traveling-wave solutions: ¢ w= λ2 √ A(k1 x + k2 y + λt) + B . aβ 2 bλ2 e−λy 1 − (t + C)2 . t) = (A1 t + B1 )x + (A2 t + B2 )y + 12 (A2 + A2 )t4 + 3 (A1 B1 + A2 B2 )t3 + 2 (B1 + B2 )t2 + Ct + D. where C1 . Suppose w(x. bλ2 4 ζ2 = 4 ¢ w(x. © 2004 by Chapman & Hall/CRC ¢ ¢ ¢ e−βx 1 − (t + C)2 . w(x. aβ 2 bλ2 4 where the function w(r) is determined by the autonomous ordinary differential equation wrr + ck −1 wp = 0. −x sin β + y cos β. t) is a solution of this equation. C1 C2 t + C5 ). where C1 and C2 are arbitrary constants. t) = W (y. y. η). Then the functions 2. t) = V (x. y. y. t) = |t| 1−p F (z1 . y. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 277 3◦ . β λ + b ∂ w k ∂w ∂y ∂y ¢ . t). Generalized separable solution linear in space variables: 2 2 1 1 1 w(x. k1 . C5 and β are arbitrary constants. w2 = w(x cos β + y sin β. B1 . Equations of the Form ∂ w = a ∂ w n ∂w ∂t2 ∂x ∂x 1. t). C. 2 4. k2 . B2 . C1 y + C4 . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). 1 2 where A1 .4.1. B. Solution for β ≠ 0 and λ ≠ 0 (generalizes the solution of Item 2◦ ): w = w(r). ζ). 2 w1 = C2 w( C1 x + C3 . . and D are arbitrary constants.2.1. .1. and λ are arbitrary constants. Integrating yields its general solution in implicit form: ¢ C1 − 2c wp+1 k(p + 1) −1/2 dw = C2 r. 2 2 k1 + k2 where A. . There are “two-dimensional” solutions of the following forms: w(x. t) = U (ξ. 3◦ . r2 = 4k e−βx e−λy 1 + − (t + C)2 .3. . ∂x ∂x ∂y ∂y 1◦ . w(x. ∂t2 ∂x2 w ∂2w ∂t2 = ∂ w ∂w + ∂ w ∂w ∂2w =a ∂2w +b ∂ ∂w . 2 ξ2 = 4 η2 = 4 e−βx e−λy + . aβ 2 4 2 2 z1 = x + ln |t|. z2 ). . A2 .1 with c = 0. ∂y ∂y This is a special case of equation 4. 2◦ .

C4 and λ are arbitrary constants. 1/2 w(x. For other solutions. and h are solved independently of the other three (see Item 6 ◦ ).2: A. There is a “two-dimensional” solution in multiplicative separable form w(x. 7◦ . t) = w(x. . 2 C2 (x + C4 )2 . . t) = 4 t−2 (x + C1 )2 + (y + C2 )2 .1. w(x. ϕ(t). t) = 1 2 2 C1 + C2 C1 x + C2 y + C 3 t + C4 2 2 . ψtt = 2gϕ + 2(f + 3h)ψ. Solutions: HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 3 w(x.7 with a = b = n = m = 1. F. where the functions f (t). t) = (C1 t + C2 ) |U (x.1. 9◦ . . y. D. Polyanin and V. y. ¥ ¦¤ Reference for equation 4.1. y. 2 (C1 y + C2 t + C3 )2 + C1 (x + C4 )2 1/2 w(x. gtt = 6(f + h)g. htt = 6h2 + 2f h + g 2 . see equation 4. w(x. y)|. y) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. A particular solution of system (1)–(3) is given by (the general solution for f can be written out in implicit form). χtt = ϕ2 + ψ 2 + 2(f + h)χ.2. h(t). g(t) = 2f (t). g. 5◦ . g(t).6 with a = b = n = 1 and equation 4. ∂x2 ∂y 2 For this linear equation. t) = t C1 exp λx sin λy + C2 + C3 where C1 . 8◦ . £ (1) (2) (3) h(t) = f (t). y. y. t) = f (t)x2 + g(t)xy + h(t)y 2 . y. y. where the functions f (t). g(t). © 2004 by Chapman & Hall/CRC . t) = t−2 x sin λ + y cos λ + C1 . There is a generalized separable solution quadratic in space variables: w(x. .278 4◦ . . y. where the function U = U (x. y). and χ(t) are determined by the system of ordinary differential equations ftt = 6f 2 + 2f h + g 2 . t) = (At + B)−2 Θ(x. y. y. htt = 6h2 + 2f h + g 2 . and h(t) are determined by the autonomous system of ordinary differential equations ftt = 6f 2 + 2f h + g 2 . t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t).2. Zaitsev (2002). see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).2. gtt = 6(f + h)g. “Two-dimensional” solution in multiplicative separable form (generalizes the last two solutions of Item 4◦ ): w(x. t) = t C1 ln x2 + y 2 + C2 . ϕtt = 2(3f + h)ϕ + 2gψ. 6◦ . There is a generalized separable solution of the form w(x. ψ(t). where ftt = 12f 2 The ﬁrst three equations for f .

. y. y. ∂ ∂2w =a 2 ∂t ∂x 1 ∂w √ w ∂x +b ∂ ∂y 1 ∂w . y. η) are determined by the system of differential equations ∂2f ∂2f + = 0. η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. Traveling-wave solutions: § w(x. η) and g = g(ξ. where C1 . . . . t) = 2 2 aC1 + bC2 2 C3 C 1 x + C2 y + C 3 t + C4 where C1 . t) = t C1 exp λ b x sin λ a y + C2 + C3 .3 with f (w) = aw −1/2 and g(w) = bw−1/2 . y. ξ = √ . t) = (C1 t + C2 )U 2 (ξ. t).2. (1) ∂ξ 2 ∂η 2 ∂2g ∂2g + = f 2.7 with n = k = −1/2 and equation 4. . C4 are arbitrary constants. y. . C5 and β are arbitrary constants. t) = t C1 ln bx2 + ay 2 + C2 . t) = t4 w(x. y.1. . ∂ξ 2 ∂η 2 For this linear equation. y. a b where the function U = U (ξ. . 4◦ . For these linear equations. © 2004 by Chapman & Hall/CRC § § § 2 . y.2. η) . 2 . 5◦ . √ √ 2 w(x. 2◦ . . η)t + g(ξ. . Then the functions 4 w1 = C1 w( C1 C2 x + C3 .1. 3◦ . t) = (aC1 + bC2 )2 w(x. −x b/a sin β + y cos β. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 279 3. t) = cos λ sin λ √ x + √ y + C1 a b + ay 2 )−2 .4. C2 t + C5 ). see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). “Two-dimensional” solution in multiplicative separable form (generalizes the last two solutions of Item 3◦ ): y x w(x. C4 and λ are arbitrary constants. Suppose w(x.4. . t) = f (ξ. w2 = w(x cos β + y a/b sin β. . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). . y. 1◦ . where C1 . η = √ . and (2) is a Helmholtz equation (wherever f is known). t) = 2 a(C1 y + C2 t + C3 )2 + bC1 (x + C4 )2 4 C2 (x + C4 )4 2 w(x. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). “Two-dimensional” generalized separable solution quadratic in t: y x 2 w(x. Solutions: w(x. η = √ . η). C1 C2 y + C4 . ξ = √ . a b where the functions f = f (ξ. √ w ∂y This is a special case of equation 4. t) is a solution of the equation in question. . y. t + C4 C1 x + C2 y + C 3 4 −4 2 4 2 2 3 ab t (bx 2 2 w(x. . (2) ∂ξ 2 ∂η 2 Equation (1) is the Laplace equation.

t) = (sin y + Cex )2 2 C1 (at2 + At + B) . Traveling-wave solution in implicit form: 2 2 a(k1 + k2 ) ln |w| − λ2 w = A(k1 x + k2 y + λt) + B. see equation 4. w(x. © 2004 by Chapman & Hall/CRC .280 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 6◦ . y.6 with n = −1/2 and equation 4. C1 . and λ are arbitrary constants. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). y. y. t) = 2x 1 2 e cos (C1 e−x sin y + C2 ) where A. y. B 2 (x + C2 )2 at2 + At + B . w(x. t) = (C1 t + C2 ) exp Aeλx sin(λy + B) .1. w(x. t) = f2 (x. y. . t) = 2x sinh2 (C e−x sin y + C ) e 1 2 2 C1 (−at2 + At + B) . y. y)t2 + f1 (x. 4. y. y) . 7◦ . y. y.3 with f (w) = a/w and g(w) = b/w. B. y) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. w(x. k2 . Suppose w(x. ¨ ¨ ¨ 2 ∂2w =a ∂ 1 ∂w + ∂ 1 ∂w where C1 . t) = (C1 t + C2 ) exp A(x2 − y 2 ) . ∂x w ∂x ∂y w ∂y This is a special case of equation 4.1. “Two-dimensional” solution in multiplicative separable form (generalizes the ﬁrst three solutions of Item 3◦ ): w(x. C2 t + C5 ). . ∂t2 1◦ . Then the functions 2 w1 = C1 w( C1 C2 x + C3 . C5 and β are arbitrary constants. t) = (C1 t + C2 )eU (x.2. . C1 C2 y + C4 . t) = (x sin λ + y cos λ + C)2 at2 + At + B .y) . w2 = w(x cos β + y sin β. There is a “two-dimensional” generalized separable solution of the form w(x.4. where A. t) = e2x cosh2 (C1 e−x sin y + C2 ) C 2 (at2 + At + B) . t) = (C1 t + C2 )eAx+By . n = k = −1 and equation 4. B. y)t + f0 (x. .2. and λ are arbitrary constants.7 with a = b. 2◦ . a[(Ay + Bt + C1 )2 + A2 (x + C2 )2 ] . C. ∂x2 ∂y 2 For solutions of this linear equation. y. w(x.1. −x sin β + y cos β.2. w(x. . y. 3◦ . Solutions: w(x. w(x. t) is a solution of the equation in question.7 with n = m = −1/2.2. C2 . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). t) = 4◦ . t). where the function U = U (x. k1 . For other solutions. y. w(x.

2.1. t) = y n C1 x 2a w(x. 3◦ . t) = 2 2 bC1 bC1 C1 y + C 2 t + C 3 x + C4 where C1 .2.6 with a = b.2. and C are arbitrary constants. where C1 .1 with g(w) = bw n . y.1. N. For solutions of this equation. B. see 5. ¦© References for equation 4. Then the functions n −2 w1 = C1 w( C2 x + C3 .1. Solutions: 1 √ √ w(x. t) = yϕ(x − t a ) + ψ(x − t a ) n+1 .1. y) is a solution of the stationary equation ∂2Θ ∂2Θ + = AeΘ . 1 −n . t) = 1 2 bC2 2 2 t aC1 + b + C2 1 n 2 y n a(t + C1 )2 − (x + C2 )2 2 C1 x + C2 y + C 3 t + C4 2 a C2 − w(x.4. 4◦ . . 1 √ √ w(x. y. C2 t + C5 ). . C5 and λ are arbitrary constants. − 2 aC1 2 bC2 1 n . .2. . F. y. C4 are arbitrary constants. Polyanin and V.1. t) = ( 2 Aat2 + Bt + C)eΘ(x.2. 2◦ . A. y. Suppose w(x. t) = b(n + 2) w(x. Solutions: w(x. 1 n . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 281 5◦ . . 1 √ √ w(x. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). A.y) . and the function Θ(x. y. Baikov (1990). 6◦ . y. . “Two-dimensional” solution in multiplicative separable form (generalizes the last four solutions of Item 3◦ ): 1 w(x. C1 C2 y + C4 . Zaitsev (2002). . y. y. ∂x2 ∂y 2 which occurs in combustion theory. t) = yϕ(x + t a ) + ψ(x + t a ) n+1 . where ϕ(z1 ) and ψ(z2 ) are arbitrary functions. © 2004 by Chapman & Hall/CRC 2 −n . where ψ(w) is an arbitrary function and λ is an arbitrary constant. t) = y n+1 ϕ(x − t a ) + ψ(x + t a ) .7 with a = b. n = −1 and equation 4. For other solutions. . t) is a solution of the equation in question. xa−1/2 sinh λ + t cosh λ). 1◦ .1. D. Solutions in implicit form: √ √ 2λ a (y + λt) + (t a x)(bwn − λ2 ) = ψ(w). Ibragimov (1994).4. y. ∂2w ∂ ∂w ∂2w =a +b wn . ∂t2 ∂x2 ∂y ∂y This is a special case of equation 4. where A. see equation 4. y.4: V. n = m = −1. 5. . w2 = w(x cosh λ + ta1/2 sinh λ.

t) = |t|−2/n H(y. t) is a solution of the equation in question. Suppose w(x. . where the function u = u(x. in which n should be set equal to zero and k should be renamed n. C1 y + C4 . Then the functions w1 = (C2 /C1 )2/n w( C1 x + C3 . 8◦ . t) = |t|2λ G(ξ. z2 ). There are “two-dimensional” solutions of the following forms: w(x. “Two-dimensional” solution in multiplicative separable form: w(x. C5 and β are arbitrary constants. y x w(x. ¦ References for equation 4. where C1 . D. ρ2 = x exp − .2. y. the ﬁrst two solutions of Item 3◦ . η). w(x. Zaitsev (2002). y. . t) = exp − n+1 n+1 n+1 where k1 . 6◦ . . C2 t + C5 ). y. and λ are arbitrary constants. U (z1 . There is an exact solution of the form w(x. r). y.282 5◦ . w2 = w(x cos β + y a/b sin β.1. A. t) = V (z)y 2/n. where the function V = V (z) is determined by the ordinary differential equation 2an2 (zVz + Vz ) + b(n + 2)V n+1 = 0. η = nλ+1 . where θ = θ(y) is determined by the autonomous ordinary differential equation (θ n θy )y = Cθ. ρ1 = t exp − .1. 9◦ . t) is determined by the differential equation ∂ 2 u 2b(n + 2) n+1 ∂2u =a 2 + u . 7◦ . The ﬁrst solution of Item 2◦ . 1◦ .5: N. Polyanin and V. . and the solutions of Items 5◦ and 6◦ are special cases of a multiplicative separable solution w = u(x. z = (x2 − at2 )y −2 . ∂2w ∂t2 =a ∂ ∂x wn ∂w ∂x +b ∂ ∂y wn ∂w ∂y . k2 . see equation 4. t)θ(y). Solution: HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES w(x. t |t| w(x. 2 ∂t ∂x n2 For n = −1 and n = −2. Ibragimov (1994). y.2. y. this equation is linear. t) = W (z). z1 = t + k1 ln |y|. z). z2 = x + k2 ln |y|. F. t). ny ny 2y V (ρ1 . z = x2 − at2 . w(x. © 2004 by Chapman & Hall/CRC .1. ρ2 ). y. For other solutions. t) = |y| 2/n z = x/t. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). t) = u(x. Remark. This is a special case of equation 4.2.7. t)y 2/n . 6. ξ = . y.7 with n = k. −x b/a sin β + y cos β. t) = F (y. r = x2 − at2 . y.

3◦ . t) = t C1 ln bx2 + ay 2 + C2 n+1 . “Two-dimensional” solution in multiplicative separable form (generalizes the ﬁfth and sixth solutions of Item 2◦ ): w(x. 4◦ . t) = w(x. 5◦ . . y. Traveling-wave solution in implicit form: 1 n+1 . k2 . “Two-dimensional” solution in multiplicative separable form (generalizes the ﬁrst and second solutions of Item 2◦ ): w(x. 2 2 ak1 + bk2 n+1 w − λ2 w = C1 (k1 x + k2 y + λt) + C2 . t) = 2ab(n + 1) w(x. y. where the function f (t) is determined by the autonomous ordinary differential equation ftt = λf n+1 . η= √ a y. Solutions: 2 w(x. 1 n+2 w(x. t) = t C1 exp λ b x sin λ a y + C2 + C3 where C1 . y. t) = (C1 t + C2 ) U (ξ. 2 a(C1 y + C2 t + C3 )2 + bC1 (x + C4 )2 w(x. 2(n + 2) © 2004 by Chapman & Hall/CRC .4. t) = 2 (aC1 t− n (bx2 + ay 2 ) n . . ξ= √ b x. η) 1 n+1 .1. t) = f (t)Θ(x. y. C1 x + C2 y + C 3 t + C4 (x + C4 )2/n 1/n 1 2/n 1 2 + bC2 )1/n C2 2/n . k1 . where the function U = U (ξ. (1) λ is an arbitrary constant. y. (2) a ∂x ∂x ∂y ∂y A particular solution to equation (1) is given by (C is an arbitrary constant): f = (C kt)−2/n . y. . √ √ w(x. C2 . and the function Θ = Θ(x. y) is a solution of the two-dimensional stationary equation ∂Θ ∂ ∂Θ ∂ Θn +b Θn − λΘ = 0. . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). . C4 and λ are arbitrary constants. y. t) = t− n cos λ sin λ √ x + √ y + C1 a b 1 n 2 2 n . and λ are arbitrary constants. y). y. n+1 where C1 . ∂ξ 2 ∂η 2 For solutions of this linear equation. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 283 2◦ . η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. k=n λ .

see equation 4. Solutions in implicit form: C1 x + C2 y + C 3 t + C4 a b C1 y + C 2 t + C 3 x + C4 C1 x + C2 t + C 3 y + C4 2 2 2 = aC1 wn + bC2 wk . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). t) = F (r. w(x. η= y tnλ+1 “two-dimensional” solution. ρ2 ). 2◦ . where C1 . There are solutions of the following forms: r = bx2 + ay 2 x w(x. . 7. t w(x. . 2 2 wk + aC1 wn = C2 . t). “two-dimensional” solution.7 with k = n. w(x. t) = t2λ G(ξ. y.3 with f (w) = aw n and g(w) = bwk . “Two-dimensional” solution (c1 and c2 are arbitrary constants): w(x. 2 2 where C1 . w(x. y. and λ are arbitrary constants. This is a special case of equation 4. F. where the function u = u(z.284 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 6◦ . C1 C2 y + C4 . and λ are arbitrary constants. Ibragimov (1994). . C2 . z1 = x + k1 ln |t|. ξ = nλ+1 . y. Suppose w(x. y.2.4. 7◦ .1. ¦ References for equation 4. k2 . η). z = y/x w(x. t) is a solution of the equation in question. t) = e−2t V (ρ1 . y. y. ∂2w ∂t2 =a ∂ ∂x wn ∂w ∂x +b ∂ ∂y wk ∂w ∂y .6: N. t) = y 2/n H(z.4. 4◦ . C4 are arbitrary constants.4.6: ∂ ∂u ∂2u = ϕ(u) . z = c1 x + c2 y. Polyanin and V. ρ2 = yent “two-dimensional” solution. For other solutions. 2 2 wn + bC1 wk = C2 . t) = W (θ). Traveling-wave solution in implicit form: 2 2 bβ2 k+1 aβ1 n+1 w + w − λ2 w = C1 (β1 x + β2 y + λt) + C2 . “one-dimensional” solution. 1 2 . C2 t + C5 ). z2 = y + k2 ln |t| “two-dimensional” solution. z2 ).2. .1. t) = |t|−2/n U (z1 . 2 2 −2 “two-dimensional” solution. A. C5 are arbitrary constants. t) is determined by a differential equation of the form 3. t). β2 . © 2004 by Chapman & Hall/CRC ϕ(u) = ac2 un + bc2 uk . y.2. θ = (bx + ay )t where k1 . β1 . ∂t2 ∂z ∂z which can be reduced to a linear equation. . y. . 1◦ . . Zaitsev (2002). Then the functions n k −2 w1 = C1 w( C1 C2 x + C3 . t) = u(z. 3◦ . ρ1 = xent . D. . t). n+1 k+1 where C1 .

∂2w ∂t2 =a ∂2w ∂x2 + ∂ ∂y (bw + c) ∂w ∂y . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 285 5◦ .1. where C1 . Polyanin and V. There is a “two-dimensional” solution of the form (generalize the solutions of Items 3 ◦ and 4◦ ): w(x. t) = yϕ(x − t a ) + ψ(x − t a ) b √ 1/2 c √ − . t). There are exact solutions of the following forms: w(x. w(x. t) = t2λ F (ξ. ξ = s1 y + s2 t. 2 bC1 w2 = w(x cosh λ + ta1/2 sinh λ. y. y.4. D. A. t) = U (z1 . y. t) is a solution of the equation in question. z2 ).2. √ √ c w(x. © 2004 by Chapman & Hall/CRC 2◦ . ζ = x−k/n y z1 = xe . 2/n z2 = ye k/n−1 w(x. t) = v(x.8: a ∂v ∂ vn ∂x ∂x + ∂ ∂v ψ(v) ∂z ∂z = 0. t) = e −2t ξ= x tnλ+1 .7: N. Solutions: . xa−1/2 sinh λ + t cosh λ). “one-dimensional” solution. z1 = a1 x + b1 y + c1 t. nt η= y tkλ+1 kt “two-dimensional” solution.1. z2 = a2 x + b2 y + c2 t. ξ) is determined by a differential equation of the form 5.1 with g(w) = bw + c. C2 t + C5 ) + ! ! ! 2 c(1 − C1 ) . y. Ibragimov (1994). 2 2 ψ(v) = bs1 v k − s2 .1. C1 C2 y + C4 . “two-dimensional” solution. y. y. 6◦ . yt where λ is an arbitrary constant. y. which can be reduced to a linear equation. t) = x2/n G(ζ. y. t) = yϕ(x + t a ) + ψ(x + t a ) b where ϕ(z1 ) and ψ(z2 ) are arbitrary functions.2. Zaitsev (2002). C5 and λ are arbitrary constants.3. ¦ References for equation 4. where the function v = v(x. y. ξ). b √ 1/2 c √ − . y. w(x. θ=x −k/n H(z1 . t) = |y|1/2 ϕ(x − t a ) + ψ(x + t a ) − . “Two-dimensional” solution (s1 and s2 are arbitrary constants): w(x. Equations of the Form ∂ w = ∂ f (w) ∂w + ∂ g(w) ∂w ∂t2 ∂x ∂x ∂y ∂y 1. F. . η). w(x.4.4. 1◦ . . . Then the functions −2 w1 = C1 w( C2 x + C3 . Suppose w(x. y. w(x. “two-dimensional” solution. 7◦ . 2 4. t) = (x/t) U (θ). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). z2 ).4. . This is a special case of equation 4.

y. which means that the equation is integrable by quadrature. B3 . b 2 2 c a C1 y + C 2 t + C 3 C2 − . 5◦ . t) = w(x. where ϕ(w) is an arbitrary function and λ is an arbitrary constant. B2 . t) = w(x. 4◦ . the equation can be made homogeneous. and B5 are arbitrary constants. t) = C 1 x + C2 y + C 3 t + C4 + (y + C1 )2 " C2 x t 2 aC2 + b + C3 2 c − . and the function U (ξ) is determined by the ﬁrst-order ordinary differential equation 2 2 (bU + c + ab2 B4 − b2 B5 )Uξ + 2b(aB1 − B2 )U = 2(A2 − aA1 )ξ + C1 . y. © 2004 by Chapman & Hall/CRC " . A3 = 4bB3 (B2 − aB1 ). which means that the equation is integrable by quadrature. . 2 2 A2 = b(4B2 − aB3 ). Solution: 2 2 w = v(r) − 4abC1 x2 + 4bC2 t2 . 2 b bC2 2 2 2 C2 − aC1 b y + C4 C1 x + C2 t + C 3 c − . − 2 2 x + C4 b bC1 bC1 where A. the equation can be made homogeneous. With appropriate translations in both variables. . 2 b bC2 w(x. A5 = 2b(2B2 B5 − aB3 B4 ). . Solution: 6◦ . where B1 . where C1 and C2 are arbitrary constants and the function v(r) is determined by the ﬁrst-order ordinary differential equation 2 2 (bv + c)vr + 2b(aC1 − C2 )v = 8b(a2 C1 + C2 )r + C3 . A4 = 2b(B3 B5 − 2aB1 B4 ). the equation can be made homogeneous. b c 2a(y + C1 )2 − . y. 7◦ . r = y + bC1 x2 + bC2 t2 . t) = w(x. and the coefﬁcients An are expressed in terms of Bn as 2 2 A1 = b(B3 − 4aB1 ). which means that the equation is integrable by quadrature. t) = w(x. . y. where C1 and C2 are arbitrary constants and the function u(z) is determined by the ﬁrst-order ordinary differential equation 2 2 (bu + c − C2 )uz + 2abC1 u = 8a2 bC1 z + C3 . With appropriate translations in both variables. 2 w = u(z) − 4abC1 x2 . Solution (generalizes the solutions of Items 5◦ and 6◦ ): w = U (ξ) + A1 x2 + A2 t2 + A3 xt + A4 x + A5 t. t) = A w(x. y.286 3◦ . y. 2 − (x + C )2 b 3b a(t + C2 ) 3 1 2 bC2 C1 x + C2 y + C 3 t + C4 2 − 2 aC1 c − . C1 . Solutions in implicit form: √ √ 2λ a (y + λt) + (t a x)(bw + c − λ2 ) = ϕ(w). z = y + bC1 x2 + C2 t. ξ = y + b(B1 x2 + B2 t2 + B3 xt + B4 x + B5 t). Solutions: HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 2 2 C3 − aC1 c − . C4 are arbitrary constants (the ﬁrst solution is of the traveling-wave type). With appropriate translations in both variables. B4 .

1. “Two-dimensional” solution: 2 w = V (η. t) − 4abC1 x2 − 4abC1 C2 x. t). Generalized separable solution linear in y: w = F (x. The general solution of system (1)–(2) is given by F (x. Here. √ ρ = y + bC1 x − t a + ϕ(ξ) dξ. where ϕ1 (ξ). a ∂x2 ∂ζ ∂ζ ∂ζ 12◦ +. ψ1 (ξ). “Two-dimensional” generalized separable solution quadratic in y (generalizes the second and third solutions of Item 2◦ ): w = f (x. ζ) + 4bC1 t2 + 4bC1 C2 t. b b η ϕ2 (ξ) dξ − ξ G(x. t) = ψ1 (ξ) + ψ2 (η) − 1 4a 4a √ √ ξ = x + t a. t)y + h(x. t). ∂t2 ∂x (1) (2) Equation (1) is a linear homogeneous wave equation. © 2004 by Chapman & Hall/CRC . where the functions f = f (x. gtt = agxx + 6bf g. where the functions F and G are determined by the system of differential equations ∂2F ∂2F − a 2 = 0. g = g(x. and h = h(x. ∂t2 ∂η ∂η ∂η 11◦ +. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 287 8◦ . t) are determined by the system of differential equations ftt = afxx + 6bf 2 . the subscripts denote partial derivatives. where C1 and C2 are arbitrary constants and the function W (ζ. and ψ2 (η) are arbitrary functions. t)y 2 + g(x. t). ∂t2 ∂x ∂2G ∂2G − a 2 = bF 2 . Given F = F (x. “Two-dimensional” solution: 2 w = W (x. √ ξ = x + t a. t) = ϕ1 (ξ) + ϕ2 (η). η = y + bC1 x2 + bC2 x. where C1 and C2 are arbitrary constants and the function V (η. t)y + G(x. (2) represents a linear nonhomogeneous wave equation. ϕ2 (η) dη − 2 b 2a ϕ1 (ξ) dξ ϕ2 (η) dη. Solution: w = R(ρ) − 4aC1 ϕ(ξ). η = x − t a. t) is determined by the differential equation ∂ ∂V ∂2V 2 ∂V 2 = (bV + c + ab2 C2 ) + 2abC1 − 8a2 bC1 . t). t). ϕ2 (η).4. htt = ahxx + bg 2 + 2bf h + 2cf . t) is determined by the differential equation ∂ ∂W ∂2W 2 ∂W 2 + (bW + c − b2 C2 ) − 2bC1 − 8bC1 = 0. ζ = y + bC1 t2 + bC2 t. 10◦ . 9◦ .

Solution: √ A2 2 2 a AB 1 A2 t − t − (Aη + 4aB)ψ(η).2. ∂x ∂x ∂y ∂y The substitution U = aw + b leads to an equation of the form 4. z = (x2 − at2 )y −2 w(x. b With appropriate translations in both variables. (bU + c)Uz + AU − 15◦ .5 with n = 1. © 2004 by Chapman & Hall/CRC . ∂t2 ∂x ∂x ∂y ∂y 1◦ . and the function U (z) is determined by the ﬁrst-order ordinary differential equation (C is an arbitrary constant) A2 z + C = 0. t) is a solution of this equation.2. ϕ = ϕ(ξ). y. t) = t2λ G(ξ. Another solution can be obtained by substituting −t for t in (3). and the function R(ρ) is determined by the simple ordinary differential equation [(bR + c)Rρ ]ρ = 0. 14◦ . η = x − t a. 16◦ . .4. ψ = ψ(η) is an arbitrary function.3 with f (w) = a and g(w) = bw + c. Integrating yields a solution of the original equation in the form √ b(w + 4aC1 ϕ)2 + 2c(w + 4aC1 ϕ) = C2 y + bC1 C2 x − t a + C2 13◦ +. . C2 .2: ∂t2 ∂ ∂U ∂2U = U ∂t2 ∂x ∂x 3. y. t) = H(z). √ where C1 . 2. y.2. z=y+ 8a (3) where A and B are arbitrary constants. .288 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES where C1 is an arbitrary constant.1. C1 y + C3 . Then the functions w1 = w( C1 x + C2 . For other solutions.1. ψ(η) is an arbitrary function. r). η = x − t a. where λ is an arbitrary constant. Solution (obtained in the same way as in Item 12◦ ): √ b(w + 4aC1 ψ)2 + 2c(w + 4aC1 ψ) = C2 y + bC1 C2 x + t a + C2 ψ dη + C3 . . the equation can be made homogeneous. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). which means it is integrable by quadrature. w = U (z) − √ xt + 2b b b 2 ab √ √ √ A 2 x + 2 a xt − 3at2 + B(x + a t) + ψ(η) dη. 17◦ . ϕ(ξ) is an arbitrary function. “two-dimensional” solution. The substitution u = w + (c/b) leads to a special case of equation 4. ∂2w + ∂ ∂U U ∂y ∂y . Suppose w(x. . y. # # # where C1 . ϕ dξ + C3 . There are solutions of the following forms: r = x2 − at2 x y c w(x. “one-dimensional” solution. ξ = . ∂2w = ∂ (aw + b) ∂w + ∂ (aw + b) ∂w “two-dimensional” solution. see equation 4. t) = F (y. ∂ ∂w ∂ ∂w = (a1 w + b1 ) + (a2 w + b2 ) . η = λ+1 b t t w(x. C4 are arbitrary constants. η) − . and C3 are arbitrary constants. C1 t + C4 ).

C4 are arbitrary constants. . t) = (A1 t + B1 )x + (A2 t + B2 )y 2 2 1 1 + 12 (a1 A2 + a2 A2 )t4 + 1 (a1 A1 B1 + a2 A2 B2 )t3 + 2 (a1 B1 + a2 B2 )t2 + Ct + D. . © 2004 by Chapman & Hall/CRC $ $ $ λ= 2 2 b 1 C1 + b2 C2 . y. . . Traveling-wave solutions: where C1 . t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t). 2 (x + C )2 + a (C y + C t + C )2 a 2 C1 4 1 1 2 3 2 2 b1 C1 + b2 C2 . ∂t2 ∂x U ∂x ∂y U ∂y 4. t) = 1 2 + a C2 a 1 C1 2 2 C1 x + C2 y + C 3 t + C4 2 $ ∂2w = ∂ (a1 wn + b1 ) ∂w + ∂ (a2 wn + b2 ) ∂w w(x. . Traveling-wave solution: w(x. B2 . Generalized separable solution linear in space variables: w(x. y. Suppose w(x. B1 . 2 + a C2 a 1 C1 2 2 5◦ . 2 2 (C2 − b2 C1 )(x + C4 )2 − b1 (C1 y + C2 t + C3 )2 2 a2 C1 (x + C4 )2 + a1 (C1 y + C2 t + C3 )2 2 2 (C2 − b1 C1 )(y + C4 )2 − b2 (C1 x + C2 t + C3 )2 2 (y + C )2 + a (C x + C t + C )2 a 1 C1 4 2 1 2 3 where C1 . 2 ∂t ∂x aw + b ∂x ∂y aw + b ∂y The substitution U = aw + b leads to an equation of the form 4. 3◦ . 3◦ . y. C1 t + C4 ). C. − 2 2 b1 C1 + b2 C2 2 2 a 1 C1 + a 2 C2 1/n 1/n . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 289 2◦ .4.4. y. y. see equation 4. and λ are arbitrary constants. y. . Solutions: w(x. t) = w(x. k1 . and D are arbitrary constants. . A2 . 1/n . 2 2 a1 k1 + a2 k2 where A.3 with f (w) = a1 w + b1 and g(w) = a2 w + b2 . t) = w(x. y. 6◦ . t) = A k1 x + k2 y + λt + B + 2 2 λ2 − b 1 k 1 − b 2 k 2 . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). B.2. There is a generalized separable solution of the form w(x.1. t) = C1 x + C2 y + λt + C3 1 n+1 . . . Solutions: w(x. 1 2 3 where A1 . Then the functions w1 = w( C1 x + C2 . C4 are arbitrary constants. 4◦ . . t) = w(x. ∂x ∂x ∂y ∂y 1◦ .4: ∂ 1 ∂U ∂ 1 ∂U ∂2U = + . . y. ∂t2 2◦ . k2 . y. For other solutions. C2 . y. . . t) is a solution of this equation. y. t) = w(x.3 with f (w) = a1 wn + b1 and g(w) = a2 wn + b2 . For other solutions. . and C3 are arbitrary constants. C1 y + C3 . C4 are arbitrary constants.2.1. . 4◦ . ∂ 1 ∂w ∂ 1 ∂w ∂2w = + . where C1 . 2 2 (C2 − b2 C1 )(x + C4 )2 − b1 (C1 y + C2 t + C3 )2 .2. 2 a1 C1 (y + C4 )2 + a2 (C1 x + C2 t + C3 )2 where C1 . see equation 4. 5.4. t) = 1 2 + a C2 a 1 C1 2 2 C1 x + C2 y + C 3 t + C4 2 − 2 2 (C2 − b1 C1 )(y + C4 )2 − b2 (C1 x + C2 t + C3 )2 .

z2 = a2 x + b2 y + c2 t. Equation (3) is linear in g(t). w2 = w(x cos β + y sin β. ∂2w ∂2w ∂2w = (α + βw) + ∂t2 ∂x2 ∂y 2 + γw2 + δw + ε. © 2004 by Chapman & Hall/CRC % = C2 t. y) satisﬁes the two-dimensional Helmholtz equation ∆Θ + κΘ = 0. where C1 .4. t). are also solutions of the equation (the plus or minus signs are chosen arbitrarily).1. (2) is a constant-coefﬁcient linear equation. the function Θ(x. where the an . −x sin β + y cos β. . y + C2 . For solutions of this linear equation. Then the functions w1 = w( x + C1 . gtt = (γf + δ − ακ)g. Then the functions 2 2 −2 w1 = C1 w( C1 x + C2 . 3◦ . % % . 2◦ . ∂2w ∂t2 = αw ∂2w ∂x2 + ∂2w ∂y 2 –α ∂w ∂x 2 + ∂w ∂y 2 – β. t + C3 ). y. For γ ≠ 0. y. “Two-dimensional” generalized separable solution: w(x. t) is a solution of this equation. bn . t). C4 and β are arbitrary constants. .290 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 4. z2 ). it is a constant-coefﬁcient linear equation. ∂x2 ∂y 2 (1) % % % where κ = γ/β (β ≠ 0). The functions f (t) and g(t) in (1) are determined from the autonomous system of nonlinear ordinary differential equations ftt = γf 2 + δf + ε. t) = f (t) + g(t)Θ(x. −x sin β + y cos β. Particular solutions of the equation are given by f = const. C2 . and cn are arbitrary constants (n = 1. y. . Suppose w(x. and β are arbitrary constants. t) is a solution of this equation. C1 y + C3 . w2 = w(x cos β + y sin β. t) = U (z1 . where f satisﬁes the quadratic equation γf 2 + δf + ε = 0. y. For γ = 0. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). (2) (3) Equation (2) is independent of g(t). 2). 1◦ . the general solution of (2) can be written out in implicit form as df 2 3 3 γf + δf 2 + 2εf + C1 where C1 and C2 are arbitrary constants. For particular solutions of the form f = const. C1 t + C4 ). 2. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). . Suppose w(x. 1◦ . y). Other Equations 1. ∆= ∂2 ∂2 + . where C1 . z1 = a1 x + b1 y + c1 t. To the special case U = U (z1 ) there corresponds a traveling wave solution. Here. There is a “two-dimensional” solution of the form w(x. C3 .

If A1 = 0 (A2 ≠ 0). There are generalized separable solutions of the form w(x. t) = f (t) + g(t)ϕ(x) + h(t)ψ(y) + u(t)θ(x)χ(y). The coefﬁcients a1 . 2A1 a2 = 2 2 D1 − sD2 . then a2 = D1 D2 /B2 . 2B1 a 3 = A2 2 2 C1 − sC2 . ψyy = −4νψ. and B2 are arbitrary constants) ϕ(x) = A1 cosh µx + A2 sinh µx. a3 . y. C1 . where ν is an arbitrary constant. θxx = νθ. htt = −ανf h. In particular. and D1 D2 ≠ 0. one should set in (1): if ν = µ2 > 0 ϕ(x) = A1 cosh 2µx + A2 sinh 2µx ψ(y) = B1 cos 2µy + B2 sin 2µy θ(x) = C1 cosh µx + C2 sinh µx χ(y) = D1 cos µy + D2 sin µy if ν = −µ2 < 0 ϕ(x) = A1 cos 2µx + A2 sin 2µx ψ(y) = B1 cosh 2µy + B2 sinh 2µy θ(x) = C1 cos µx + C2 sin µx χ(y) = D1 cosh µy + D2 sinh µy The functions f (t). C1 C2 a 4 = B2 2 2 D1 + sD2 . y.1. D1 . g(t). A2 . we have (A1 . The functions f (t). 1 2 2 2 gtt = −4ανf g + ανa1 (D1 + sD2 )u2 . B1 . if ϕxx = νϕ and ψyy = −νψ. h(t). utt = −2αν(a3 g − a4 h)u. © 2004 by Chapman & Hall/CRC . 2 2 htt = 4ανf h − ανa2 (C1 − sC2 )u2 . C2 .4. The arbitrary constants A1 . then a4 = −B1 . If C1 = 0 (C2 ≠ 0). 2 2 2B1 D1 D2 = B2 (D1 − sD2 ). If D2 = 0 (D1 ≠ 0). and u(t) are determined by the following system of ordinary differential equations (s = sign ν): 2 2 ftt = −4αν(A2 − sA2 )g 2 + 4αν(B1 + sB2 )h2 − β. ψ(y) = B1 cos µy + B2 sin µy ϕ(x) = A1 cos µx + A2 sin µx. If B1 = 0 (B2 ≠ 0). where s = sign ν. A2 . ψ(y) = B1 cosh µy + B2 sinh µy (ν = µ2 > 0). 3◦ . where ν is an arbitrary constant. C1 C2 ≠ 0. then one should set a1 = C1 C2 /A2 . and χyy = −νχ. (1) For ϕxx = 4νϕ. g(t). then a3 = −A1 . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 291 2◦ . and D2 are related by the two constraints 2 2 2A1 C1 C2 = A2 (C1 + sC2 ). and h(t) are determined by the autonomous system of ordinary differential equations 2 2 ftt = αν(A2 − sA2 )g 2 − αν(B1 + sB2 )h2 − β. then a4 = B1 . B1 ≠ 0. D1 D2 with A1 ≠ 0. B2 . a2 . There are generalized separable solutions of the form w(x. and a4 are expressed as a1 = 2 2 C1 + sC2 . B1 . then a3 = A1 . If D1 = 0 (D2 ≠ 0). If C2 = 0 (C1 ≠ 0). 1 2 gtt = ανf g. (ν = −µ2 < 0). t) = f (t) + g(t)ϕ(x) + h(t)ψ(y).

There is a “two-dimensional” solution of the form z1 = a1 x + b1 y + c1 t. C1 t + C2 + & 2 2 2◦ . m ≠ 2. and cn are arbitrary constants (n = 1. 2 2 a(2 − n) b(2 − m) 4 . y. Solution for n ≠ 2. t) is a solution of the equation in question. C12−m y. w(x.2. ∂2w ∂t2 = ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y + ceλw . h(t). y. 2). Solution for n ≠ 2 and m ≠ 2 (generalizes the solution of Item 2◦ ): w = w(r). y. . This is a special case of equation 4. 4. χtt = 2α(f + h)χ − β. z1 = p−n−1 xt 1−p η = β1 x + β2 y + β3 t. Equations with Two Space Variables Involving Exponential Nonlinearities 2 4. λ where C1 and C2 are arbitrary constants. In the special case ϕ(t) = ψ(t) ≡ 0. . . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). . gtt = −2αg(f + h). where the an .2 with f (w) = ceλw . t) = U (z1 . Equations of the Form ∂ w = ∂ f (x) ∂w + ∂ g(y) ∂w +aeλw ∂t2 ∂x ∂x ∂y ∂y 1. . There are “two-dimensional” solutions of the following forms: w(x. There is a generalized separable solution of the form w(x.1. and λ ≠ 0: w=− 2cλ(2 − n)(2 − m) x2−n y 2−m 1 1 ln + − (t + C)2 2 λ 4 − nm a(2 − n) b(2 − m)2 4 y 2−m 1 x2−n + − (t + C)2 . 5◦ . . 3◦ . y. t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t).1. z2 = a2 x + b2 y + c2 t. To the special case U = U (z1 ) there corresponds a traveling wave solution. Then the functions 2 ln C1 . z2 = yt p−k−1 1−p U (z1 . y. C4 are arbitrary constants. Then the functions 3. t) = 2 t 1−p ξ = α1 x + α2 y + α3 t. w(x. r2 = 4k © 2004 by Chapman & Hall/CRC . g(t). C 1 y + C3 . the functions f (t). 1◦ . htt = α(2f h − 2h2 − g 2 ). Suppose w(x. ∂t2 ∂x ∂x ∂y ∂y 1◦ . z2 ). 2◦ . bn . t) is a solution of this equation. η). and χ(t) are determined by the autonomous system of ordinary differential equations ftt = α(2f h − 2f 2 − g 2 ). Suppose w(x. w1 = C1 w( C1 & & & ∂2w where C1 . ∂ ∂w ∂ ∂w = a1 wn + a2 wk + bwp . C1 t + C4 ). y. p−n−1 p−k−1 p−1 2 x + C2 .4. are also solutions of the equation. t) = F (ξ.2. w1 = w C12−n x. z2 ).292 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 4◦ .

4. r r A= 2(4 − n − m) . t) = F (z1 . There are “two-dimensional” solutions of the following forms: w(x. t) = V (x. 3◦ . t) = W (y. λ © 2004 by Chapman & Hall/CRC .4. Solution for n ≠ 2. r 2−n 4◦ . t) = V (x. Suppose w(x. w(x. 2 η2 = 4 ζ2 = 4 2 ln |t|. bλ2 4 1 x2−n − (t + C)2 . y − 2 2 2 ln C1 .1. z2 ) − 2. t) = F (z1 . 2◦ . w(x. are also solutions of the equation. η). Then the functions ' w1 = w C12−n x. a(2 − n)2 bλ2 e−λy 1 − (t + C)2 . wrr + wr + ck −1 eβw = 0. ζ). λ ∂w ∂y ' z2 = y|t| m−2 . y. t) = U (ξ. t). 2 beλy This is a special case of equation 4. t). t) is a solution of the equation in question. ∂2w ∂t2 = ∂ ∂x axn ∂w ∂x + ∂ ∂y ξ2 = 4 ' y 2−m x2−n + . z2 ) − ξ2 = 4 ' e−λy x2−n + . t) = U (ξ. r2 = 4k e−λy 1 x2−n + − (t + C)2 . ζ). EQUATIONS WITH TWO SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 293 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation wrr + A w + ck −1 eλw = 0. and λ ≠ 0: w=− 2cβ(2 − n) x2−n e−λy 1 1 ln + − (t + C)2 2 β n a(2 − n) bλ2 4 . y. y. w(x. w(x. y. y.2. a(2 − n)2 4 z1 = x|t| n−2 . Solution for n ≠ 2 and λ ≠ 0 (generalizes the solution of Item 2◦ ): w = w(r). (2 − n)(2 − m) 4◦ . y. y. + ceβw . y. η). 2 a(2 − n) b(2 − m)2 1 y 2−m − (t + C)2 . λ β where C1 and C2 are arbitrary constants. β ≠ 0.3 with f (w) = ceβw . 1◦ . t) = W (y. 2 η2 = 4 ζ2 = 4 2 ln |t|. b(2 − m)2 4 1 x2−n − (t + C)2 . w(x. C1 t + C2 + ln C1 . w(x. There are “two-dimensional” solutions of the following forms: w(x. β ' z2 = y + 2 ln |t|. y. a(2 − n)2 bλ2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 2 A A= . a(2 − n)2 4 z1 = x|t| n−2 .

. aβ 2 bλ2 e−λy 1 − (t + C)2 . 2 C2 w2 = w(x cosh λ + ta1/2 sinh λ.4. e−βx e−λy 1 + − (t + C1 )2 . There are “two-dimensional” solutions of the following forms: w(x. η). . if ckµ < 0. aβ 2 bλ2 4 r2 = 4k where C1 and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the autonomous ordinary differential equation wrr + ck −1 eµw = 0. 2 ∂t ∂x ∂x ∂y ∂y This is a special case of equation 4. t) = U (ξ. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂2w = aeβx + beλy + ceµw . y. z2 = y + ln |t|. if ckµ > 0. are also solutions of the equation. Its general solution is given by − 1 µ 1 − µ w= 1 − µ 1 − µ cµ (r + C3 )2 2k cµ ln − sin2 (C2 r + C3 ) 2 2kC2 cµ sinh2 (C2 r + C3 ) ln − 2 2kC2 cµ cosh2 (C2 r + C3 ) ln 2 2kC2 ln − if ckµ < 0. 1◦ . t) = F (z1 . y. y. e−βx e−λy + . ξ2 = 4 η2 = 4 ζ2 = 4 ( ( e−βx 1 − (t + C)2 . Equations of the Form ∂ w = a ∂ eβw ∂w ∂t2 ∂x ∂x + b ∂ eλw ∂w ∂y ∂y 1. t). y. xa−1/2 sinh λ + t cosh λ). .1.294 3. Suppose w(x. C5 and λ are arbitrary constants. Then the functions ( w1 = w x − 2◦ . . y − ln C1 . ∂y ∂y 1◦ .2. µ β λ 2 4. y. if ckµ < 0. bλ2 4 3◦ . where C2 and C3 are arbitrary constants. © 2004 by Chapman & Hall/CRC . . β λ µ where C1 and C2 are arbitrary constants. ζ). w(x. 2 2 2 ln C1 . aβ 2 4 2 2 2 w(x. C1 t + C2 + ln C1 . where C1 .4 with f (w) = ceµw . t) = V (x. C2 y + C4 . z1 = x + ln |t|. C1 t + C5 ) + ln ( ∂2w =a ∂2w +b ∂ ∂w 2 C1 . t) = W (y. Then the functions ∂t2 ∂x2 ew w1 = w(C1 x + C3 . t) is a solution of the equation in question. y.2. w(x. Suppose w(x. z2 ) − ln |t|. t) is a solution of this equation. y. Solution for β ≠ 0 and λ ≠ 0: w = w(r). are also solutions of the equation.

y. and D are arbitrary constants. B. t) = ln b sinh2 (Ax + Bt + C) 4aC − 2 ln (x + A)2 − a(t + B)2 + C + 2 ln |y + D|. y. t) = ln 1 bB 2 Ax + By + C t+D 2 − aA2 . © 2004 by Chapman & Hall/CRC χ = x/t. ξ = xt−1 . ξ = x − a t. w(x. t) = ln b w(x. ζ2 = x + k2 ln |y|. t) = U (x.2. 4◦ . 3◦ . Solutions in implicit form: √ √ 2λ a (y + λt) + (t a x)(bew − λ2 ) = ϕ(w). y. where f = f (ξ) and g = g(η) are arbitrary functions and k is an arbitrary constant. ζ1 = t + k1 ln |y|. ρ1 = tey . y. y. y. b(Ax + Bt + C)2 (B 2 − aA2 )(y + D)2 . C. bB 2 2 a B2 − bA2 bA2 Ay + Bt + C x+D . w(x. t) + 2 ln |y + C|. √ √ w(x. t) = f (ξ) + g(η) − 2 ln k ef (ξ) dξ − 4ak √ √ η = x + a t. y. η = y|t|k−1 . y. 2 cos2 (Ax + Bt + C) 2◦ . t) = U (ρ1 .1.4. where C1 and C2 are arbitrary constants and ϕ(z1 ) and ψ(z2 ) are arbitrary functions. ρ2 = xey . t) = G(ξ. 6◦ . There are “two-dimensional” solutions of the following forms: w(x. y. w(x. y. w(x. y. and k2 are arbitrary constants. “Two-dimensional” solution: w(x.1: ∂2U ∂2U = a 2 + 2beU . b cosh2 (Ax + Bt + C) (B 2 − aA2 )(y + D)2 . . w(x. t) = ϕ(x − t a ) + ψ(x + t a ) + ln |C1 y + C2 |. t) = V (χ) + 2 ln |y/t|. t) = ln w(x. √ √ w(x. ζ2 ) + 2 ln |y|. where the function U = U (x. Solutions: (aA2 − B 2 )(y + D)2 . y. y. k1 . t) = ln w(x. t) = F (y.2. η) − 2k ln |t|. ∂t2 ∂x Integrating yields a solution of the original equation in the form b eg(η) dη + 2 ln |y + C|. r = x2 − at2 . y. y. r). t) = ln (B 2 − aA2 )(y + D)2 . where A. y. 5◦ . Solutions: w(x. where ϕ(w) is an arbitrary function and λ is an arbitrary constant. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 295 √ √ w(x. t) = H(ζ1 . t) = ln yϕ(x − t a ) + ψ(x − t a ) . y. t) is determined by a differential equation of the form 3. ρ2 ) + 2y. w(x. where k. t) = ln w(x. ) w(x. t) = ln yϕ(x + t a ) + ψ(x + t a ) .

8◦ . ∂x ∂x ∂y ∂y 1◦ . t) = C1 t + C2 + ln C3 (bx2 − ay 2 ) + C4 xy + C5 . C1 y + C4 . ∂t2 eλw eλw w1 = w(C1 x + C3 . t) = C1 t + C2 + b C3 x + ln cos a C3 λy + C4 . t) = C1 t + C2 + ln C3 ln(bx2 + ay 2 ) + C4 . F. where C1 . λ cos2 (bC1 t + C5 ) √ √ 2 aC1 x2 + C2 exp b C3 x cos a C3 y + C4 1 . y. bC1 x2 + C2 xy + Ky 2 + C3 x + C4 y + C5 1 ln . y. t) = w(x. t) = C1 t + C2 + ln C3 exp b C4 x cos a C4 y + C5 + C6 . Solutions: 1 ln(C3 x + C4 y + C5 ). y. λ 1 w(x. λ √ √ 1 w(x. C2 t + C5 ) + w2 = w x cos β + y 2◦ . t) = w(x. are also solutions of the equation. y. Then the functions 2. λ √ √ 1 w(x. t) = ln 2 2 λ t + C4 aC1 + bC2 w(x. t) = w(x. t) = W (z). λ C1 a/b sin β. t) = w(x. t . b C2 bC1 x2 + C2 xy − Ky 2 + C3 x + C4 y + C5 1 . .296 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 7◦ . . . References for equation 4. . ln λ cosh2 (aC1 t + C5 ) © 2004 by Chapman & Hall/CRC . For other exact solutions. −x b/a sin β + y cos β. y. y. y. t) = C1 t + C2 + w(x. A. Ibragimov (1994).2. y.1: N. ln λ sinh2 (bC1 t + C5 ) √ √ 2 −aC1 x2 + C2 exp b C3 x cos a C3 y + C4 1 . t) = w(x. C5 and β are arbitrary constants. y. y. y. Zaitsev (2002). t) = 2 C2 (x + C4 )2 1 . Polyanin and V. 1 ¦0 z = (x2 − at2 )y −2 . t) is a solution of this equation. ln λ b cosh2 (C1 t + C6 ) √ √ 2 2 aC1 x + C2 exp b C3 x cos a C3 y + C4 1 ln . ln λ sinh2 (aC1 t + C5 ) √ √ 2 bC1 y 2 + C2 exp b C3 x cos a C3 y + C4 1 . y. y. y. t) = w(x. Suppose w(x. b 2 C1 − aC1 . λ 1 w(x.2. λ cos2 (aC1 t + C5 ) √ √ 2 bC1 y 2 + C2 exp b C3 x cos a C3 y + C4 1 ln . t) = w(x.4. y. y. There is an exact solution of the form w(x. ln λ sinh2 (C1 t + C6 ) K= K= 2 C1 − aC1 . λ cos2 (C1 t + C6 ) bC1 x2 + C2 xy + Ky 2 + C3 x + C4 y + C5 1 . λ 2 1 C1 x + C2 y + C 3 1 . K = 1 + aC1 . ln 2 λ a(C1 y + C2 t + C3 )2 + bC1 (x + C4 )2 2 ∂2w =a ∂ ∂w +b ∂ ∂w 2 1 C2 ln 2 . w(x. y. . see equation 4. D.2. t) = w(x.3 with f (w) = a and g(w) = be w .

t). . Zaitsev (2002).2. 2 λ 2C1 f (t) = − 1 ln λ sinh2 (C t + C ) if λ > 0. (2) ∂ξ 2 ∂η 2 For this linear equation. 4 ¦3 4◦ . F. η). w(x. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). t) = W (z) + ln λ t x bx2 + ay 2 “one-dimensional” solution. λ x w(x. . 1 λ 2 1 − ln λ cos2 (C1 t + C2 ) if λ > 0. t) = R(ζ). 2 λ 2C1 5◦ . The general solution of equation (1) is given by − 1 ln 1 λ(t + C )2 if λ > 0. k1 . η) − ln |t|. η = y + k2 ln |t| “two-dimensional” solution. ρ2 ) − t. r = bx2 + ay 2 “two-dimensional” solution. ζ = t2 where k.2. x y 1 ln V (ξ. z2 ) − λ 2 w(x. “Two-dimensional” solution (generalizes the ﬁrst ﬁve solutions of Item 2 ◦ ): x y 1 w(x. D. ρ2 = yet λ x y 2 . t) = V (ρ1 . t) = f (t) + ftt = eλf . and k2 are arbitrary constants. ∆V − λ = 0. 3◦ . z= “one-dimensional” solution. “Two-dimensional” solution: (1) References for equation 4. . t) = G(r. y. t) = C1 t + C2 + ln U (ξ.2: N. λ a b where C1 and C2 are arbitrary constants and the function U = U (ξ. Polyanin and V. C6 are arbitrary constants.4.2. y. w(x. y. w(x. y. y. There are solutions of the following forms: y 2 “two-dimensional” solution. η = √ . see equation 4. ∆= 6◦ . y. . y. η) is a solution of the Poisson equation ∂2 ∂2 + . t) = H(z1 . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). 1 2 λ 2 2C1 1 − ln − λ cosh2 (C1 t + C2 ) if λ < 0. w(x. ξ = x + k1 ln |t|.4.2. η = √ . ξ = √ . Ibragimov (1994). © 2004 by Chapman & Hall/CRC . t) = F (z. y.3 with f (w) = ae λw and g(w) = beλw . y. A. and the function V = V (ξ. ρ1 = xet . t) = ln λ cosh2 (bC1 t + C5 ) where C1 . . t) + ln x. z2 = y|t|k−1 “two-dimensional” solution. λ a b where the function f = f (t) is determined by the autonomous ordinary differential equation w(x. w(x. η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. y. ∂ξ 2 ∂η 2 For this linear equation. η). For other exact solutions. z = . z1 = x|t|k−1 . t) = U (ξ. 2k ln |t|. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 297 √ √ 2 −bC1 y 2 + C2 exp b C3 x cos a C3 y + C4 1 w(x. ξ = √ . λ 2 “two-dimensional” solution.

z = x−λ y. ξ) is determined by a differential equation of the form 5. “Two-dimensional” solution (c1 and c2 are arbitrary constants): w(x. “Two-dimensional” solution (s1 and s2 are arbitrary constants): w(x. z2 = a2 x + b2 y + c2 t. 5 5 5 2 C1 x + C2 t + C 3 y + C4 where C1 . where the function u = u(z.6: ∂ ∂u ∂2u = ϕ(u) . Traveling-wave solution in implicit form: 2 2 ak1 ew + bk2 λ−1 eλw − β 2 w = C1 (k1 x + k2 y + βt) + C2 . . A. ξ = s1 y + s2 t. 4◦ . 2 ∂t ∂x ∂x ∂y ∂y 1◦ . where the function U = U (ξ) is determined by the ordinary differential equation aλ2 ξ 2 eU + beλU − (λ − 1)2 ξ 2 Uξξ + λ aλξ 2 eU + beλU (Uξ )2 + ξ aλ(λ − 3)eU − (λ − 1)(λ − 2) Uξ + 2(aeU − 1) = 0. 6◦ . z1 = a1 x + b1 y + c1 t. z2 ). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). t) = v(x. . Solutions in implicit form: C1 x + C2 y + C 3 t + C4 a b C1 y + C 2 t + C 3 x + C4 2 2 2 = aC1 ew + bC2 eλw . Then the functions λ w1 = w( C1 C2 x + C3 .2. . ϕ(u) = ac2 eu + bc2 eλu . t) is a solution of this equation. C5 are arbitrary constants. There is a “two-dimensional” solution of the form (generalize the solutions of Items 3 ◦ and 4◦ ): w(x. a ∂x ∂x ∂z ∂z which can be reduced to a linear equation.298 3. y. 5◦ . 1 2 ∂t2 ∂z ∂z which can be reduced to a linear equation.3: N. © 2004 by Chapman & Hall/CRC . t) = U (ξ) + 2 ln(x/t). t) is determined by a differential equation of the form 3. “Two-dimensional” solution: w(x. 7 ¦6 2 References for equation 4. Ibragimov (1994).8: ∂v ∂ ∂v ∂ 2 2 ev + ψ(v) = 0. where the function v = v(x.4. ψ(v) = bs1 eλv − s2 . Zaitsev (2002). C2 . y. Solution: w(x. C1 C2 y + C4 .4. t) is determined by the differential equation ∂2u ∂u ∂u ∂2u = aλ2 z 2 eu + beλu + λ aλz 2 eu + beλu + aλ(λ − 3)zeu + 2aeu . ξ = x−λ ytλ−1 .4. 3◦ .3 with f (w) = ae w and g(w) = beλw . 2 2 2 eλw + aC1 ew = C2 . z = c1 x + c2 y. . y. . k2 . F. y. 8◦ . . For other exact solutions. where C1 . Suppose w(x. ◦ 7 .2. k1 . C1 t + C5 ) − 2 ln |C2 |. t) = u(z. . ξ). see equation 4.2. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂2w =a ew +b eλw . D. where C1 . 2 2 ew + bC1 eλw = C2 . and β are arbitrary constants. 2 ∂t ∂z 2 ∂z ∂z 9◦ . y. y. 2◦ . t) + 2 ln x.4. Polyanin and V. C4 are arbitrary constants. where the function u = u(z. . t). t) = u(z. t) = U (z1 .4.

Solutions: w(x. . t) = w(x. y. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). NONLINEAR TELEGRAPH EQUATIONS WITH TWO SPACE VARIABLES 299 4. t) = w(x. y. 2 ∂t ∂x ∂x ∂y ∂y 1◦ . η1 = a1 x + b1 y + c1 t. Suppose w(x. Then the functions 1. ∂t ∂y ∂y 1◦ . β−λ β−λ β w1 = w( C1 1 x + C2 . Traveling-wave solutions: 8 8 8 8 1 ln C1 x + C2 y + βt + C3 . There are “two-dimensional” solutions of the following forms: λ1 −β 2 ln |t|. . t) is a solution of this equation. C4 are arbitrary constants. 2 + a C2 a 1 C1 2 2 where C1 . For other solutions.1. 2◦ .4. y. ∂ ∂w ∂ ∂w ∂2w =a eλ1 w +b eλ2 w + ceβw . . C4 are arbitrary constants. where C1 . y. w1 = w( C1 x + C2 . and C3 are arbitrary constants. C1 t + C4 ) + 2 ln |C1 |.2. t) = 3◦ . C1 2 y + C3 .3. C1 x + C2 y + C 3 t + C4 2 − 2 2 (C2 − b2 C1 )(x + C4 )2 − b1 (C1 y + C2 t + C3 )2 1 . 4◦ . Other Equations ∂ ∂w ∂ ∂w ∂2w = (a1 eλw + b1 ) + (a2 eλw + b2 ) . Then the functions 2. y. . . C2 . . t) is a solution of this equation. . t) = 1 1 ln 2 2 λ a 1 C1 + a 2 C2 β= 2 2 b 1 C1 + b2 C2 . . . t)y + g(x. t) = V (η1 . “Two-dimensional” generalized separable solution linear in y: ∂t2 ∂x2 w = f (x. 8 8 8 where C1 . β w(x. .3. t) = U (ξ. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). Equations Involving Power›Law Nonlinearities 1. C1 t + C4 ). η2 ). 4. © 2004 by Chapman & Hall/CRC ∂2w +k ∂w =a ∂2w + ∂ (bw + c) ∂w . . y. t). 2 ∂t ∂x ∂x ∂y ∂y 1◦ . η2 = a2 x + b2 y + c2 t. ln 2 λ a2 C1 (x + C4 )2 + a1 (C1 y + C2 t + C3 )2 2 2 (C2 − b1 C1 )(y + C4 )2 − b2 (C1 x + C2 t + C3 )2 1 .3. C4 are arbitrary constants. y. . see equation 4. y. 2◦ .3. λ2 −β β w(x. ln 2 λ a1 C1 (y + C4 )2 + a2 (C1 x + C2 t + C3 )2 2 2 b1 C1 + b2 C2 . C1 y + C3 . Suppose w(x. Nonlinear Telegraph Equations with Two Space Variables 4. λ where C1 . .3 with f (w) = a1 eλw + b1 and g(w) = a2 eλw + b2 . η) − η = y|t| .4.2. ξ = x|t| β . w(x.

t). w2 = w(x cos β + y sin β. . Suppose w(x. Generalized separable solution linear in the space variables: w(x. t). ∂w ∂ ∂2w +k =a 2 ∂t ∂t ∂x 1 ∂w w ∂x + ∂ ∂y 1 ∂w w ∂y . C2 . y. the second one represents a linear nonhomogeneous telegraph equation. ∂t2 ∂t ∂x The ﬁrst equation is a linear homogeneous telegraph equation. 1◦ . C4 and β are arbitrary constants. Then the functions 2 w1 = C1 w( C1 x + C2 . y.4 with m = 1. For these equations. and D are arbitrary constants. 3. 4◦ . β1 . t)y 2 + g(x. k k where A1 . where C1 . 3◦ . B1 . C. Additive separable solution: w(x. This is a special case of equation 4. where C1 . A2 . t) = Akx + Bky + Ce−kt + k(A2 a1 + B 2 a2 )t + D. C1 .1. The substitution u = w + (c/b) leads to the special case of equation 4. B2 . and λ are arbitrary constants. t)y + h(x. There is a “two-dimensional” generalized separable solution quadratic in y: w = f (x.300 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES where the functions f and g are determined by the one-dimensional equations ∂f ∂2f ∂2f +k =a 2. 2◦ . and C2 are arbitrary constants. β2 . t) is a solution of the equation in question. 3◦ . t) = (A1 e−kt + B1 )x + (A2 e−kt + B2 )y + − 1 (a1 A2 + a2 A2 )e−2kt 1 2 2k 2 1 2 2 2 (a1 A1 B1 + a2 A2 B2 )te−kt + C1 e−kt + (a1 B1 + a2 B2 )t + C2 . Traveling-wave solution in implicit form (k ≠ 0): 2 2 2 2 2 2 kλ(a1 β1 + a2 β2 )w + [kλ(b1 β1 + b2 β2 − λ2 ) − C1 (a1 β1 + a2 β2 )] ln(kλw + C1 ) = k 2 λ2 (β1 x + β2 y + λt) + C2 . . There is a generalized separable solution of the form w(x. −x sin β + y cos β. t) = f (t)x2 + g(t)xy + h(t)y 2 + ϕ(t)x + ψ(t)y + χ(t). ∂2w ∂t2 +k ∂w ∂t = ∂ ∂x (a1 w + b1 ) ∂w ∂x + ∂ ∂y (a2 w + b2 ) ∂w ∂y . B. C1 y + C3 . 1◦ . . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). Given f = f (x. © 2004 by Chapman & Hall/CRC 9 9 . 2. t).6 with n = m = −1.3. where A. . 2◦ . y. t + C4 ). see the book by Polyanin (2002).3.1. 2 ∂t ∂t ∂x ∂g ∂2g ∂2g +k = a 2 + bf 2 . y.

y) . this equation is linear. C1 y + C3 . w(x.6 with n = 0. H. t) is determined by the telegraph equation Reference: N. k(sin y + Cex )2 2 C1 (2at + A + Be−kt ) . w(x. if m = −1. and C2 are arbitrary constants. “Two-dimensional” multiplicative separable solution: w(x.1. where A. . Ibragimov (1994). t) = 2x 1 2 ke cos (C1 e−x sin y + C2 ) where A. © 2004 by Chapman & Hall/CRC . . and C are arbitrary constants and the function Θ(x.3. y. see the book by Polyanin (2002). Suppose w(x. C1 . B. y. B B ∂2w +k ∂w =a ∂2w +b ∂ ∂w where C1 .1. Then the functions m −2 w1 = C1 w( x + C2 . 2◦ . C4 are arbitrary constants. “Two-dimensional” solution: w(x. t) = u(x. y.2. y. A ¦@ 2◦ . ∂t2 ∂t ∂x For solutions of this linear equation. . see 5. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). For solutions of this equation.3. ∆Θ − AkeΘ = 0. 3◦ . where the function u(x. ∂t2 ∂x2 wm 1◦ .4. t) exp(Cy) ∂U ∂2U ∂2U +k =a 2. t)y 2/m . y. t) = (Aat + B + Ce−kt )eΘ(x. t) = U (x. Ibragimov (1994). Solutions: Reference: N. y. ∆= ∂x2 ∂y 2 which occurs in combustion theory. The exact solutions speciﬁed in Item 2◦ are special cases of a more general solution in the form of the product of functions with different arguments: w(x. H. t) is determined by the differential equation ∂u ∂ 2 u 2b(m + 2) m+1 ∂2u +k =a 2 + u . . C. t)|y + C|1/(m+1) U (x. y. where C is an arbitrary constant and the function U (x. t) = ke2x sinh2 (C1 e−x sin y + C2 ) 2 C1 (−2at + A + Be−kt ) . . 4. t) is a solution of the equation in question. NONLINEAR TELEGRAPH EQUATIONS WITH TWO SPACE VARIABLES 301 2at + A + Be−kt . w(x. t + C4 ). t) = ke2x cosh2 (C1 e−x sin y + C2 ) C 2 (2at + A + Be−kt ) . ∂t ∂y ∂y This is a special case of equation 4. w(x. y) is a solution of the stationary equation ∂2 ∂2 + . y. B.1. 2 ∂t ∂t ∂x m2 For m = −2 and m = −1. A ¦@ if m ≠ −1. t) = 3◦ .

2◦ . t + C4 ). . y) satisﬁes the stationary equation a ∂Φ ∂ Φn ∂x ∂x +b ∂ ∂Φ Φn ∂y ∂y = CΦ. For C = 0. . Suppose w(x. =a 2 ∂t ∂t ∂x ∂x ∂y ∂y 1◦ . 2 ∂t ∂t r ∂r ∂r 4◦ . y. C1 y + C3 . ¯ ¯ a b 3◦ . t) = u(r. . H. t) is a solution of this equation. t + C4 ). y. . C2 . ∂2w +k ∂w =a ∂ ∂w +b where C1 . t). −x where C1 . it follows from (1) that F = Ae−kt + B. . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). w2 = w x cos β + y a/b sin β. y). . Suppose w(x. Then the functions 6. (2) Example. and λ are arbitrary constants. Then the functions n n −2 w1 = C1 w( C1 x + C2 . C1 y + C3 . © 2004 by Chapman & Hall/CRC C C C C b/a sin β + y cos β. y. . . t) = U (t)(bx2 + ay 2 )1/n . y. 2◦ . y. (1) and the function Φ(x. ∂ x2 ¯ ∂ y2 ¯ x y where Ψ = Φn+1 . t . . For C = 0. Multiplicative separable solution: w(x. β2 . “Two-dimensional” solution: w(x. C4 and β are arbitrary constants. where A and B are arbitrary constants. t) is a solution of this equation. n2 ∂w where the function U (t) is determined by the autonomous ordinary differential equation Utt + kUt = E ¦D Reference: N. Traveling-wave solution in implicit form: 2 2 a 1 β1 w n + a 2 β2 w m − λ 2 dw = β1 x + β2 y + λt + C2 . kλw + C1 where C1 . Ibragimov (1994). t) is determined by the differential equation ∂u ab ∂ ∂u ∂2u +k = run .302 5. where the function u(r. 4ab(n + 1) n+1 U . t) = F (t)Φ(x. ∂ . Solution: w(x. x = √ . where the function F (t) is determined by the autonomous ordinary differential equation (C is an arbitrary constant) Ftt + kFt = CF n+1 . ∂t2 wn wm n m −2 w1 = C1 w( C1 x + C2 . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂w ∂w ∂ ∂w ∂ ∂2w +k wn +b wn . r= bx2 + ay 2 . ∂t ∂x ∂x ∂y ∂y 1◦ . y = √ . equation (2) is reduced to the Laplace equation ∂2Ψ ∂2Ψ + = 0. C4 are arbitrary constants. β1 .

where the βi . t) is determined by the differential equation ∂U ∂ ∂U ∂2U +k = aβ 2 U n + bµ2 U m . t) = U (ξ. y. and λ = 1.3. ∂t2 ∂t ∂x2 ∂y ∂y This is a special case of equation 4. . “Two-dimensional” solution: w(x. +k =a 1◦ . where β and µ are arbitrary constants. Then the functions w1 = w( x + C2 .4. where C is an arbitrary constant and the function U (x. µi . t + C4 ) − 2 ln |C1 |. “two-dimensional” solution of the form w(x. t) = y 2/m u(z.3. “Two-dimensional” solution: w(x. ∂t2 ∂t ∂ξ ∂ξ Remark. where the function u = u(z. C1 y + C3 . ∂u ∂2u +k ∂t2 ∂t ∂2u ∂z 2 ∂u ∂z 2 ξ1 = β1 x + µ1 y + λ1 t. ξ = βx + µy. C4 are arbitrary constants. t). ∂z Reference: N. Ibragimov (1994). t) = u(x. t) is a solution of the equation in question. . t) = V (ξ1 . . . where C is an arbitrary constant and the function u(x. y. y. 4. “Two-dimensional” solution: w(x. t) is determined by the differential equation m2 = am2 un + bn2 z 2 um + nm amun−1 + bnz 2 um−1 G ¦F + bn(n − 3m − 4)zum ∂u + 2b(m + 2)um+1 .4. H. t) + 2 ln |y + C|. y. ξ2 = β2 x + µ2 y + λ2 t. 4◦ . t) is determined by the differential equation ∂U ∂2U ∂2U +k = a 2 + 2beU . ∂t2 ∂t ∂x ◦ 3 . g(t) = a. t) = U (x. t) + ln |y + C|. and the function U = U (ξ. “Two-dimensional” solution: w(x. Suppose w(x. Equations Involving Exponential Nonlinearities ∂ ∂w +b ew . y. ∂t2 ∂t ∂x G ¦F H H ∂2w ∂w ∂2w Reference: N. y. H. t) is determined by the linear telegraph equation ∂u ∂ 2u ∂2u +k =a 2. z = xy −n/m . t). 2◦ . There is a more general.3. ξ2 ). where C1 . NONLINEAR TELEGRAPH EQUATIONS WITH TWO SPACE VARIABLES 303 3◦ .2. Ibragimov (1994). 1.10 with f (t) = k. and λi are arbitrary constants. © 2004 by Chapman & Hall/CRC . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). h(t) = b.

y = √ . y)]. β2 . kγw + C1 where C1 . t) is determined by the differential equation ∂u ab ∂ ∂u ∂2u +k = reu . C2 . . Traveling-wave solution in implicit form: 2 2 a1 β1 ew + a2 β2 eλw − γ 2 dw = β1 x + β2 y + γt + C2 . y. y. “Two-dimensional” solution: w(x. β1 . and γ are arbitrary constants. t) = ϕ(t) + ln[ψ(x. . 5◦ . Suppose w(x. t) is a solution of this equation. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). r= bx2 + ay 2 . There is a “two-dimensional” solution of the form w(x. t) = u(z. y) satisﬁes the Poisson equation x y where x = √ . y. =a 2 ∂t ∂t ∂x ∂x ∂y ∂y 1◦ .304 2. H. Then the functions 3. y. 2◦ . −x b/a sin β + y cos β. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂w ∂w ∂ ∂w ∂ ∂2w +k ew +b ew . ∂t2 ew eλw λ w1 = w( C1 x + C2 . Solution: w(x. C1 y + C3 . . . . Q ¦P I I a/b sin β. t) is a solution of this equation. t) = u(t) + ln(bx2 + ay 2 ). ¯ ¯ a b For solutions of this linear equation. ∂2w +k ∂w =a ∂ ∂w +b ∂ ∂w where C1 . t). t + C4 ) − 2 ln |C1 |. where the function u(r. t . ∂t ∂x ∂x ∂y ∂y 1◦ . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). 2 ∂t ∂t r ∂r ∂r 4◦ . Suppose w(x. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). . Then the functions w1 = w( C1 x + C2 . Reference: N. ∂2ψ ∂2ψ + = C. C4 and β are arbitrary constants. . ∂ x2 ¯ ∂ y2 ¯ where the function u(t) is determined by the autonomous ordinary differential equation z = y/x. w2 = w x cos β + y where C1 . “Two-dimensional” additive separable solution: w(x. Ibragimov (1994). . C1 y + C3 . t) + 2 ln |x|. t + C4 ) − 2 ln |C1 |. 3◦ . © 2004 by Chapman & Hall/CRC I I . and the function ψ(x. t) = u(r. y. 2◦ . y. utt + kut = 4abeu . where the function u(t) is determined by the autonomous ordinary differential equation (C is an arbitrary constant) ϕtt + kϕt = Ceϕ . C4 are arbitrary constants. .

ξ = βx + µy. ∂t2 ∂t ∂ξ ∂ξ Remark. C2 . Then the functions w1 = w( x + C1 . r2 = A (x + C1 )2 (y + C2 )2 + − (t + C3 )2 . xa sinh λ.4. where β and µ are arbitrary constants. t) is determined by the differential equation ∂U ∂ ∂U ∂2U +k = aβ 2 ew + bµ2 eλw . © 2004 by Chapman & Hall/CRC . t) = V (ξ1 . y. y. w4 = w x. t . t). y. w2 = w x cos β + y w3 = w x cosh λ + ta a/b sin β. Equations with Two Space Variables Involving Arbitrary Functions 2 4. y cosh λ + tb 1/2 −1/2 sinh λ + t cosh λ . F (w) = f (w) dw. Ibragimov (1994). Traveling-wave solution in implicit form: C1 + 2 F (w) 2 2 λ2 − ak1 − bk2 −1/2 dw = k1 x + k2 y + λt + C2 . 4. −1/2 sinh λ. t) = U (ξ.4. ξ2 ). where the βi . 2◦ . and λ are arbitrary constants. t) = u(z. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). yb sinh λ + t cosh λ . A and the expression in square brackets must have like signs. y + C2 . + f (w). −x 1/2 T T T ∂2w =a ∂2w +b ∂2w b/a sin β + y cos β. µi . and λ are arbitrary constants. Suppose w(x. There is a “two-dimensional” solution of the form w(x. C2 . y.4.4. and σi are arbitrary constants. S ¦R z = y|x|−λ . ξ2 = β2 x + µ2 y + σ2 t. where C1 . Solution (C1 . Equations of the Form ∂ w = ∂ f (x) ∂w + ∂ g(y) ∂w +h(w) ∂t2 ∂x ∂x ∂y ∂y 1. “Two-dimensional” solution: w(x. a b Here. There is a more general. and the function U = U (ξ.1. k1 . and the function w(r) is determined by the ordinary differential equation wrr + 2r−1 wr + A−1 f (w) = 0. k2 . “two-dimensional” solution of the form w(x. Reference: N. β. t) + 2 ln |x|. t) is a solution of this equation. 3◦ . C2 . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 305 3◦ . ∂t2 ∂x2 ∂y 2 1◦ . t + C3 ). where C1 . y. 4◦ . H. C3 . ξ1 = β1 x + µ1 y + σ1 t. and C3 are arbitrary constants): w = w(r).

“Two-dimensional” solution for n ≠ 2 and m ≠ 2: w = U (ξ. η) is determined by the equation 1+ 1 C2 ∂2U ∂2U ∂2U ∂U − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) + f (U ) = 0. (2 − n)(2 − m) 2◦ . D. b(2 − m)2 4 2 . 2−n where the function W (y. a) (y. z2 = C3 x + C4 y + λ2 t. a(2 − n)2 b(2 − m)2 4 − nm . Zaitsev (2002). 2−m + f (U ). ξ 2 = 4k y 2−m x2−n + . B1 = where the function V (x. ∂ ∂w ∂ ∂w ∂2w = axn + by m 2 ∂t ∂x ∂x ∂y ∂y 1◦ . © 2004 by Chapman & Hall/CRC . t). “Two-dimensional” solution for n ≠ 2: w = W (y. + f (w). r2 = 4k y 2−m 1 x2−n + − (t + C)2 . y x +√ . 2 2 a(2 − n) b(2 − m) 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation wrr + A w + k −1 f (w) = 0. t) is determined by the differential equation ∂ 2 U B1 ∂U ∂ 2U =k + ∂t2 ∂ξ 2 ξ ∂ξ 3◦ . Polyanin and A. r r A= 2(4 − n − m) . 2. ζ 2 = 4k 1 x2−n − (t + C)2 . Polyanin and V. η). (2 − n)(2 − m) where the function U (ξ. 2 ∂ξ ∂ξ∂η ∂η ∂η (2) Remark. ζ) is determined by the differential equation ∂w ∂ by m ∂y ∂y V ¦U +k ∂ 2 W B3 ∂W + ∂ζ 2 ζ ∂ζ + f (W ) = 0. 5◦ . η). y. ξ= √ aC b η = (C 2 − 1) x2 xy − 2C √ − C 2 t2 . z2 ). t) = u(z1 . “Two-dimensional” solution: w = U (ξ. Solution for n ≠ 2 and m ≠ 2: w = w(r). η 2 = 4k 1 y 2−m − (t + C)2 . Relations (1) and equation (2) can be used to obtain other “two-dimensional” solution by means of the following rename: (x. a(2 − n)2 4 2 . and the function U = U (ξ.306 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 4◦ . η) is determined by the differential equation ∂w ∂ axn ∂x ∂x +k ∂2V B2 ∂V + ∂η 2 η ∂η + f (V ) = 0. B3 = References: A. Zhurov (1998). A. “Two-dimensional” solution for m ≠ 2: w = V (x. There is a “two-dimensional” solution of the form w(x. F. b). B2 = 4◦ . a ab (1) where C is an arbitrary constant (C ≠ 0). D. z1 = C1 x + C2 y + λ1 t. ζ). I.

B= n . aβ 2 bλ2 4 where the function w(r) is determined by the autonomous ordinary differential equation wrr + k −1 f (w) = 0. “Two-dimensional” solution for n ≠ 2: w = W (y. + f (w). Integrating yields its general solution in implicit form: W C1 + 2k −1 where C1 and C2 are arbitrary constants. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 307 ∂ ∂w ∂ ∂w ∂2w = axn + beλy 2 ∂t ∂x ∂x ∂y ∂y 1◦ . Solution for n ≠ 2 and λ ≠ 0: 3. w = w(r). a(2 − n)2 4 2 . wrr + wr + k −1 f (w) = 0. r2 = 4k + f (w). r 2−n ◦ 2 . Solution for β ≠ 0 and λ ≠ 0: w = w(r). 2 a(2 − n) bλ2 where the function U (ξ. ζ). t). ∂2w ∂t2 = ∂ aeβx ∂w +k + ∂ 2 W A ∂W + ∂ζ 2 ζ ∂ζ ∂ beλy ∂w ∂y + f (W ) = 0. ∂η 2 where the function W (y. η) is determined by the differential equation ∂w ∂ axn ∂x ∂x 4◦ .4. η 2 = 4k e−λy 1 − (t + C)2 . t) is determined by the differential equation ∂ 2 U B ∂U ∂2U =k + 2 ∂t ∂ξ 2 ξ ∂ξ 3◦ . ζ 2 = 4k 1 x2−n − (t + C)2 . “Two-dimensional” solution for λ ≠ 0: w = V (x. . η). 2−n +k ∂2V + f (V ) = 0. ξ 2 = 4k e−λy x2−n + . 2−n where the function V (x. © 2004 by Chapman & Hall/CRC −1/2 f (w) dw dw = C2 r. “Two-dimensional” solution for n ≠ 2 and λ ≠ 0: w = U (ξ. A= ∂x ∂x ∂y ◦ 1 . e−λy 1 x2−n + − (t + C)2 .4. 2 a(2 − n) bλ2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 2 A A= . bλ2 4 + f (U ). ζ) is determined by the differential equation ∂w ∂ beλy ∂y ∂y 4. r2 = 4k e−βx e−λy 1 + − (t + C)2 .

F. k1 . Zaitsev (2002). ξ 2 = 4k e−βx e−λy + .3. k2 .6 with g(t) = b and h1 (x) = h2 (y) = 0 and a special case of equation 4. .3. C1 y + C3 . ζ). “Two-dimensional” solution for λ ≠ 0: w = V (x. η) is determined by the differential equation ∂w ∂ aeβx ∂x ∂x 4◦ . D. A.7 with f (x. . ∂ζ 2 References: A. “Two-dimensional” solution for β ≠ 0 and λ ≠ 0: w = U (ξ. ∂y ∂y 1◦ . w2 = w(x cosh λ + ta1/2 sinh λ.2. ∂ ∂w ∂ ∂w ∂2w = f (x) + g(y) + aw ln w + bw. “Two-dimensional” solution for β ≠ 0: w = W (y. aβ 2 bλ2 where the function U (ξ. y) = g(y). Polyanin and A. where C1 . y) = f (x) and g(x. 2 4. C4 and λ are arbitrary constants. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). . ` ` ` ∂2w =a ∂2w + ∂ ∂w where C1 . 5. η).4. aβ 2 4 +k ∂2V + f (V ) = 0. η 2 = 4k e−λy 1 − (t + C)2 . Polyanin and V. © 2004 by Chapman & Hall/CRC . t). y. Then the functions ∂t2 ∂x2 g(w) w1 = w( C1 x + C2 . ∂η 2 where the function W (y. . t) is determined by the differential equation ∂ 2 U 1 ∂U ∂2U =k − 2 ∂t ∂ξ 2 ξ ∂ξ 3◦ . ζ 2 = 4k e−βx 1 − (t + C)2 . 2 ∂t ∂x ∂x ∂y ∂y This is a special case of equation 4. I. 2◦ . C2 .4.4. Suppose w(x. y. t) is a solution of this equation. xa−1/2 sinh λ + t cosh λ). . Zhurov (1998). bλ2 4 + f (U ). and λ are arbitrary constants. Equations of the Form ∂ w = ∂ f (w) ∂w + ∂ g(w) ∂w +h(w) ∂t2 ∂x ∂x ∂y ∂y 1. Traveling-wave solution in implicit form: 2 k2 2 g(w) dw + (ak1 − λ2 )w = C1 (k1 x + k2 y + λt) + C2 . ζ) is determined by the differential equation ∂w ∂ beλy ∂y ∂y Y ¦X +k ∂2W + f (W ) = 0. C1 t + C4 ). D. where the function V (x.308 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 2◦ .

EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 309 3◦ . √ √ 2λ a (y + λt) + (t a x)[g(w) − λ2 ] = ψ(w). and C1 and C2 are determined by (2). z). and Moussiaux.4. y. and λ is an arbitrary constant. t) = v(ζ). y. y. η) is determined by the ﬁrst-order partial differential equation √ ∂U ∂U = ϕ(η). where λ is an arbitrary constant and the function U = U (ξ. t) = u(y. where F (C1 . To the special case ϕ(η) = 0 in (1) there corresponds the second group of solutions speciﬁed in Item 3◦ . where ϕ1 (z). η). © 2004 by Chapman & Hall/CRC b b Φ(η) = C1 . C2 ) is an arbitrary function of two variables. 7◦ . In the general case. ∂y ∂y ζ = (x2 − at2 )y −2 . the characteristic system corresponding to equation (1) has the form (Polyanin. equation (1) is an ordinary differential equation in ξ that can be readily integrated to obtain the ﬁrst group of solutions speciﬁed in Item 3◦ . 2002) b a a a a dξ = g(U ) − λ2 Its independent integrals are given by a a U where ξ g C1 Φ(η) dη 1 √ ϕ(η) dη. t) = U (ξ. ξ = y + λt. The general solution of equation (1) has the form Φ(η) = F (C1 . 1 √ 2λ a b dU dη √ = . 4◦ . η = x t a. C2 ) = 0.2. “Two-dimensional” solutions (generalize the solutions of Item 3 ◦ ): √ w(x. Solutions in implicit form: √ √ g(w) dw = yϕ1 x t a + ϕ2 x t a . where the function u = u(y. 2λ a We ﬁrst calculate the integral in the second relation of (2) and then. Solution: + ∂ ∂u g(u) = 0. “Two-dimensional” solution: w(x. see equation 4. where the function v = v(ζ) is determined by the ordinary differential equation 2aζvζζ + 2avζ + 2ζ 2 [g(v)vζ ]ζ + 3ζg(v)vζ = 0. in the resulting expression. substitute the left-hand side of the ﬁrst relation of (2) for C1 .3 with f (w) = a.4. For other exact solutions. 2 a (2) . ∂ 2 u ∂u + ∂z 2 ∂z z = x2 − at2 . z) is determined by the differential equation 4a z 6◦ . 5◦ . and ψ(w) are arbitrary functions. ϕ2 (z). (1) 2λ a g(U ) − λ2 ∂ξ ∂η with ϕ(η) being an arbitrary function. In the special case λ = 0. Zaitsev. w(x.4. 2λ a ϕ(η) λ √ η = C2 .

. . C4 and β are arbitrary constants. k1 . . w2 = w(x cos β − y sin β. For other exact solutions. ∂x ∂x ∂y ∂y 1◦ . Traveling-wave solution in implicit form: 2 2 k1 f (w) + k2 g(w) dw − λ2 w = C1 (k1 x + k2 y + λt) + C2 . y. 3. and λ are arbitrary constants. where the function U = U (ζ) is determined by the ordinary differential equation 2ζ 2 Uζζ + 3ζUζ = 2[ζf (U )Uζ ]ζ . y. Suppose w(x. t) is a solution of this equation. 2 2 g(w) + C1 f (w) = C2 . 3◦ . . t) is determined by the differential equation ∂u ∂2u 1 ∂ = rf (u) . y. where C1 . . where C1 . t) = u(r.310 2. . t) = U (ζ). Solution: w(x. r= x2 + y 2 . 3◦ .3 with f (w) = g(w). C1 y + C3 . Solutions in implicit form: C1 x + C2 y + C 3 t + C4 C1 y + C 2 t + C 3 x + C4 C1 x + C2 t + C 3 y + C4 where C1 . © 2004 by Chapman & Hall/CRC 2 2 2 = C1 f (w) + C2 g(w). where the function u = u(r. t) is a solution of this equation. is also a solution of the equation (the plus or minus signs in w1 are chosen arbitrarily). are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). Traveling-wave solution in implicit form: 2 2 (k1 + k2 ) c c c f (w) dw − λ2 w = C1 (k1 x + k2 y + λt) + C2 . C1 t + C4 ). where C1 . ζ = (x2 + y 2 )t−2 . C2 . 2◦ . . Then the functions w1 = w( C1 x + C2 . C4 are arbitrary constants. c c c ∂2w = ∂ f (w) ∂w + ∂ ∂w where C1 . y. Suppose w(x. k2 . see equation 4. k1 . k2 . C4 are arbitrary constants. . and λ are arbitrary constants. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂2w = f (w) + f (w) . . 4◦ . . C1 y + C3 . 2 . C1 t + C4 ). . t).2. . Then the function ∂t2 g(w) w1 = w( C1 x + C2 . 2 ∂t ∂x ∂x ∂y ∂y 1◦ . 2◦ . t). 2 2 2 f (w) + C1 g(w) = C2 . x sin β + y cos β. “Two-dimensional” solution with axial symmetry: w(x. ∂t2 r ∂r ∂r 5◦ . .4. C2 .

ξ= (ξ 2 wξ )ξ = [ϕ(w)wξ ]ξ . . t) = U (z. and f g. one obtains a Riccati equation for ξ = ξ(w): 2 2 (2) C5 ξw = ξ 2 − C1 f (w) − C2 g(w). . For C5 ≠ 0. © 2004 by Chapman & Hall/CRC z = ax + by. EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 311 C1 x + C2 y + C 3 . . . Solution: (3) To the special case C5 = 0 there corresponds the second solution of Item 3◦ . see the book by Polyanin and Zaitsev (2003). 5◦ . . which is reduced to a second-order linear equation. 7◦ . η= which admits the ﬁrst integral 2 2 [η 2 f (u) + C1 g(u) − C2 ]uη = C5 . C4 are arbitrary constants and the function u(ξ) is determined by the ordinary differential equation w = w(ξ). ∂ ∂U ∂2U = ϕ(U ) . For exact solutions of equation (2).4. C1 y + C 2 t + C 3 . w = u(η). . x + C4 where C1 .4. 2 2 ϕ(w) = C1 f (w) + C2 g(w). and the function v(ζ) is determined by the ﬁrst-order ordinary differential equation w = v(ζ). . treating w in (1) as the independent variable. y. t) is determined by a differential equation of the form 3. . t + C4 where C1 . 4◦ . ∂t2 ∂z ∂z which can be reduced to a linear equation. For C5 ≠ 0. . ζ= 2 2 [ζ 2 g(v) + C1 f (v) − C2 ]vζ = C5 . (4) For exact solutions of equation (4). see the book by Polyanin and Zaitsev (2003).6: ϕ(U ) = a2 f (U ) + b2 g(U ).4. . C4 are arbitrary constants and the function u(η) is determined by the ordinary differential equation 2 2 C2 uηη = [η 2 f (u)uη ]η + C1 [g(u)uη ]η . 6◦ . there corresponds the third solution of Item 3◦ . “Two-dimensional” solution (a and b are arbitrary constants): w(x. t). . Solution: (1) To the special case C5 = 0 there corresponds the ﬁrst solution of Item 3◦ . The inverse function ζ = ζ(v) is determined by the Riccati equation that can be obtained from (4) with the renaming u → v. one obtains a Riccati equation for η = η(u): 2 2 C5 ηu = η 2 f (u) + C1 g(u) − C2 . which is reduced to a second-order linear equation. treating u in (3) as the independent variable. which admits the ﬁrst integral 2 2 ξ 2 − C1 f (w) − C2 g(w) wξ = C5 . C4 are arbitrary constants. C1 x + C2 t + C 3 . Solution: To the special case C5 = 0. . where the function U = U (z.4. η → ζ. . y + C4 where C1 .

where the function R = R(ξ. 4. C2 . 10◦ . 2◦ . “Two-dimensional” solution: w(x.4. C2 . ξ2 ξ = xt−1 . z = x2 − at2 . “Two-dimensional” solution (a and b are arbitrary constants): w(x. η = A2 x + B2 y + C2 t. are also solutions of the equation (the plus or minus signs are chosen arbitrarily).4. η = yt−1 . y. t + C3 ). f (w)g(w) dw where the functions ϕ(z) and ψ(z) are arbitrary functions. t) = Q(z1 . 12◦ . η). ∂y ∂x ∂η ∂η which can be reduced to a linear equation. where C1 . η) is determined by the differential equation ∂2R ∂2R ∂R ∂R ∂ ∂R ∂ ∂R ∂2R + 2ξη + η 2 2 + 2ξ + 2η = f (R) + g(R) . Suppose w(x. ∂2w ∂ ∂w ∂2w =a + f (w) + g(w). 2 2 h(w) = k2 f (w) + ak1 − λ2 . y. η) is determined by a differential equation of the form 5. There is a “two-dimensional” solution of the form (generalizes the solutions of Items 7 ◦ –9◦ ): w(x. t) = W (y. and λ are arbitrary constants. t) = V (ξ. Traveling-wave solution in implicit form: h(w) C1 − 2 −1/2 d d d h(w)g(w) dw dw = k1 x + k2 y + λt + C2 . see Ibragimov (1994).312 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 8◦ . w2 = w(x cosh λ + ta1/2 sinh λ. t) is a solution of this equation. Solutions in implicit form: d d √ f (w) ϕ x + t a − 2 √ f (w) ϕ x − t a − 2 −1/2 f (w)g(w) dw −1/2 √ dw = ψ x + t a √ dw = ψ x − t a y. y. where C1 .8: ∂V ∂ ∂V ∂ f (V ) + ψ(V ) = 0. w(x.4. y. There are “two-dimensional” solutions of the following forms: w(x. where the function W = W (y.4. 9◦ . η). z2 ). t) = R(ξ. t) = U (y. For results of the group analysis of the original equation. y. η). η = ax + bt. 11◦ . ξ) is determined by a differential equation of the form 5. y. © 2004 by Chapman & Hall/CRC . y. ψ(V ) = a2 g(V ) − b2 . ∂x ∂x ∂ξ ∂ξ which can be reduced to a linear equation.8: ∂W ∂ ∂W ∂ g(W ) + χ(W ) = 0. and λ are arbitrary constants. k1 . ξ = ay + bt. ξ). y + C2 . k2 . where the function V = V (x. 3◦ . Then the functions w1 = w( x + C1 . xa−1/2 sinh λ + t cosh λ). “Two-dimensional” solution (a and b are arbitrary constants): w(x. z2 = a2 x + b2 y + c2 t. ∂t2 ∂x2 ∂y ∂y 1◦ . z). C3 . y. ∂ξ 2 ∂ξ∂η ∂η ∂ξ ∂η ∂ξ ∂ξ ∂η ∂η z1 = a1 x + b1 y + c1 t. t) = V (x. χ(W ) = a2 f (W ) − b2 . ξ = A1 x + B1 y + C1 t. y. 4◦ .

ξ 2 = 4 a(2 − n)2 bλ2 w = w(r). Remark. ζ 2 = 4 2. r 2−n ◦ 2 . 2 b(2 − m) 4 1 x2−n − (t + C)2 . w(x. y). + ∂ ∂w f (w) ∂y ∂y – a2 f (w) + b. There are “two-dimensional” solutions of the following forms: y 2−m x2−n + . 2 a(2 − n) 4 w(x.4. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).4. = axn ∂2w + by m ∂2w + f (w). t) = V (x. y). η2 = 4 e e−λy 1 − (t + C)2 . bλ2 4 1 x2−n − (t + C)2 . r2 = e ∂t2 ∂x2 ∂y 2 ◦ 1 .3. ζ). 2 a(2 − n) 4 w = W (y. Solution for n ≠ 2 and λ ≠ 0: w = V (x.4. wrr + wr + kf (w) = 0. Other Equations 1. where the function U = U (x. t). t) = W (y. ζ 2 = 4 © 2004 by Chapman & Hall/CRC . t). t) = U (ξ. w = U (ξ. k a(2 − n)2 bλ2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 2(3 − n) A A= . wrr + wr + kf (w) = 0. η). y. y) is determined by the Poisson equation ∂2U ∂2U + + b = 0. η2 = 4 e 1 y 2−m − (t + C)2 . ∂2w ∂t2 ∂x2 ∂y 2 ◦ 1 . ∂2w = axn ∂2w + beλy ∂2w + f (w). The above remains true if the constant b in the equation is substituted by an arbitrary function b = b(x. ξ 2 = 4 a(2 − n)2 b(2 − m)2 r2 = e w(x. 2 k a(2 − n) b(2 − m)2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation 1−n 1−m A A=2 + . There are “two-dimensional” solutions of the following forms: e−λy x2−n + . 4. x2−n y 2−m 1 4 + − (t + C)2 . ζ). x2−n e−λy 1 4 + − (t + C)2 . y. f 3 (w) f (w) dw = at + U (x. ∂x2 ∂y 2 For this linear equation. y. η). r 2−n 2−m 2◦ . Solution for n ≠ 2 and m ≠ 2: w = w(r). EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 313 ∂ ∂w ∂2w = f (w) 2 ∂t ∂x ∂x Solution in implicit form: 5.

ψtt = a(β 2 + γ 2 )ϕψ. ζ 2 = 4 4. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂2w ∂2w ∂2w = aeβx + beλy + f (w). ψtt = bϕ − βf (t) + g(t) ψ. w = W (y. β = b/a.314 3. y). w = V (x. see Tikhonov and Samarskii (1990) and Polyanin (2002). ∆≡ ∂2 ∂2 + . There are “two-dimensional” solutions of the following forms: w = U (ξ. ∂x2 ∂y 2 For solutions of this linear equation. y. can be found in Kamke (1977) and Polyanin and Zaitsev (1995. aβ 2 4 a ≠ 0. t) = ϕ(t) + ψ(t)eβx+γy . bλ2 4 e−βx 1 − (t + C)2 . Generalized separable solution: ∂w ∂x 2 + ∂w ∂y 2 + f (t). ∂2w ∂2w ∂2w = aw + –a ∂t2 ∂x2 ∂y 2 1◦ . The solution of the second equation. The solution of the ﬁrst equation is expressed as (C1 and C2 are arbitrary constants) ϕ(t) = t 0 (t − τ )f (τ ) dτ + C1 t + C2 . r2 = 4 e−βx e−λy 1 + − (t + C)2 . y) satisﬁes the two-dimensional Helmholtz equation ∆Θ + βΘ = 0. and the function Θ = Θ(x. where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕtt = bϕ2 + g(t)ϕ + h(t). where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations ϕtt = f (t). which is linear in ψ. y. w(x. ξ2 = 4 η2 = 4 f f e−βx e−λy + . ζ). ∂2w = aw + f (t) ∂2w + ∂2w ∂y 2 ∂t2 ∂x2 Generalized separable solution: + bw2 + g(t)w + h(t). η). 2003) for many f (t). 2◦ . k aβ 2 bλ2 4 where C and k are arbitrary constants (k ≠ 0) and the function w(r) is determined by the ordinary differential equation wrr + 4r−1 wr + kf (w) = 0. t) = ϕ(t) + ψ(t)Θ(x. 5. Solution for β ≠ 0 and λ ≠ 0: w = w(r). t). ∂t2 ∂x2 ∂y 2 1◦ . aβ 2 bλ2 e−λy 1 − (t + C)2 . w(x. © 2004 by Chapman & Hall/CRC .

and the functions ϕ(t). w(x. and χ = χ(t) are determined by the ordinary differential equations [f1 (x)ϕx ]x + aϕ ln ϕ + [h1 (x) + C1 ]ϕ = 0. y) ∂w ∂y + kw ln w. ∂2w ∂t2 = ∂ f1 (x) ∂w + ∂ f2 (y) ∂w ∂y + aw ln w + g(t) + h1 (x) + h2 (y) w. t) = ϕ(x)ψ(y)χ(t). y. y. and the function Θ(x. ψ(t). ψ(t). ∂x ∂x ∂y Multiplicative separable solution: w(x. P (y) = D1 cosh µy + D2 sinh µy. y) ∂w + ∂ g(x. y) + g(x. and µ are related by two constraints. ∂x ∂x ∂y Multiplicative separable solution: w(x. where the functions ϕ = ϕ(x).4. χ(t). B1 . 7. D2 . and C1 and C2 are arbitrary constants. and χ(t) are determined by solving an appropriate system of second-order ordinary differential equations (not written out here).4. ∂2w ∂t2 = ∂ f (x. and η(t) satisfy a system of second-order ordinary differential equations (not written out here). ψ = ψ(y). D1 . C2 . A2 . and µ are arbitrary constants. y) + kΘ ln Θ − AΘ = 0. y) satisﬁes the stationary equation ∂Θ ∂ ∂Θ ∂ f (x. y). 3◦ . t) = ϕ(t)Θ(x. A2 . 6. and the functions ϕ(t). B2 . [f2 (y)ψy ]y + aψ ln ψ + [h2 (y) + C2 ]ψ = 0. where the function ϕ(t) is determined by the ordinary differential equation ϕtt − kϕ ln ϕ − Aϕ = 0. t) = ϕ(t) + ψ(t)(A1 cos µx + A2 sin µx) + χ(t)(B1 cosh µy + B2 sinh µy). B1 . There are generalized separable solutions of the form w(x. where A1 . where F (x) = A1 cos 2µx + A2 sin 2µx. B2 . t) = ϕ(t) + ψ(t)F (x) + χ(t)G(y) + η(t)H(x)P (y). The arbitrary constants A1 . © 2004 by Chapman & Hall/CRC . EQUATIONS WITH TWO SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 315 2◦ . y. t) = ϕ(t) + ψ(t)(A1 cosh µx + A2 sinh µx) + χ(t)(B1 cos µy + B2 sin µy). y. H(x) = C1 cos µx + C2 sin µx. A is an arbitrary constant. ∂x ∂x ∂y ∂y A particular solution of equation (1) is given by (B is an arbitrary constant) ϕ(t) = exp k − 2A k (t + B)2 + . 4 2k −1/2 (1) and the general solution can be written out in implicit form as (B and C are arbitrary constants) g kϕ2 ln ϕ + (A − 1 k)ϕ2 + B 2 dϕ = C t. χtt − aχ ln χ − [g(t) − C1 − C2 ]χ = 0. There are generalized separable solutions of the following forms: w(x. G(y) = B1 cosh 2µy + B2 sinh 2µy. C1 . y.

t) is determined by the linear telegraph equation ∂u ∂2u ∂2u + f (t) = g(t) 2 . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂2w ∂2w ∂2w ∂2w = f1 (x. this equation is linear. + g1 (x. t) = u(x. where the function v(x. “Two-dimensional” solution: w(x. y) ∂x ∂y Multiplicative separable solution: w(x. t) is determined by the differential equation ∂v ∂ 2 v 2(m + 2) ∂2v + f (t) = g(t) 2 + h(t)v m+1 . 2 ∂x ∂x∂y ∂y ∂x ∂y h ∂w ∂2w ∂ ∂w ∂2w + f (t) = g(t) + h(t) wm . if m = −1. where C is an arbitrary constant and the function u(x. t) = v(x. t) exp(Cy) if m ≠ −1. t)|y + C|2/m . where the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕtt − kϕ ln ϕ − s(t) + C ϕ = 0. ∂ 2θ ∂2θ ∂θ ∂θ ∂2θ + f2 (x. y) + g2 (x. y) + f2 (x. λ where C is an arbitrary constant and the function v(x. y. t) = v(x. t) is determined by the linear telegraph equation ∂u ∂2u ∂2u + f (t) = g(t) 2 . 2 ∂t ∂t ∂x m2 For m = −2 and m = −1. 2 ∂t ∂t ∂x 2◦ . 2 2 ∂t ∂t ∂x ∂y ∂y 1◦ . y) + f3 (x. t) + ln |y + C| λ where C is an arbitrary constant and the function u(x. 2 2 ∂t ∂t ∂x ∂y ∂y 1◦ . y). y. y) + (t) w + kw ln w. y) 2 + g1 (x. t)|y + C|1/(m+1) u(x. t) = u(x. y. y) 9. y) satisﬁes the stationary equation f1 (x. y) ∂t2 ∂x2 ∂x∂y ∂y 2 ∂w ∂w + g2 (x. y) − C θ + kθ ln θ = 0. “Two-dimensional” solution: 1 w(x. y. t) = ϕ(t)θ(x. t) + ln |y + C|. ∂v ∂ 2v 2 ∂2v + f (t) = g(t) 2 + h(t)eλv . ∂w ∂2w ∂ ∂w ∂2w + f (t) = g(t) + h(t) eλw . y.316 8. 2 ∂t ∂t ∂x λ © 2004 by Chapman & Hall/CRC . “Two-dimensional” solution: 2 w(x. and the function θ = θ(x. y) + f3 (x. 2 ∂t ∂t ∂x 2◦ . t) is determined by the differential equation 10. y) + h(x. y) + h(x. “Two-dimensional” solution: w(x.

For other exact solutions. µ ≠ 0. ρ2 . z. ρ3 ). µ ρ3 = z + 2 ln |t|. ν 4◦ . y. 2 ρ1 = x|t| n−2 . and k are arbitrary constants. t) = |t| 1−p F (ρ1 .6. Solution for p ≠ 1. Equations of the Form ∂ 2 w = ∂ f (x) ∂w + ∂ g(y) ∂w + ∂ h(z) ∂w + aw p ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z i 1. and p ≠ 1: A w= 2s(p − 1) 1 p−1 p p−1 p−1 p−1 p−1 y 2−m z 2−k 1 x2−n + + − (t + C)2 2 2 a(2 − n) b(2 − m) c(2 − k)2 4 2 2 2 1+p + + + . and ν ≠ 0: s(p − 1)2 (r + C1 )2 w= − 2k(1 + p) p 1 1−p . z.6. λ µ ν where C1 and C2 are arbitrary constants. 2◦ . Solution for n ≠ 2. z. C1 2 t + C2 .1. t) is a solution of this equation. where C1 and C2 are arbitrary constants.5. 2 ρ1 = x + 2 ln |t|. aλ2 bµ2 cν 2 4 where C1 . Then the functions w1 = C1 w C12−n x. 1◦ . 3◦ . There is a “three-dimensional” solution of the form w(x. C1 2 t + C2 . λ ≠ 0.1. 3◦ . ∂2w ∂t2 = ∂ ∂x aeλx ∂w ∂x + ∂ ∂y beµy ∂w ∂y + ∂ ∂z ceνz ∂w ∂z + wp . λ ρ2 = y + 2 ln |t|. ρ2 . C2 . Suppose w(x. There is a “three-dimensional” solution of the form w(x. m ≠ 2. are also solutions of the equation. 2 ρ3 = z|t| k−2 . ρ3 ). y. 2 4◦ . i 1◦ . are also solutions of the equation. y.5. C12−m y.2 with f (w) = sw p . 2. Then the functions p w1 = C1 w x + p−1 1−p 1−p 1−p ln C1 . see equation 4. see equation 4. k ≠ 2. z + ln C1 . Suppose w(x. z. y + ln C1 . 2 ρ2 = y|t| m−2 . r2 = 4k e−λx e−µy e−νz 1 + + − (t + C2 )2 . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 317 4.3 with f (w) = sw p . t) = |t| 1−p F (ρ1 . C12−k z. © 2004 by Chapman & Hall/CRC . ∂2w ∂t2 = ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y + ∂ ∂z cz k ∂w ∂z + wp . t) is a solution of this equation. Equations with Three Space Variables Involving Arbitrary Parameters 4. A= 1−p 2−n 2−m 2−k 1 1−p . For other exact solutions. y.5. 2◦ .1.4.

Suppose w(x. y. see equation 4. µ ν where C1 and C2 are arbitrary constants. and p ≠ 1: A w= 2s(p − 1) 1 p−1 p−1 p−1 p−1 1−p ln C1 . Then the functions r w1 = C1 w C12−n x. Suppose w(x. C1 t + C2 + r 2 2 2 2 ln C1 . y + p−1 p−1 1−p 1−p ln C1 . z. ν where C1 and C2 are arbitrary constants. ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z 1◦ .1. 2 ρ1 = x|t| n−2 . . + wp . Solution for n ≠ 2. t) is a solution of this equation. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂ ∂w ∂2w = axn + by m + ceνz + wp . z + 2◦ . y.318 3. There is a “three-dimensional” solution of the form w(x. ρ3 = z + 2 ln |t|.6. 4◦ . and p ≠ 1: 1 w= 2s(p − 1) 2 1 p−1 1 1−p 1+p 2 + 1−p 2−n e−µy e−νz 1 x2−n + + − (t + C)2 a(2 − n)2 bµ2 cν 2 4 2 3◦ . © 2004 by Chapman & Hall/CRC q . ∂ ∂w ∂ ∂w ∂ ∂w ∂2w = axn + by m + cz k + eλw .1. y. C1 2 t + C2 . z. ν . λ where C1 and C2 are arbitrary constants.5. t) = |t| 1−p F (ρ1 . ρ2 = y + 2 ln |t|. ρ3 ). are also solutions of the equation. Solution for n ≠ 2.4 with f (w) = sw p . Then the functions ∂t2 axn beµy ceνz r ∂2w = ∂ ∂w + ∂ ∂w + ∂ ∂w w1 = C1 w C12−n x. C12−m y. q 4. There is a “three-dimensional” solution of the form w(x. ρ3 ). ρ1 = x|t| n−2 . A= 1−p 2−n 2−m 2 2 3◦ . y. ρ2 . z. ρ2 = y|t| m−2 . For other exact solutions.6. ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . ρ2 . ν 4◦ . Suppose w(x. C1 2 t + C2 . 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . see equation 4. 1 1−p y 2−m e−νz 1 x2−n + + − (t + C)2 2 2 a(2 − n) b(2 − m) cν 2 4 2 2 1+p + + . z + ln C1 . µ ≠ 0. t) is a solution of this equation. z.5 with f (w) = sw p . ν ≠ 0. Then the functions w1 = w C12−n x. t) = |t| 1−p F (ρ1 . m ≠ 2. C12−m y. For other exact solutions. y. ν ≠ 0. C12−k z. t) is a solution of this equation.2. 2◦ . 4. z. µ ρ3 = z + 2 ln |t|. are also solutions of the equation. are also solutions of the equation. Equations of the Form ∂ 2 w = ∂ f (x) ∂w + ∂ g(y) ∂w + ∂ h(z) ∂w + aeλw ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z q 1.

ν 4◦ . t) = F (ρ1 . λ 2 2 . C2 . z − 2 2 2 2 ln C1 . C1 t + C2 + ln C1 . For other exact solutions.5. are also solutions of the equation. y. ρ3 = z|t| k−2 . Solutions for λ ≠ 0.1. if skβ < 0. where C1 . and k are arbitrary constants and r2 = 4k e−λx e−µy e−νz 1 + + − (t + C3 )2 . y. t) is a solution of this equation. ρ2 . t) = − β 1 w(x. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 319 2◦ . t) = − β 1 w(x. 2 4◦ .3 with f (w) = se βw . t) = − β sβ (r + C1 )2 2k sβ sin2 (C1 r + C2 ) ln − 2 2kC1 sβ sinh2 (C1 r + C2 ) ln − 2 2kC1 sβ cosh2 (C1 r + C2 ) ln 2 2kC1 ln − if skβ < 0. s 3.6. ν β where C1 and C2 are arbitrary constants. z. z. For other exact solutions. z − ln C1 . are also solutions of the equation. Solution for n ≠ 2. µ ρ3 = z + 2 ln |t|. β ρ1 = x + 2 ln |t|. z. if skβ > 0. Then the functions ∂t2 axn by m ceνz t ∂2w = ∂ ∂w + ∂ ∂w + ∂ ∂w w1 = w C12−n x. 3◦ . y. and λ ≠ 0: w=− y 2−m z 2−k 1 x2−n 2sλ 1 + + − (t + C)2 ln 2 2 2 λ A a(2 − n) b(2 − m) c(2 − k) 4 2 2 2 + + − 1. There is a “three-dimensional” solution of the form w(x. There is a “three-dimensional” solution of the form w(x. λ µ ν β where C1 and C2 are arbitrary constants. if skβ < 0. Suppose w(x.1. ρ3 ) − 2 ln |t|. t) = − 1 β 1 w(x. λ ρ2 = y + 2 ln |t|. z. t) = F (ρ1 . C12−m y. y − ln C1 . ρ2 . µ ≠ 0. Suppose w(x. s 2. 2◦ . 3◦ . © 2004 by Chapman & Hall/CRC . y. Then the functions t w1 = w x − 2 2 2 2 ln C1 . z. see equation 4. y. y. z. ρ2 = y|t| m−2 . z. z. C1 t + C2 + ln C1 . ∂x ∂x ∂y ∂y ∂z ∂z 1◦ .6.4. k ≠ 2. y. A= 2−n 2−m 2−k 2 ln |t|. ρ3 ) − ρ1 = x|t| n−2 . y. 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . t) is a solution of this equation. ∂ ∂w ∂ ∂w ∂ ∂w ∂2w = aeλx + beµy + ceνz + eβw . see equation 4. m ≠ 2. aλ2 bµ2 cν 2 4 where k and the expression in square brackets must have like signs. + eβw .2 with f (w) = se λw . ν ≠ 0. and β ≠ 0: w(x.

µ 2 ln |t|. ν ≠ 0. C6 . xa1 sinh λ + t cosh λ). v v v v w2 = w(x cos β + y w3 = w(x cosh λ + a1 /a2 sin β. H. are also solutions of the equation. t) is a solution of this equation. z. C1 C2 z + C5 . t). C2 t + C6 ). w4 = w(x. y. Then the functions v w1 = w C12−n x. y. ν 3◦ . and µ are arbitrary constants.6. ρ3 = z + + sw p .3. Suppose w(x. m ≠ 2. Solution for n ≠ 2. ρ3 ) − ρ1 = x|t| n−2 . . . ν 4◦ . y − 2 2 2 ln C1 . t) = F (ρ1 .5. β 2 ln |t|. ∂2w ∂t2 = a1 ∂2w ∂x2 + a2 ∂2w ∂y 2 + a3 ∂ ∂z wk ∂w ∂z . + c ∂ w k ∂w ∂z ∂z 1◦ . y. z − ln C1 . Solution for n ≠ 2.1. µ ν β where C1 and C2 are arbitrary constants. β. . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). t) is a solution of this equation. Then the functions k −2 w1 = C1 w( C2 x + C3 . 3◦ . β a(2 − n)2 bµ2 cν 2 4 β 2sβ(2 − n) 2 ln |t|. see equation 4. z. © 2004 by Chapman & Hall/CRC u + ∂ beµy ∂w + ∂ ceνz ∂w + eβw . There is a “three-dimensional” solution of the form w(x. ya2 where C1 .320 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 2◦ . z.5 with f (w) = se βw . z. x ¦w Reference: N. and β ≠ 0: w=− x2−n e−µy e−νz 1 1 n 1 ln + + − (t + C)2 + ln . λ.4 with f (w) = se βw . Equations of the Form ∂ 2 w = a ∂ w n ∂w + b ∂ w m ∂w ∂t2 ∂x ∂x ∂y ∂y 1. β 2 2 . There is a “three-dimensional” solution of the form w(x. −x 1/2 a2 /a1 sin β + y cos β.6. see equation 4.1. ρ2 = y|t| m−2 . and β ≠ 0: w=− 2sβ x2−n y 2−m e−νz 1 1 ln + + − (t + C)2 2 2 β A a(2 − n) b(2 − m) cν 2 4 2 2 + − 1. ν ≠ 0. C2 y + C4 . ρ2 . y cosh µ + ta2 sinh µ. Suppose w(x. For other exact solutions. C1 t + C2 + ln C1 . ρ3 ) − ρ1 = x|t| n−2 . ∂2w ∂t2 = ∂ ∂x axn ∂w ∂x 2 ∂y ∂y ∂z ∂z 1◦ . 2 ρ2 = y + 4◦ . −1/2 −1/2 1/2 ta1 sinh λ. z. A= 2−n 2−m 2 ln |t|. z. 4. For other exact solutions. 4. µ ≠ 0. z. . 2◦ . y. y. Ibragimov (1994). sinh µ + t cosh µ). t) = F (ρ1 . ρ2 . ρ3 = z + 2 ln |t|.

6 with n = m = 0. ξ = C 1 x + C2 y t 2 2 a 1 C1 + a 2 C2 . 2/k 2λ −1 −1 −kλ−1 y y 1 −1/2 2λ 2 2 a1 C1 + a2 C2 z + λt C1 x + C2 y t 2 2 a 1 C1 + a 2 C2 r = a 2 x + a1 y ξ = x − a1 t 2 2 2 2 2 w(x. © 2004 by Chapman & Hall/CRC y 1 k+1 . t) = A|z| k+1 sin λ1 x + C1 ) sin λ2 y + C2 ) sin γt + C3 ). y. “Three-dimensional” solution (generalizes the ﬁrst four solutions of Item 2 ◦ ): w(x. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 321 2◦ . y. ρ = t−1 2 a2 x2 + a1 y 2 . z. Solutions: w(x. y. 3◦ . . z. t) = |z| k+1 u(x. z. . y. t) = |z| k+1 Ax2 + By 2 + (a1 A + a2 B)t2 + C1 xy + C2 xt + C3 yt + C4 x + C5 y + C 6 t + C 7 . C1 . y. y. y. . w(x. y. z. C7 . y = a2 y. “two-dimensional” solution. x = a1 x. z. ∂t2 ∂x2 ∂y 2 For this equation.3. Solutions in implicit form: where Φ(w) is an arbitrary function. z. z. . y. y. p = a2 x + a1 y − a1 a2 t where λ is an arbitrary constant. There are solutions of the following forms: w(x. z. 2 1/k 2 2 2 z + C5 C3 − a 1 C1 − a 2 C2 a3 C1 x + C2 y + C 3 t + C 4 where A. t) = |t|2λ V (ρ. γ= a 1 λ2 + a 2 λ2 . z. λ1 . “three-dimensional” solution. . ζ). C1 .4. t) = w(x. t) = U (ξ. t) = w(x. 5◦ . 2 2 a1 C3 − a 2 C1 − 2 2 a 3 C2 a 3 C2 2 2 a2 C3 − a 1 C1 − 2 2 a 3 C2 a 3 C2 C1 y + C 2 z + C 3 t + C 4 x + C5 C1 x + C2 z + C 3 t + C 4 y + C5 2 1/k 2 1/k . t) = |z| Q(p). η = yt . t). B. “two-dimensional” solution. y. z. z. t) = w(x. z. t) = A|z| k+1 exp λ1 x + λ2 y 1 γt + B. y. C2 .5. see equation 4. . ζ = z|t|−kλ−1 2 2 −2 2 2 2 ζ = a 2 x + a1 y − a1 a2 t 2 2 η = (a2 x + a1 y − a1 a2 t )z w(x. ζ = z|t| w(x. w(x. ϕ(ξ) and ψ(ξ) are arbitrary functions. z). z. t) = R(η). 1 2 γ= 1/k w(x. “three-dimensional” solution. y. 6◦ . ζ). t) = |z|2/k F (x. z. t) w(x. z. “one-dimensional” solution. t). y. 4◦ . 1 y 1 1 w(x. 1 2 − 2 2 a 1 C1 + a 2 C2 2 a 3 C3 . y. t) = |t| G(ξ. t) = |z| k+1 A(a2 x2 + a1 y 2 − a1 a2 t2 )−1/2 + B . w(x. and λ2 are arbitrary constants. w(x. and λ are arbitrary constants. z. y. η. −1/2 a3 wk − λ2 = Φ(w). y. y. “one-dimensional” solution. y. z. t) = 1 2 a 3 C3 C1 x + C2 y + C 3 z + C 4 t + C5 2 a 1 λ2 + a 2 λ2 . y. t) = H(r. t) is determined by the linear wave equation ∂2u ∂2u ∂2u = + . Solutions: w = zϕ(ξ) + ψ(ξ) where C1 and C2 are arbitrary constants. t) = W (ζ. “three-dimensional” solution. w(x. where the function u = u(x. z. ξ = xt . y. “three-dimensional” solution. For other exact solutions. . z). 7◦ . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002).5.

2 2 ∂t ∂x ∂y ∂y ∂z ∂z 1◦ . w2 = w x. z. ∂y 2 ∂z 2 For this linear equation. . z. ζ) is determined by the Laplace equation ∂2v ∂2v + = 0. and λ are arbitrary constants. y. y. ∂y 2 ∂z 2 which is implicitly independent of the cyclic variable ζ (the constants of integration that appear in the solution are arbitrary functions of ζ). 5◦ . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). . t) is a solution of this equation. t) = R(x. Solution: 1 √ √ w(x. Solutions: w(x. z). . Then the functions k k −2 w1 = C1 w( C2 x + C3 . z. y. −1/2 y. y. t . C1 C2 z + C5 . z. t)Q(y. . 1/k . y. z = a3 −1/2 z. Ibragimov (1994). ζ=x = AQ. Suppose w(x. −1/2 w3 = w x cosh λ + ta1 sinh λ. Multiplicative separable solution (generalizes the solution of Item 3 ◦ ): w(x. C5 are arbitrary constants. where the functions R = R(x. . z. z. z. z). C6 . C2 t + C6 ). . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂2w ∂ ∂ ∂w ∂w ∂2w = a1 + a2 wk + a3 wk . y cos β + z 1/2 a2 /a3 sin β. z = a3 −1/2 z. 1/k . y = a2 −1/2 y. √ t a1 . 4◦ . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). xa1 sinh λ + t cosh λ . t) = w(x. z. y = a2 where ϕ(ρ1 ) and ψ(ρ2 ) are arbitrary functions and the function u(y. z) are determined by the differential equations ∂2R ∂2R = a1 2 + ARk+1 . y. t) = ϕ(x + t a1 ) + ψ(x − t a1 ) u k+1 (y. . where C1 . . t) and Q = Q(y. y. t) = w(x. . 3◦ . C1 C2 y + C4 . . “Three-dimensional” solutions: where the function v(y. ζ) 1 k+1 . β. −y a3 /a2 sin β + z cos β. H. z. 2 ∂t ∂x ∂Q ∂ ∂Q ∂ Qk + a3 Qk a2 ∂y ∂y ∂z ∂z and A is an arbitrary constant.322 2. t) = . © 2004 by Chapman & Hall/CRC w = v(y. y. ¦ Reference: N. z) is determined by the Laplace equation ∂ 2 u ∂ 2u + = 0. 1/k 2 2 (C3 − a1 C1 )(y + C5 )2 2 a2 (C1 x + C2 z + C3 t + C4 )2 + a3 C2 (y + C5 )2 2 2 (C3 − a1 C1 )(z + C5 )2 2 a3 (C1 x + C2 y + C3 t + C4 )2 + a2 C2 (z + C5 )2 where C1 . ◦ 2 (C1 x + C2 y + C3 z + C4 )2 − a1 C1 (t + C5 )2 2 + a C 2 )(t + C )2 (a2 C2 3 3 5 2 C3 (x + C5 )2 − a1 (C1 y + C2 z + C3 t + C4 )2 2 2 (a2 C1 + a3 C2 )(x + C5 )2 1/k 2 . t) = w(x. z.

r = x 2 − a 1 t2 . t). t) = t2λ F (ξ. .3.5. z. C1 C2 z + C5 . where C1 . y. 3◦ . z. z. y. 7◦ . where λ is an arbitrary constant. t) = G(r. y. xa1 1/2 −1/2 sinh λ + t cosh λ). t) = G(x.4.5. ξ = xt−1 . q). t) = Θ(η). z. Then the functions m k −2 w1 = C1 w( C2 x + C3 . For other exact solutions. y. z. η. where ϕ(ρ1 ) and ψ(ρ2 ) are arbitrary functions and C1 and C2 are arbitrary constants. t) = y 2/m V (r. y. η. z. ζ). z. t). y. ρ = t−kλ−1 r = a 3 y + a2 z . y. ξ = xt−1 .6 with n = 0 and m = k. see equation 4.6 with n = 0. w(x. © 2004 by Chapman & Hall/CRC . ξ = xt−1 . ρ). s = zy −k/m . C1 C2 y + C4 . y. where λ is an arbitrary constant. s = zy −k/m . w(x. ∂2w ∂ ∂w ∂2w = a1 + a2 wm 2 2 ∂t ∂x ∂y ∂y + a3 ∂ ∂w wk . p = (x2 − a1 t2 )y −2 . t) = U (p. 3. y. q = zy −1 . 5◦ . s. z. s). z. There are solutions of the following forms: w(x. 2◦ . . t) = W (p. H. z. z. t) = t2λ F (ξ. t) = ϕ(x + t a1 ) + ψ(x − t a1 ) |y + C1 | m+1 |z + C2 | k+1 . z). η = yt−kλ−1 . 4◦ . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 323 6◦ . ξ). Ibragimov (1994). y. r = x 2 − a 1 t2 . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). t) = U (ξ. “two-dimensional” solution. w(x. η = yt−mλ−1 . ξ = x − a1 t q = xt 2 2 −1 w(x. t) = H(ξ. Multiplicative separable solution: 1 1 √ √ w(x. z. t) is a solution of this equation. ∂z ∂z 1◦ . “two-dimensional” solution. w(x. . w(x. C6 and λ are arbitrary constants. C2 t + C6 ). z. ζ = zt−kλ−1 .5. y. see equation 4. “one-dimensional” solution. z. For other exact solutions. Suppose w(x. w(x. −1 η = (a3 y + a2 z )(x − a1 t ) where λ is an arbitrary constant. t) = V (r. z). p = (a3 y + a2 z )t .3. y. w(x. y. There are “three-dimensional” solutions of the following forms: w(x. ζ). y. ξ = x 2 − a 1 t2 . y. a 3 y 2 + a2 z 2 2 2 “two-dimensional” solution. . w2 = w(x cosh λ + ta1 sinh λ. y. y. 2 2 2 2 2 2 −2 r = a 3 y 2 + a2 z 2 . z. t) = y 2/m H(x. q). ¦ Reference: N. 8◦ . r. There are “two-dimensional” solutions of the following forms: w(x. There are “three-dimensional” solutions of the following forms: w(x. ζ = zt−kλ−1 .

y. C1 C2 z + C5 . z. z. y. −x a2 /a1 sin β + y cos β. ζ). y = a2 −1/2 where the function Θ = Θ(x. . Suppose w(x. y. x = a1 −1/2 x. y. ξ = x|t|−kλ−1 . η. z. z. t) = w(x. . C1 C2 y + C4 . C2 t + C6 ). t) = H(ρ. y. t) = w(x.6 with n = m = k. 1/k . z = a3 −1/2 z. ρ = a 2 x2 + a 1 y 2 . 6◦ . w(x. . . t). y. t) = V (χ). t) = (C1 t + C2 ) w(x. z. 4◦ . t). r = a 2 a 3 x2 + a 1 a 3 y 2 + a 1 a 2 z 2 χ = (a2 a3 x + a1 a3 y + a1 a2 z )t 2 2 2 −2 p = xz −1 . t) = |t|2λ F (ξ. 1/k 2 1/k . t) = C3 a 2 a 3 x2 + a 1 a 3 y 2 + a 1 a 2 z 2 + C4 1 k+1 1 k+1 1 2 2 2 a 1 C1 + a 2 C2 + a 3 C3 C1 x + C2 y + C 3 z + C 4 t + C5 2 C3 (x + C5 )2 2 2 a1 (C1 y + C2 z + C3 t + C4 )2 + (a2 C1 + a3 C2 )(x + C5 )2 2 C3 (y + C5 )2 2 2 a2 (C1 x + C2 z + C3 t + C4 )2 + (a1 C1 + a3 C2 )(y + C5 )2 2 C3 (z + C5 )2 2 2 a3 (C1 x + C2 y + C3 t + C4 )2 + (a1 C1 + a2 C2 )(z + C5 )2 where C1 . y. y. z. y. . For other exact solutions. . w(x. z. q. y. t) = |z|2/k G(p. “one-dimensional” solution. 2◦ . y. where λ is an arbitrary constant. where C1 . “two-dimensional” solution. . η = y|t|−kλ−1 . see equation 4. z. H. ζ = z|t|−kλ−1 . 5◦ . 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂ ∂ ∂w ∂w ∂w ∂2w = a1 wk + a2 wk + a3 wk . C6 and β are arbitrary constants. t) = U (r. There are “three-dimensional” solutions of the following forms: w(x. z. ∂x2 ∂y 2 ∂z 2 For this linear equation. . . ¦ Reference: N. 3◦ . z. z) 1 k+1 . t . z. Solutions: w(x.3.5. Ibragimov (1994). t) = w(x. y. y. z. Multiplicative separable solution: w(x. w(x. Then the functions k k k −2 w1 = C1 w( C1 C2 x + C3 . w2 = w x cos β + y a1 /a2 sin β. z. q = yz −1 . y. . y. © 2004 by Chapman & Hall/CRC . C5 are arbitrary constants. z.324 4. There are solutions of the following forms: w(x. z. 1/k . t). t) = (C1 t + C2 ) a3 C3 x2 + a3 C4 y 2 − (a1 C3 + a2 C4 )z 2 + C5 w(x. t) = (C1 t + C2 ) Θ(x. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). t) is a solution of this equation. y. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). z) is determined by the Laplace equation ∂ 2 Θ ∂ 2 Θ ∂ 2Θ + + = 0.

3. . C6 are arbitrary constants. ζ). z. w2 = w x cos β + y a1 /a2 sin β. and λ are arbitrary constants. ∂w w(x. y.6 with m = n. . t). C2 t + C6 ). Then the functions 5. g(w) = a2 wm . ζ = zt−kλ−1 . ◦ 2 . see equation 4.2. w(x. C1 C2 y + C4 . w(x. 6. C2 t + C6 ). ∂x ∂x ∂y ∂y ∂z ∂z This is a special case of equation 4. χ = (a2 x2 + a1 y 2 )t2n/k−2 z −2n/k 4 . 2 2 2 + a 1 C1 w n + a 2 C2 w m = C3 .5. z. Suppose w(x. There are “three-dimensional” solutions of the following forms: w(x. z. z. © 2004 by Chapman & Hall/CRC 2 ¦ . t) = G(r. t) = U (r. y. “one-dimensional” solution. H. z. . 3◦ . ¦ Reference: N.6. . ∂2w ∂t2 = a1 ∂ wn ∂w + a2 ∂ wm ∂w + a3 ∂ . Then the functions n m k −2 w1 = C1 w( C1 C2 x + C3 . . t . b3 . t) is a solution of the equation in question. where C1 . C1 C2 z + C5 . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). where λ is an arbitrary constant. H. y. s = zt−1 “two-dimensional” solution.6 with f (w) = a1 wn . C1 y + C 2 z + C 3 t + C 4 x + C5 C1 x + C2 z + C 3 t + C 4 y + C5 2 C1 x + C2 y + C 3 t + C 4 z + C5 where C1 . w(x. p = xz −n/k . s). C2 . η = yt−nλ−1 . t) = t2λ F (ξ. 2 2 2 + a 1 C1 w n + a 3 C2 w k = C3 . t) is a solution of this equation. . y. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 325 ∂ ∂ ∂ ∂w ∂w ∂w ∂2w = a1 wn + a2 wn + a3 wk . Ibragimov (1994). t). 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . z. C1 C2 y + C4 . z. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). z. Ibragimov (1994). ◦ 3◦ . C1 C2 z + C5 . −x a2 /a1 sin β + y cos β. Solutions in implicit form: C1 x + C2 y + C 3 z + C 4 t + C5 a1 w n a2 w m a3 w k 2 2 2 2 = a 1 C1 w n + a 2 C2 w m + a 3 C3 w k . b1 .4. . y. n n k −2 w1 = C1 w( C1 C2 x + C3 . Suppose w(x. Multiplicative separable solution: w(x. y. r = a 2 x2 + a 1 y 2 . 4◦ . C5 are arbitrary constants. q. z. For other exact solutions. . y.5. b2 . There are solutions of the following forms: r = (a2 x2 + a1 y 2 )t−2 . Reference: N. y. n+1 m+1 k+1 where C1 . ◦ 1 1 1 2 . where C1 . t) = (C1 t + C2 )|x + C3 | n+1 |y + C4 | m+1 |z + C5 | k+1 . t) = t−2/k z 2/k V (χ). q = yz −n/k . wk 1◦ . . Traveling-wave solution in implicit form: a2 b2 m+1 a3 b2 k+1 a1 b2 n+1 2 3 1 w + w + w − λ2 w = C1 (b1 x + b2 y + b3 z + λt) + C2 . z. . . t) = z 2/k H(p. η. and h(w) = a3 wk . C6 and β are arbitrary constants. ξ = xt−nλ−1 . 2 2 2 2 + a 2 C1 w m + a 3 C2 w k = C3 . .

9◦ . a3. z. y. Then the functions r = yx−m/n . ∂w 7. z. Suppose w(x. ∂w ∂ χ = x−k/n zt(k−n)/n . 1 2 3 ∂t2 ∂ξ ∂ξ which can be reduced to a linear equation. C 1 y + C3 . 8◦ . a2 .4. a1 . ∂2w ∂t2 = a1 ∂ wn ∂w + a2 ∂ wm + a3 wk + bwp . w1 = C1 w( C1 where C1 . η = yt p−m−1 1−p . w(x. q). y. 2◦ .8: ∂U ∂ ∂U ∂ Φ(U ) + Ψ(U ) = 0. η) is determined by a differential equation of the form 5. η).4. k) ←− (y. p−n−1 p−m−1 p−k−1 p−1 2 x + C2 . as shown in Item 6◦ . t) = U (ζ. © 2004 by Chapman & Hall/CRC . “Two-dimensional” solution (b1 . . where the bn and cn are arbitrary constants. ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . y.4. q = c1 x + c2 y + c3 z + c4 t. y. . y. w(x. where the function U = U (ζ. where λ is an arbitrary constant.4. t) is determined by a differential equation of the form 3. c2 . “Two-dimensional” solution (the bn and cn are arbitrary constants): ◦ w(x. s. η = yt−mλ−1 . ρ) is determined by a differential equation of the form 5. ζ). (2) a1 2 3 1 ∂x ∂x ∂η ∂η which can be reduced to a linear equation. Remark. z. . t) = x2/n G(r. t).4. t) = v(x. η = c1 t + c2 y + c3 z. n) (z. η. t) = H(p. There are “three-dimensional” solutions of the following forms: w(x. ζ = zt−kλ−1 . ρ = x−m/n yt(m−n)/n . where the function u = u(ξ. t) is a solution of this equation. t) = t2λ F (ξ. (1) where the function v = v(x. η. C1 t + C5 ). . C5 are arbitrary constants. b2 . z. There is a “three-dimensional” solution of the form w(x. Formula (1) and equation (2) can be used to obtain two other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. 2 ξ = xt p−n−1 1−p . p = b1 x + b2 y + b3 z + b4 t. 2 1 1 2 ∂ζ ∂ζ ∂ρ ∂ρ which can be reduced to a linear equation. ζ = b1 t + b2 x. and c3 are arbitrary constants): w(x. ξ = xt−nλ−1 . and b3 are arbitrary constants): w(x. z. t) = t 1−p U (ξ. C1 z + C4 . ρ = c1 y + c2 z. χ). s = zx−k/n .326 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 5◦ . ζ = zt p−k−1 1−p . are also solutions of the equation (the plus or minus signs are chosen arbitrarily).4. y. ξ = b1 x + b2 y + b3 z. y. m) 7 . ζ). z.8: ∂ ∂v ∂v ∂ vn + a 2 c2 v m + a 3 c2 v k − c2 = 0. Ψ(U ) = a2 c2 U m + a3 c2 U k . The solution speciﬁed in Item 7◦ can be used to obtain other “two-dimensional” solutions by means of cyclic permutations of variables and determining parameters. t) = t−2/n x2/n U (ρ. z. ρ). z.6: ∂ ∂u ∂2u = (a1 b2 un + a2 b2 um + a3 b2 uk ) . Φ(U ) = a1 b2 U n − b2 . y. There are “two-dimensional” solutions of the following forms: w(x. t). y. 6◦ . “Two-dimensional” solution (c1 . t) = u(ξ. z.

C6 . y. 2 a 3 C3 2 2 2 a1 C3 − a 2 C1 − 2 2 a 3 C2 a 3 C2 2 2 a2 C3 − a 1 C1 − 2 2 a 3 C2 a 3 C2 2 2 2 C3 − a 1 C1 − a 2 C2 a3 C1 y + C 2 z + C 3 t + C 4 x + C5 C1 x + C2 z + C 3 t + C 4 y + C5 z + C5 C1 x + C2 y + C 3 t + C 4 where the Cn are arbitrary constants. y. β. y. . C1 t + C6 ) − 2 ln |C2 |. y. y. y = a2 −1/2 y. 1 2 γ= a 1 λ2 + a 2 λ2 . and µ are arbitrary constants. y. t) = u(x. 1 2 . z. y. t) = C1 exp λ1 x + λ2 y γt + C2 + ln |z|. z. x = a1 −1/2 w = ln |zϕ(ξ) + ψ(ξ)|. t) + ln |z|. γ= w(x. −1/2 ya2 where C1 . z. t) = ln w(x. Equations of the Form ∂ 2 w = a ∂ eλ1 w ∂w + b ∂ eλ2 w ∂w + c ∂ eλ3 w ∂w + seβw ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z 1. t) = ln w(x. Then the functions w1 = w( C1 x + C3 . y cosh µ + a1 /a2 sin β. t) = C1 sin λ1 x + C2 ) sin λ2 y + C3 ) sin γt + C4 ) + ln |z|. z. λ. “Three-dimensional” solution (generalizes the ﬁrst four solutions of Item 2 ◦ ): w(x. w(x. and ϕ(ξ) and ψ(ξ) are arbitrary functions. w(x. 2◦ . x. Solutions: where C1 and C2 are arbitrary constants. w(x. y. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). see Tikhonov and Samarskii (1990) and Polyanin (2002). t) is determined by the linear wave equation ∂2u ∂2u ∂2u = + . y. Suppose w(x. Solutions: w(x. © 2004 by Chapman & Hall/CRC a 1 λ2 + a 2 λ2 . . y. 4◦ . C1 C2 z + C5 . z. t) = C1 x2 + C2 y 2 + (a1 C1 + a2 C2 )t2 + C3 xy + C4 xt + C5 yt + C6 x + C7 y + C8 t + C9 + ln |z|. sinh λ + t cosh λ . y. t) = ln 1 2 a 3 C3 C1 x + C2 y + C 3 z + C 4 t + C5 2 − 2 2 a 1 C1 + a 2 C2 . w2 = w x cos β + y w4 = w x. t) = ln w(x. z. t .5. . Ibragimov (1994). 2 2 2 ∂t ∂x ∂y ∂z ∂z 1◦ . . 2 . z. y. y. H. 3◦ . xa1 sinh µ. t 2 2 a 1 C1 + a 2 C2 . sinh µ + t cosh µ . 2 . ξ = C 1 x + C2 y where the function u = u(x. ∂2w ∂2w ∂ ∂w ∂2w = a1 + a2 + a3 ew . −1/2 w3 = w x cosh λ + ta1 sinh λ. t) is a solution of this equation.4. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 327 4. z. . z. C1 y + C4 . z. z.5. z. z.4. −x 1/2 1/2 ta2 a2 /a1 sin β + y cos β. t) = C1 (a2 x2 + a1 y 2 − a1 a2 t2 )−1/2 + C2 + ln |z|. ∂t2 ∂x2 ∂y 2 For this equation. ¦ Reference: N.

ζ2 . x = x a1 leads to a solvable equation of the form 5.2. and h(w) = a3 ew . z. w(x. w(x. and C1 . t) + 2 ln |z|. z). t) + 2 ln |z|. ρ = x 2 − a 1 t2 . “Two-dimensional” solution: w(x. k2 . t) is determined by a solvable equation of the form 3. t) is determined by the equation 2 2 (C1 − a2 C2 ) ∂2v ∂2v = a1 2 + 2a3 ev . z. w(x. t) = v(x. © 2004 by Chapman & Hall/CRC . ξ2 . where Φ(w) is an arbitrary function. ∂η 2 ∂x (2) 2 2 For σ = C1 − a2 C2 > 0. t) = U (r. 9◦ . z. r = a 2 x2 + a 1 y 2 ρ = x − a1 t 2 2 2 2 2 −1 “two-dimensional” solution. “Two-dimensional” solution: w(x. r = a 2 x2 + a 1 y 2 . z. g(w) = a2 .1. and k3 are arbitrary constants. z. z). w(x. z. y. (1) where C1 and C2 are arbitrary constants and the function v = v(η. z. η3 ) + 2 ln |z|. There are solutions of the following forms: w(x.1: 2 ∂2U 2 2 ∂ U = (a1 C1 + a2 C2 ) 2 + 2a3 eU . and λ are arbitrary constants. w(x. η2 = x + k2 ln |z|. y. “one-dimensional” solution. η2 . w(x. where C1 and C2 are arbitrary constants and the function U = U (ξ.1: ∂2v ∂2v + = −2a3 ev . a1 ) (y. w(x.2. C2 . For other exact solutions. t) = T (χ). y. Relations (1) and equation (2) can be used to obtain another “two-dimensional” solution by means of the following renaming: (x. z. η = C1 y + C2 t. y. t) = V (ρ. t) = F (x. y. There are “three-dimensional” solutions of the following forms: w(x. k1 .6.6 with f (w) = a 1 . y. y. ξ2 = yt 2 2 −2 θ = a 2 x + a1 y − a1 a2 t χ = (a2 x + a1 y − a1 a2 t )z 10◦ +. 8◦ . z. t) = W (θ. y. ζ1 = tez . t) = S(θ) + 2 ln |z|. “two-dimensional” solution. ζ2 = xez . Solutions in implicit form: 2λ 2 2 a1 C1 + a2 C2 z + λt C1 x + C2 y t 2 2 a 1 C1 + a 2 C2 a3 ew − λ2 = Φ(w).328 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 5◦ . ξ = C1 x + C2 y. y. t). on dividing by σ. ∂x2 ∂η 2 Remark. y. η3 = y + k3 ln |z|. y. z. “two-dimensional” solution.2. w(x. y.1. z.2. where k. y) + 2 ln |z|. z. z. η1 = t + k1 ln |z|. y. ξ3 ) − 2k ln |t|. t) = G(ξ1 . t) + 2 ln |z|. ξ2 = yt−1 . ζ3 ) + 2z. see equation 4. one obtains a solvable equation of the form 3. y. ξ3 = z|t|k−1 . y. w(x. “one-dimensional” solution.1. t) = U (ξ. t) = H(η1 . t) = E(ζ1 . 7◦ . ζ3 = yez . t) = Q(ρ. y. 2 θ = a 2 x + a1 y − a1 a2 t −1 2 2 w(x.1. “two-dimensional” solution. a2 ). t) = P (r. z. the transformation η = η |σ|. η) + 2 ln |z|. t) = R(ξ1 . ξ1 = xt−1 . ξ2 ) + 2 ln |z/t|. For √ √ 2 2 σ = C1 − a2 C2 < 0. 2 ∂t ∂ξ 6◦ . z. 2 ξ1 = xt .

Solution (generalizes the ﬁrst three solutions of Item 2◦ ): √ √ w(x. w(x. ζ =x √ t a1 . 2 2 ∂t ∂x ∂y ∂y ∂z ∂z 1◦ . z. t) = ln w(x. t) + ln Q(y. z. y cos β + z w3 = w x cosh λ + a2 /a3 sin β. 1/2 ta1 −1/2 z. y. . z. see Tikhonov and Samarskii (1990) and Polyanin (2002). where the function v(y. t) = ln 2 (C1 x + C2 y + C3 z + C4 )2 − a1 C1 (t + C5 )2 . y. y. C5 are arbitrary constants. where ϕ(ρ1 ) and ψ(ρ2 ) are arbitrary functions and the function u(y. t) = ln w(x. 2 2 (a2 C1 + a3 C2 )(x + C5 )2 2 2 (C3 − a1 C1 )(y + C5 )2 . z. C6 . + t cosh λ . t) = ϕ(x + t a1 ) + ψ(x − t a1 ) + ln(a3 C1 y 2 + C2 yz − a2 C1 z 2 + C3 y + C4 z + C5 ). z). . 5◦ . w2 = w x. C1 t + C6 ) − 2 ln |C2 |. z.4. z = a3 z. . z. © 2004 by Chapman & Hall/CRC −1/2 −1/2 −1/2 x = a1 x. z. z) is determined by the Laplace equation ∂ 2 u ∂ 2u + = 0. . t) = ϕ(x + t a1 ) + ψ(x − t a1 ) + ln u(y. y = a2 −1/2 y. y = a2 −1/2 y. −y sinh λ. y = a2 y. 2◦ . (1) ∂y 2 ∂z 2 For this linear equation. z. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 329 ∂2w ∂ ∂ ∂w ∂w ∂2w = a1 + a2 ew + a3 ew . . H. y.5. z. t . ζ) is determined by the Laplace equation ∂2v ∂2v + = 0. t) = ln w(x. C1 C2 y + C4 . Additive separable solution (generalizes the solution of Item 3 ◦ ): w(x. and λ are arbitrary constants. Suppose w(x. y. y. xa1 sinh λ where C1 . y. √ √ √ √ w(x. 2 a3 (C1 x + C2 y + C3 t + C4 )2 + a2 C2 (z + C5 )2 where ϕ(ρ1 ) and ψ(ρ2 ) are arbitrary functions and C1 . ∂y 2 ∂z 2 which is implicitly independent of the cyclic variable ζ (the constants of integration that appear in the solution will be arbitrary functions of ζ). z). 2 2 (a2 C2 + a3 C3 )(t + C5 )2 2 C3 (x + C5 )2 − a1 (C1 y + C2 z + C3 t + C4 )2 . a3 /a2 sin β + z cos β. t) is a solution of this equation. y. ζ) . t) = ϕ(x + t a1 ) + ψ(x − t a1 ) + ln C1 exp(C2 a2 z) sin(C2 a3 y + C3 ) + C4 . Ibragimov (1994). t) = ϕ(x + t a1 ) + ψ(x − t a1 ) + ln C1 exp(C2 a3 y) sin(C2 a2 z + C3 ) + C4 . . Solutions: √ √ w(x. z. 3◦ . C1 C2 z + C5 . y. z = a3 −1/2 z. √ √ √ √ w(x. y. 4◦ . . . β. ¦ Reference: N. y. 2 a2 (C1 x + C2 z + C3 t + C4 )2 + a3 C2 (y + C5 )2 2 2 (C3 − a1 C1 )(z + C5 )2 . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). Then the functions 2. w1 = w( C1 x + C3 . t) = R(x. . z. z = a3 −1/2 z. z. “Three-dimensional” solutions: w = ln v(y.

y 2 ln . t) = F (ρ1 . see equation 4. “one-dimensional” solution.6. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). y. q = zy −1 2 2 −k ξ = xt−1 . w(x. z. . y. ρ). C1 C2 z + C5 . H. “two-dimensional” solution. There are solutions of the following forms: w(x. For other exact solutions. see equation 4.330 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES where the functions R = R(x. η = y|t|λ−1 . z. y. z. y. η. z.6 with f (w) = a 1 . 2 6◦ . where λ is an arbitrary constant. “two-dimensional” solution. t) = G(x. . z. 2 2 ρ = x 2 − a 1 t2 χ2 = z/y q = xt 2 −1 −1 −2 “two-dimensional” solution. t) = F (ξ. xa1 1/2 −1/2 sinh λ + t cosh λ).2. z). y. w(x. There are solutions of the following forms: y . g(w) = a2 ew . ρ = x 2 − a 1 t2 . χ2 ) + 2 ln |y/t|. r = x 2 − a 1 t2 . q). For other exact solutions. y. η = (a3 y + a2 z )(x − a1 t ) ∂ ∂w ∂ 8 .2. g(w) = a2 ew . t) is a solution of this equation. y. y. y. Ibragimov (1994). ∂2w = a1 ∂2w + a2 + a3 ∂w w2 = w(x cosh λ + ta1 sinh λ. χ = |x2 − a1 t2 |k−1 |y|−2k z 2 “one-dimensional” solution. η = y|t|β−1 . . t) = W (η). 7◦ . “two-dimensional” solution. r = x − a1 t . The general solution of equation (2) is given in 3. y. z. s) + 2 ln |y|. q). z. t) and Q = Q(y. k where ϕ(ρ1 ) and ψ(ρ2 ) are arbitrary functions and C1 and C2 are arbitrary constants. z. 4◦ . ξ = xt−1 . r = a 3 y 2 + a2 z 2 . © 2004 by Chapman & Hall/CRC d d d d ζ = z|t|kβ−1.2. t) = W (χ) − k−1 z 5◦ . z. y. ζ) − 2β ln |t|. z. t) = E(r. y. t) = V (r. w(x. 2 2 p = (a3 y + a2 z )t . where β is an arbitrary constant. w(x. t) = H(ρ. s = z|y| “two-dimensional” solution. ∂y ∂y ∂z ∂z 1◦ . ◦ 2 . t) = V (p. y. and h(w) = a3 ew .1. C1 t + C6 ) − ln C2 . C6 and λ are arbitrary constants. ◦ w(x. ρ2 ) + 2 ln t w(x. z). (3) ∂y 2 ∂z 2 and A is an arbitrary constant. the Helmholtz equation (3) can be reduced to the Laplace equation (1). t) = G(r. y. w(x. z. f ¦e Reference: N. y. 2 χ1 = x/t. z. z) are determined by the differential equations ∂2R ∂2R = + AeR . w(x. t). η. . and h(w) = a3 ekw . z. (2) ∂t2 ∂x2 ∂ 2 Q ∂ 2Q + = A.6 with f (w) = a 1 . ρ1 = xt−1 . ρ2 = |t|k−1 |y|−k z w(x. Additive separable solution: √ √ 1 w(x. y. z. t) = U (p.6. By the substitution Q = 1 Ay 2 + u. y. r. r = a 3 y 2 + a2 z 2 . Then the functions 3. t) = ϕ(x + t a1 ) + ψ(x − t a1 ) + ln |y + C1 | + ln |z + C2 |. ζ = z|t|λ−1 . . where C1 . ζ) − 2λ ln |t|. “two-dimensional” solution. w(x. . y. There are “three-dimensional” solutions of the following forms: w(x. p = (x2 − a1 t2 )y −2 . z.1. t) = U (χ1 . There are “three-dimensional” solutions of the following forms: w(x. z. ∂t2 ∂x2 ew ekw k 2 w1 = w( C1 x + C3 . t) = F (ξ. Suppose w(x. 3◦ . C1 C2 y + C4 .

2 2 a1 (C1 y + C2 z + C3 t + C4 )2 + (a2 C1 + a3 C2 )(x + C5 )2 2 C3 (y + C5 )2 . t) = G(ρ. . . z. t) = F (ξ. y. z). y. t) = C1 t + C2 + ln C3 (a2 a3 x2 + a1 a3 y 2 + a1 a2 z 2 )−1/2 + C4 . ∂ ∂w ∂2w = a1 ew 2 ∂t ∂x ∂x + a2 ∂ ∂w ew ∂y ∂y + a3 ∂ ∂w ew .6 with λ1 = λ2 = λ3 = 1. t) = V (χ). 5◦ . z. −x a2 /a1 sin β + y cos β. w(x. ξ = x|t|β−1 . 2 2 a3 (C1 x + C2 y + C3 t + C4 )2 + (a1 C1 + a2 C2 )(z + C5 )2 where C1 . C1 t + C6 ) − ln C2 . .4 with f (w) = ew . Additive separable solution: w(x. t) = ln 1 2 2 2 a 1 C1 + a 2 C2 + a 3 C3 C1 x + C2 y + C 3 z + C 4 t + C5 2 . y. z. w(x. y. where the function Θ = Θ(x. H.5.5. . t) = ln w(x. 4◦ . z.6. Ibragimov (1994). t) = ln w(x. ζ) − 2β ln |t|. x = a1 −1/2 −1/2 −1/2 x. z. For other exact solutions. z. ◦ 2 . z. t . . t) = C1 t + C2 + ln a3 C3 x2 + a3 C4 y 2 − (a1 C3 + a2 C4 )z 2 + C5 . where β is an arbitrary constant. y. . w(x. see equation 4. © 2004 by Chapman & Hall/CRC . q.4. z. z. y. η. y. t). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). C1 C2 z + C5 . z) is determined by the Laplace equation ∂ 2 Θ ∂ 2 Θ ∂ 2Θ + + = 0. y. z. ∂x2 ∂y 2 ∂z 2 For this linear equation. t) = ln w(x.2. y. q = y/z. w(x. 3◦ . t) = C1 t + C2 + ln Θ(x. see Tikhonov and Samarskii (1990) and Polyanin (2002). EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 331 4. z. . C5 are arbitrary constants. “one-dimensional” solution. w(x. 2 C3 (x + C5 )2 . y. where C1 . 6◦ . Suppose w(x. t) = U (r. “two-dimensional” solution. ∂z ∂z This is a special case of equation 4. g g g g w2 = w x cos β + y a1 /a2 sin β. t) = H(p. y. y. 1◦ . 2 2 a2 (C1 x + C2 z + C3 t + C4 )2 + (a1 C1 + a3 C2 )(y + C5 )2 2 C3 (z + C5 )2 . There are solutions of the following forms: w(x. z. ζ = z|t|β−1. y. i ¦h Reference: N. z = a3 z. Solutions: w(x. There are “three-dimensional” solutions of the following forms: w(x. z. η = y|t|β−1. r = a 2 a 3 x2 + a 1 a 3 y 2 + a 1 a 2 z 2 χ = (a2 a3 x + a1 a3 y + a1 a2 z )t 2 2 2 −2 ρ = a 2 x2 + a 1 y 2 . z. y. y. . t) is a solution of the equation in question. y = a2 y. t) + 2 ln |z|. C6 and β are arbitrary constants. Then the functions 2 w1 = w( C1 C2 x + C3 . p = x/z. t).4. C1 C2 y + C4 . z.

z. z. H. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). t) = C1 t + C2 + 1 ln |z + C3 | + ln Θ(x. j j j j w2 = w x cos β + y a1 /a2 sin β. w = H(ρ1 . C6 and β are arbitrary constants. ρ1 = y/x. y = a2 −1/2 y. ξ = x|t|β−1 . 3◦ . λ1 λ2 λ3 where C1 . t). where C1 . Ibragimov (1994). Then the functions k 2 w1 = w( C1 C2 x + C3 . . 1◦ . ◦ 2 . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂2w = a1 ew 2 ∂t ∂x ∂x + a2 ∂ ∂w ew ∂y ∂y + a3 ∂ ∂w ekw . ζ) − 2β ln |t|. Traveling-wave solution in implicit form: 2 2 2 a 1 k 1 λ1 w a 2 k 2 λ2 w a 3 k 3 λ3 w e + e + e − β 2 w = C1 (k1 x + k2 y + k3 z + βt) + C2 . t) = C1 t + C2 + 1 1 1 ln |x + C3 | + ln |y + C4 | + ln |z + C5 |. ρ2 = |t| w = U (χ) + ln (a2 x + a1 y )t 2 |x| z. “three-dimensional” solution. y. z. . k2 . y. see Tikhonov and Samarskii (1990) and Polyanin (2002). y. k1 . C1 t + C6 ) − ln C2 . . 4◦ . C2 . y). Ibragimov (1994). r = a 2 x + a1 y 2 2 −2 2 2 k−1 x = a1 −1/2 x. η = y|t|β−1. λ1 λ2 λ3 3◦ . 2 −2/k −k .4. and β are arbitrary constants. z. η. Suppose w(x. There are solutions of the following forms: w = F (ξ. where the function Θ = Θ(x. y. © 2004 by Chapman & Hall/CRC . j j j j where C1 . C1 C2 3 z + C5 . . ∂x2 ∂y 2 For this linear equation. z. t) is a solution of this equation. . C6 are arbitrary constants. y) is determined by the Laplace equation “three-dimensional” solution. ∂2w ∂t2 = a1 ∂ ∂x eλ1 w ∂w ∂x + a2 ∂ ∂y eλ2 w ∂w ∂y + a3 ∂ ∂z eλ3 w ∂w ∂z . where β is an arbitrary constant. χ = (a2 x + a1 y )|z| |t| 2/k−2 “one-dimensional” solution. Additive separable solution: w(x. ∂z ∂z 1◦ . −x a2 /a1 sin β + y cos β. . H. . C1 C2 2 y + C4 . l ¦k Reference: N. C1 t + C6 ) − ln C2 . k ∂ 2 Θ ∂ 2Θ + = 0. ρ2 ) + 2 ln |x/t|. 6. k3 .332 5. t . For other exact solutions.5. 2◦ . l ¦k Reference: N.6 with λ1 = λ2 = 1 and λ3 = k. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). see equation 4. . t) is a solution of this equation. z. “two-dimensional” solution. Additive separable solution: w(x. ζ = z|t|kβ−1 w = G(r. C1 C2 z + C5 . Then the functions λ λ λ 2 w1 = w( C1 C2 1 x + C3 . C1 C2 y + C4 . Suppose w(x.

ρ2 ) + λ1 t 10◦ +. y. ζ = z|t|βλ3 −1 .8: ∂U ∂ ∂U ∂ Φ(U ) + Ψ(U ) = 0. There are more complicated “two-dimensional” solutions of the form w(x. Ψ(U ) = a2 c2 eλ2 U + a3 c2 eλ3 U . y. λ2 ) 7 . z2 ). η) is determined by a differential equation of the form 5. ψ(v) = a2 b2 eλ2 v + a3 b2 eλ3 v − b2 . z1 = b1 x + b2 y + b3 z + b4 t. ξ = x|t|βλ1 −1 . b2 . ρ). z2 = c1 x + c2 y + c3 z + c4 t. ρ2 = |t|λ3 /λ1 −1 |x|−λ3 /λ1 z. η). ρ = c1 y + c2 z. y. Φ(U ) = a1 b2 eλ1 U − b2 . z. 2 5◦ . . (2) a1 1 2 3 ∂x ∂x ∂η ∂η which can be reduced to a linear equation. There is a “three-dimensional” solution of the form w(x.6: ∂ ∂u ∂2u 2 2 2 = ϕ(u) .4. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY PARAMETERS 333 4◦ . t) = W (ρ1 . 6◦ . where the function u = u(ξ. and b3 are arbitrary constants): w(x. C5 are arbitrary constants. (1) where the function v = v(x. y. Remark. ξ = k1 x + k2 y + k3 z. “Two-dimensional” solution (b1 . “Two-dimensional” solution (k1 . ζ = b1 t + b2 x. λ3 ) ←− (y.4. z. 2 1 1 2 ∂ζ ∂ζ ∂ρ ∂ρ which can be reduced to a linear equation. ◦ 9 . where the function U = U (ζ. a2 . t) = U (ζ. and k3 are arbitrary constants): w(x. η. There is a “two-dimensional” solution of the form x 2 ln .4. 2 2 2 + a 1 C 1 e λ1 w + a 2 C 2 e λ2 w = C 3 . ◦ 8◦ . ∂t2 ∂ξ ∂ξ which can be reduced to a linear equation. t) = v(x. ρ) is determined by a differential equation of the form 5. © 2004 by Chapman & Hall/CRC η = y|t|βλ2 −1 . Relations (1) and equation (2) can be used to obtain two other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. t) = V (z1 . t) = F (ξ. a3 . .5. ρ1 = |t|λ2 /λ1 −1 |x|−λ2 /λ1 y. t) = u(ξ. a1 . The solution speciﬁed in Item 7◦ can be used to obtain other “two-dimensional” solutions by means of cyclic permutations of variables and determining parameters as shown in Item 6◦ .4.8: ∂v ∂ ∂v ∂ e λ1 v + ψ(v) = 0. z. ζ) − 2β ln |t|. . 2 2 2 2 + a 2 C 1 e λ2 w + a 3 C 2 e λ3 w = C 3 . y. Solutions in implicit form: C1 x + C2 y + C 3 z + C 4 t + C5 a 1 e λ1 w a 2 e λ2 w a 3 e λ3 w 2 2 2 2 = a 1 C 1 e λ1 w + a 2 C 2 e λ2 w + a 3 C 3 e λ3 w . 2 2 2 + a 1 C 1 e λ1 w + a 3 C 2 e λ3 w = C 3 . z. where β is an arbitrary constant. . k2 . w(x. η = b1 y + b2 z + b3 t. . z.4. y. C1 y + C 2 z + C 3 t + C 4 x + C5 C1 x + C2 z + C 3 t + C 4 y + C5 2 C1 x + C2 y + C 3 t + C 4 z + C5 where C1 .4. z. λ1 ) (z.4. t). ϕ(u) = a1 k1 eλ1 w + a2 k2 eλ2 w + a3 k3 eλ3 w . t) is determined by a differential equation of the form 3. “Two-dimensional” solution (bn and cn are arbitrary constants): w(x.

z. z. There are “three-dimensional” solutions of the following forms: λ1 −β λ2 −β λ3 −β 2 ln |t|. w(x. ζ = z|t| β . . β w(x. η2 . ρ2 = A (x + C1 )2 + (y + C2 )2 + (z + C3 )2 − (t + C4 )2 . t) = U (ξ. Then the functions β−λ β−λ β−λ β w1 = w( C1 1 x + C2 . . t) is a solution of this equation. x2 + y 2 + z 2 . C1 3 z + C4 . y. t) = V (η1 . ξ=y+ x . k2 .6. 2◦ .1. For the case of axisymmetric solutions. ∂t2 ∂x2 ∂y 2 ∂z 2 The equation admits translations in any of the variables x. Suppose w(x.334 7. Equations with Three Space Variables Involving Arbitrary Functions 4. C2 . “Three-dimensional” solution: w = u(ξ. and the function w(ρ) is determined by the ordinary differential equation wρρ + 3ρ−1 wρ + A−1 f (w) = 0. η3 ). k3 . y. C5 are arbitrary constants. C η = (C 2 − 1)x2 − 2Cxy + C 2 z 2 . 2◦ . η. C1 t + C5 ) + 2 ln |C1 |. + f (w). y. ∂z 2 + r2 r= x2 + y 2 . k1 . ηn = an x + bn y + cn z + dn t (n = 1. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). t). . Equations of the Form ∂ 2 w = ∂ f (x) ∂w + ∂ f (y) ∂w + ∂ f (z) ∂w + g(w) ∂t2 ∂x 1 ∂x ∂y 2 ∂y ∂z 3 ∂z 1. where the arbitrary constant A and the expression in square brackets must have like signs. y. η. η = y|t| β . 2. C1 2 y + C3 . Traveling-wave solution in implicit form: C1 + 2 2 2 2 λ2 − k 1 − k 2 − k 3 −1/2 ∂2w = ∂2w + ∂2w + ∂2w f (w) dw dw = k1 x + k2 y + k3 z + λt + C2 . ξ = x|t| β . where C1 . y. 1 ∂ ∂w sin θ . 1◦ . z. + ∂ 2w . r= sin θ ∂θ ∂θ © 2004 by Chapman & Hall/CRC . ζ) − 4. . t) = w(ρ). HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂2w = a1 eλ1 w 2 ∂t ∂x ∂x + a2 ∂ ∂w eλ2 w ∂y ∂y + a3 ∂ ∂w eλ3 w ∂z ∂z + beβw . the Laplace operator on the right-hand side of the equation is expressed in cylindrical and spherical coordinates as ∂w ∂ 2w ∂ 2 w ∂ 2 w 1 ∂ + + = r ∂x2 ∂y 2 ∂z 2 r ∂r ∂r 1 ∂ ∂w ∂ 2w ∂ 2 w ∂ 2 w + + = 2 r2 2 2 2 ∂x ∂y ∂z r ∂r ∂r 4◦ . z.6. 3). m m m m where C1 . and λ are arbitrary constants. 1◦ . Solution: w(x. 3◦ . and t.

η). ξ = Ax + By + Cz. The solutions speciﬁed in Items 4◦ and 5◦ can be used to obtain other “threedimensional” solutions by means of the cyclic permutations of the space variables. η. Solution for n ≠ 2. C 5◦ . b(2 − m)2 c(2 − k)2 w = W (ζ. x2−n y 2−m z 2−k 1 4 + + − (t + C)2 . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 335 where C is an arbitrary constant (C ≠ 0). Polyanin and A. Zhurov (1998). ξ. ρ). There are “two-dimensional” solutions of the following forms: w = U (ξ. ζ) is determined by the equation 1 ∂2v + 1+ 2 2 ∂z C ∂2v ∂2v ∂2v ∂v − 4ξ + 4C 2 (ξ 2 + ζ) 2 + 2(2C 2 − 1) + f (v) = 0. where A. r2 = + ∂ ∂w cz k ∂z ∂z + f (w). a. t) is determined by the equation ∂ 2 U ∂ 2 U 1 ∂U ∂2U = (A2 + B 2 + C 2 ) + + + f (U ). and C are arbitrary constants and the function U = U (ξ. I.6. η. 2 2 2 B a(2 − n) b(2 − m) c(2 − k) 4 where B and C are arbitrary constants (B and the expression in square brackets must have like signs). where C is an arbitrary constant (C ≠ 0). ξ2 = 4 η2 = 4 n n y 2−m z 2−k x2−n + + . t). a(2 − n)2 4 ρ2 = 4 y 2−m z 2−k + . t) is determined by the equation ∂2u = ∂t2 1+ 1 C2 ∂2u ∂2u ∂u ∂2u − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) + f (u). 2 b(2 − m) c(2 − k)2 4 1 x2−n − (t + C)2 . k) ←− (y. m ≠ 2. b. 2−n 2−m 2−k 2◦ . ζ). © 2004 by Chapman & Hall/CRC . D. and the function v = v(z. “Three-dimensional” solution: w = v(z. 2 ∂ξ ∂ξ∂η ∂η ∂η x . 2 ∂ξ ∂ξ∂ζ ∂ζ ∂ζ Remark. ζ 2 = 4 The second and third solutions can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. ξ. c. dr2 r dr A= 2 2 2 + + . η. and k ≠ 2: w = w(r). ξ=y+ ζ = (C 2 − 1)x2 − 2Cxy − C 2 t2 . ∂ ∂w ∂ ∂w ∂2w = axn + by m 2 ∂t ∂x ∂x ∂y ∂y 1◦ . ∂t2 ∂ξ 2 ∂η 2 η ∂η 2. w = V (x. and the function w(r) is determined by the ordinary differential equation d2 w A dw + + Bf (w) = 0. t). m) p ¦o Reference: A. n) (z. and the function u = u(ξ. η= (Bx − Ay)2 + (Cy − Bz)2 + (Az − Cx)2 . “Three-dimensional” solution: w = U (ξ. B. 6◦ . 2 2 a(2 − n) b(2 − m) c(2 − k)2 z 2−k 1 y 2−m + − (t + C)2 .4.

r2 = x2−n y 2−m e−νz 1 4 + + − (t + C1 )2 . c.336 3. and ν ≠ 0: w = w(r). η1 ). ∂2w ∂t2 = ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y + ∂ ∂z ceνz ∂w ∂z 1◦ . m ≠ 2. Zhurov (1998). ζ 2 = 4 The second and third solutions can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. Integrating yields its solution in implicit form: C2 − 2B where C2 and C3 are arbitrary constants. ν) ←− (y. B aλ2 bµ2 cν 2 4 where B and C1 are arbitrary constants and the function w(r) is determined by the autonomous ordinary differential equation wrr + Bf (w) = 0. There are “two-dimensional” solutions of the following forms: w = U (ξ. B a(2 − n)2 b(2 − m)2 cν 2 4 where B and C are arbitrary constants and the function w(r) is determined by the ordinary differential equation 2 2 A A= + . Solution for n ≠ 2. ξ2 = 4 y 2−m e−νz x2−n + + . HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂2w = aeλx 2 ∂t ∂x ∂x + ∂ ∂w beµy ∂y ∂y + ∂ ∂w ceνz ∂z ∂z + f (w). a(2 − n)2 b(2 − m)2 cν 2 e−νz 1 y 2−m + − (t + C)2 . There are “two-dimensional” solutions of the following forms: w = U (ξ. and ν ≠ 0: w = w(r). w = V (x. µ ≠ 0. 1◦ . 2◦ . aλ2 bµ2 cν 2 e−µy e−νz 1 + − (t + C)2 . r 2−n 2−m 2◦ . η2 = 4 bµ2 cν 2 4 e−λx 1 − (t + C)2 . bµ2 cν 2 q w = W (ζ. . a. t). + f (w). b(2 − m)2 cν 2 4 2 w = V1 (x. µ) s ¦r Reference: A. ξ2 = 4 q −1/2 e−λx e−µy e−νz + + . Polyanin and A. Solution for λ ≠ 0. r2 = 4 e−λx e−µy e−νz 1 + + − (t + C1 )2 . I. 4. λ) (z. t). D. η1 = 4 q © 2004 by Chapman & Hall/CRC q f (w) dw dw = C3 r. wrr + wr + Bf (w) = 0. aλ2 4 ρ2 = 4 e−µy e−νz + . η). b. ρ).

cν 2 4 ∂ ∂w + ∂ ρ2 = 4 1 ρ2 = 4 2 y 2−m e−νz + . ρ3 ). w = U (ξ. r2 = 2 B a(2 − n) bµ2 cν 2 4 where B and C are arbitrary constants and the function w(r) is determined by the ordinary differential equation 2 1 w + Bf (w) = 0. ρ3 ).4. y. 2 η2 = 4 2 η3 = 4 t 2 w = W1 (ζ1 . a(2 − n)2 cν 2 ρ2 = 4 3 x2−n y 2−m + .6. 2 a(2 − n) b(2 − m)2 ∂w t e−νz 1 x2−n + − (t + C)2 . ζ3 = 4 t t 5. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 337 2 w = W1 (ζ1 . Multiplicative separable solution: w(x. a(2 − n)2 4 e−µy 1 − (t + C)2 . [h(z)Zz ]z + aZ ln Z + C3 Z = 0. ζ1 = 4 2 w = W2 (ζ2 . ρ2 ). Solution for n ≠ 2. a(2 − n)2 bµ2 . ϕtt − aϕ ln ϕ + (C1 + C2 + C3 − b)ϕ = 0. ρ1 ). η3 ). η2 ). © 2004 by Chapman & Hall/CRC t t w = V3 (z. where the functions X(x). wrr + 2−n r r 2◦ . and ϕ(t) are determined by the ordinary differential equations [f (x)Xx ]x + aX ln X + C1 X = 0. 2 a(2 − n) bµ2 4 1 x2−n − (t + C)2 . t). µ ≠ 0. There are “two-dimensional” solutions of the following forms: e−µy e−νz x2−n + + . bµ2 4 e−νz 1 − (t + C)2 . ∂ ∂w ∂ ∂w ∂ ∂w ∂2w = f (x) + g(y) + h(z) + aw ln w + bw. η3 ). Z(z). a(2 − n)2 cν 2 4 y 2−m 1 x2−n + − (t + C)2 . a(2 − n)2 cν 2 x2−n e−µy + . a(2 − n)2 b(2 − m)2 4 1 x2−n − (t + C)2 . [g(y)Yy ]y + aY ln Y + C2 Y = 0. w = V1 (x. cν 2 4 ρ2 = 4 1 e−µy e−νz + . ξ2 = 4 2 a(2 − n) bµ2 cν 2 e−µy e−νz 1 2 + − (t + C)2 . ζ2 = 4 2 w = W3 (ζ3 . ζ2 = 4 2 w = W3 (ζ3 . 2 b(2 − m) 4 e−νz 1 − (t + C)2 . + f (w). 2 a(2 − n) 4 1 y 2−m − (t + C)2 . w = w(r). η2 ). Y (y). η1 = 4 bµ2 cν 2 4 ∂t2 axn beµy ceνz t ∂2w = ∂ ∂w + w = V2 (y. t) = X(x)Y (y)Z(z)ϕ(t). b(2 − m)2 cν 2 x2−n e−νz + . 2 a(2 − n) cν 2 4 e−µy 1 x2−n + − (t + C)2 . ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . w = V3 (z. ζ3 = 4 t t t 6. z. 2 η3 = 4 t w = V2 (y. η1 ). ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . ζ1 = 4 2 w = W2 (ζ2 . 2 η2 = 4 e−νz 1 x2−n + − (t + C)2 . ρ1 ). bµ2 cν 2 ρ2 = 4 2 ρ2 = 4 3 x2−n e−νz + . and ν ≠ 0: x2−n e−µy e−νz 1 4 + + − (t + C)2 . ρ2 ).

−x 1/2 a2 /a1 sin β + y cos β. and C3 are arbitrary constants. w2 = w x cos β + y a1 /a2 sin β. ∂2w ∂2w ∂ ∂w ∂2w = a1 + a2 + h(w) . ∂x ∂x ∂y ∂y ∂z ∂z A particular solution of equation (1) is given by ϕ(t) = exp a − 2A a (t + B)2 + . z. η) is determined by the ﬁrst-order partial differential equation x h(U ) − λ2 and ϕ(η) is an arbitrary function. w ¦v Reference: N. xa1 where C1 . y. y. 2◦ . 4. . t 2 2 a 1 C1 + a 2 C2 . η = C 1 x + C2 y where C1 and C2 are arbitrary constants. t) = ϕ(t)Θ(x. A particular and the general solutions of the last equations can be obtained from the formulas of Item 2◦. Solutions in implicit form: u h(w) dw = zϕ(η) + ψ(η). “Two-dimensional” solution (generalizes the solutions of Item 2 ◦ ): u w(x. Equations of the Form ∂ 2 w = ∂ f (w) ∂w + ∂ f (w) ∂w + ∂ f (w) ∂w + g(w) ∂t2 ∂x 1 ∂x ∂y 2 ∂y ∂z 3 ∂z 1. H. 4 2a (1) where B is an arbitrary constant. Ibragimov (1994). C1 y + C3 . ϕ(η) and ψ(η) are arbitrary functions. C2 . ∂η (1) . Then the functions w1 = w( C1 x + C2 . C1 t + C5 ). where A should be set equal to b−C1−C2 −C3 . ◦ 2 .2. η). and the general solution can be written out in implicit form (C is an arbitrary constant): u aϕ2 ln ϕ + (A − 1 a)ϕ2 + B 2 −1/2 dϕ = C t. the function ϕ(t) is determined by the autonomous ordinary differential equation ϕtt − aϕ ln ϕ − Aϕ = 0. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). and the function Θ(x. C1 z + C4 . and λ are arbitrary constants.6. t . 3◦ . Here. C2 . z. y. ξ = z + λt. z. . z). y. C5 . z. .338 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES where C1 . η = C 1 x + C2 y where C1 . Suppose w(x. y. t) = U (ξ. −1/2 w3 = w x cosh λ + ta1 sinh λ. and λ are arbitrary constants. t) is a solution of this equation. where A is an arbitrary constant. and the function U = U (ξ. Multiplicative separable solution: w(x. t 2 2 a 1 C1 + a 2 C2 . y. . ∂t2 ∂x2 ∂y 2 ∂z ∂z 1◦ . β. z. © 2004 by Chapman & Hall/CRC ∂U ∂ξ 2λ 2 2 a 1 C1 + a 2 C2 u u u u sinh λ + t cosh λ . z) satisﬁes the stationary equation ∂Θ ∂ ∂Θ ∂ ∂Θ ∂ f (x) + g(y) + h(z) + aΘ ln Θ + (b − A)Θ = 0. ∂U = ϕ(η).

8: a2 ∂ ∂u ∂2u + h(u) = 0. y. w2 = w x.4. z. t) = G(ξ. t). a2 ). z. { ¦z Reference: N. Suppose w(x. ◦ r = a 2 x2 + a 1 y 2 2 2 2 2 2 −2 “three-dimensional” solution. Relations (2) and equation (3) can be used to obtain another “three-dimensional” solution by means of the following renaming: (x. 2 w(x. t) = H(ζ. (5) Remark 2. y. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 339 In the special case λ = 0. “three-dimensional” solution. 1/2 ta1 −1/2 z.4. “two-dimensional” solution. where Φ(U ) is an arbitrary function. ∂y 2 ∂z ∂z (3) y w = u(y. w(x. t . C1 z + C4 . and the function v = v(ξ. a1 ) (y. the general solution of equation (1) can be written out in implicit form: y 2 2 2λ a1 C1 + a2 C2 ξ η[h(U ) − λ2 ] = Φ(U ). © 2004 by Chapman & Hall/CRC . ∂t2 ∂x2 ∂y ∂y ∂z ∂z 1◦ . ζ=x √ t a1 . β. C1 y + C3 . z. z.4. In the general case.2. Relations (4) and equation (5) can be used to obtain another “three-dimensional” solution by means of the following renaming: (x. xa1 sinh λ y y y y where C1 . ∂2w ∂ ∂w ∂ ∂w ∂2w = a1 + a2 g(w) + a3 g(w) . − 2C √ a1 a1 a2 (4) where C is an arbitrary constant (C ≠ 0). y. z).6 with f (w) = a 1 and g(w) = a2 . w(x. y. equation (1) can be solved using a characteristic system of ordinary differential equations. Zaitsev. see equation 4. C1 t + C5 ). ζ) is determined by a differential equation of the form 5. a1 ) (y. z. . (2) which can be reduced to a linear equation. 6◦ . a3 /a2 sin β + z cos β. “one-dimensional” solution. ξ = x − a1 t 2 2 ζ = a 2 x + a1 y − a1 a2 t η = (a2 x + a1 y − a1 a2 t )z 7 . C5 . and Moussiaux (2002). . Ibragimov (1994). . t) is a solution of this equation. t) = U (η). y x +√ . ξ. Equation (3) is implicitly independent of the cyclic variable ζ (the constants of integration that appear in the solution will be arbitrary functions of ζ). 5◦ . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). In the special case ϕ(η) = 0. and λ are arbitrary constants. . “Three-dimensional” solutions: where the function u(y. y. 4◦ . equation (1) is an ordinary differential equation in ξ and can be easily integrated to obtain solutions of Item 2◦ . −y sinh λ. y cos β + z w3 = w x cosh λ + a2 /a3 sin β.6. H. z. z. t) = F (r. ξ= √ a1 C a2 η = (C 2 − 1) x2 xy − C 2 t2 . Then the functions w1 = w( C1 x + C2 . 2. “Three-dimensional” solution: w = v(z. Remark 1. There are solutions of the following forms: w(x. y. For other exact solutions. z). + t cosh λ . ζ). η) is determined by the equation 1+ 1 C2 ∂2v ∂2v ∂2v ∂v ∂ ∂v − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) + h(v) ∂ξ 2 ∂ξ∂η ∂η ∂η ∂z ∂z = 0.6. see Polyanin. a2 ). η). z. y.

y.8: ∂u ∂ ∂u ∂ g(u) + h(u) = 0. where C1 . y y= √ . z. z. . z. 2 q = zy −1 “two-dimensional” solution.4. t) = U (p. p = (a3 y + a2 z )t . y. q). ξ = x − a1 t −2 2 2 w(x. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). z. 3◦ . For other exact solutions. w(x. t) = W (ξ. There are solutions of the following forms: “three-dimensional” solution. . ζ). w(x. where the function u(y. t) = H(r. t) = G(ξ. t) = U (p. see equation 4. “two-dimensional” solution. C1 z + C4 . r = a 3 y 2 + a2 z 2 2 2 2 3◦ . 2◦ .2. 2 2 2 “two-dimensional” solution. ξ = x2 − a1 t2 w(x. xa1 1/2 −1/2 | | | | ∂2w = a1 ∂2w + ∂ ∂w + ∂ ∂w sinh λ + t cosh λ). z. z. C5 . The transformation v= brings (1) to the Laplace equation g(u) du. ∂y2 ∂z 2 For solutions of this linear equation. t) is a solution of this equation. ∂y ∂y ∂z ∂z which can be reduced to a linear equation. z. r. ζ=x √ t a1 . q). y. . w(x. y. w2 = w(x cosh λ + ta1 sinh λ. ζ) is determined by a differential equation of the form 5. y. t). g(w) h(w) w1 = w( C1 x + C2 . y. and λ are arbitrary constants. y. ζ=x √ t a1 . t) = V (η). z.340 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 2◦ . ζ) is determined by the differential equation a2 ∂u ∂ ∂u ∂ g(u) + a3 g(u) = 0. “three-dimensional” solution. There are solutions of the following forms: w(x. which is implicitly independent of the cyclic variable ζ (the constants of integration that appear in the solution are arbitrary functions of ζ). . 4 . ζ). y. “one-dimensional” solution. Then the functions 3. z). see equation 4. ξ). “Three-dimensional” solutions: where the function u(y. For other exact solutions. see Tikhonov and Samarskii (1990) and Polyanin (2002).4. z. y. w(x. The equation obtained is implicitly independent of the cyclic variable ζ (the constants of integration that appear in the solution will be arbitrary functions of ζ). in which g(w) should be renamed a2 g(w) and h(w) renamed a3 g(w). z). Suppose w(x. 2 q = xt −1 η = (a3 y + a2 z )(x − a1 t ) 2 −1 4◦ .2. z. y. z. © 2004 by Chapman & Hall/CRC . C1 t + C5 ). 2 −2 p = (x − a1 t )y . .6 with f (w) = a 1 . “Three-dimensional” solutions: w = u(y.6. ∂y ∂y ∂z ∂z (1) | | w = u(y. C1 y + C3 . a2 z z= √ a3 ∂2v ∂2v + = 0. ξ = x − a1 t 2 2 r = a 3 y + a2 z . ◦ “three-dimensional” solution. y.6. z. t) = F (x. z.6 with f (w) = a 1 . ∂t2 ∂x2 ∂y ∂y ∂z ∂z 1◦ .

t) is a solution of this equation. t) is a solution of this equation. see equation 4. C5 and β are arbitrary constants. Then the functions w1 = w( C1 x + C2 . y. 5.2. q). B. . −x a2 /a1 sin β + y cos β. There are “three-dimensional” solutions of the following forms: w = W (ρ.6. p = (a2 x + a1 y )t . and C are arbitrary constants. w1 = w( C1 x + C2 . t). in which f (w) should be renamed a 1 f (w) and g(w) renamed a2 f (w). g(w). C1 t + C5 ). and a3 f (w). ¦~ } } } } } } } } Reference: N. w2 = w x cos β + y a1 /a2 sin β. θ = a3 a1 a2 Ay Cy Bz Bx Az Cx √ −√ + √ −√ . C5 and β are arbitrary constants. H. ξ = a2 x2 + a1 y 2 w(x. For other exact solutions. © 2004 by Chapman & Hall/CRC . . t) = W (ξ. ρ = a 2 x2 + a 1 y 2 . respectively. z.2. z. y.6. t). w2 = w x cos β + y a1 /a2 sin β. a2 ) 3◦ . see equation 4. r = a2 a3 x2 + a1 a3 y 2 + a1 a2 z 2 w(x. ¦~ ∂2w Reference: N. Suppose w(x. Remark. H. in which f (w). ◦ 2 . C1 z + C4 . where C1 . There are solutions of the following forms: w(x. t . y. a1 ) (z. “one-dimensional” solution. Ibragimov (1994). t) = Ψ(χ). y. z. y x2 xy z2 x + √ . C1 y + C3 . . Suppose w(x. There are exact solutions of the following forms: w(x. For other exact solutions. Then the functions 4. ξ = √ a1 C a2 a1 a1 a2 a3 By Cz Ax w = V (ζ. z. z.6. . C1 t + C5 ). . z.6. t) = U (p. χ = (a2 a3 x + a1 a3 y + a1 a2 z )t 2 2 2 −2 “two-dimensional” solution. a2 f (w). EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 341 ∂ ∂ ∂ ∂w ∂w ∂w ∂2w = a1 f (w) + a2 f (w) + a3 f (w) . t). + √ −√ a3 a3 a1 a2 a2 a1 2 2 2 where A. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). C1 y + C3 . y. . η. . 3◦ . C1 z + C4 . t . t). where C1 . t) = Φ(r.4. z. ζ = √ + √ + √ . ∂ ∂w ∂ ∂w ∂ ∂w = a1 f (w) + a2 f (w) + h(w) . z. θ. z. The ﬁrst and second solutions speciﬁed in Item 2◦ can be used to obtain other “threedimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. η = (C 2 − 1) − 2C √ + C2 .6. q = zt −1 “two-dimensional” solution. ◦ 2 . t). w = U (ξ. −x a2 /a1 sin β + y cos β. 2 2 −2 “three-dimensional” solution. and h(w) should be renamed a1 f (w). a3) ←− (y. 4◦ . 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . ∂t2 ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . y. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). z. . Ibragimov (1994).

HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂ ∂w ∂2w = f (w) + g(w) + h(w) . . which admits the ﬁrst integral 2 2 2 ξ 2 − C1 f (w) − C2 g(w) − C3 h(w) wξ = C6 . and the function u(ξ) is determined by the ordinary differential equation w = w(ξ). see Polyanin and Zaitsev (2003). ξ= (ξ 2 wξ )ξ = [ϕ(w)wξ ]ξ . . z. y. . . treating w in (1) as the independent variable. and the function u(η) is determined by the ordinary differential equation w = u(η). . Suppose w(x. . . . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). 2 ∂t ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . . k2 . 5◦ . k3 . 2 2 2 2 = C1 f (w) + C2 g(w) + C3 h(w). For C6 ≠ 0. and λ are arbitrary constants. . 2 2 4◦ . Solutions in implicit form: C1 x + C2 y + C 3 z + C 4 t + C5 C1 y + C 2 z + C 3 t + C 4 x + C5 C1 x + C2 z + C 3 t + C 4 y + C5 C1 x + C2 y + C 3 t + C 4 z + C5 where C1 . C5 are arbitrary constants. (3) x + C5 where C1 . η= 2 2 2 C3 uηη = [η 2 f (u)uη ]η + C1 [g(u)uη ]η + C2 [h(u)uη ]η .342 6. which can be reduced to a second-order linear equation. 2 2 2 2 f (w) + C1 g(w) + C2 h(w) = C3 . k1 . C5 are arbitrary constants. C1 t + C5 ). Solution: C1 y + C 2 z + C 3 t + C 4 . . C5 are arbitrary constants. 2 2 ϕ(w) = C1 f (w) + C2 g(w) + C3 h(w). 2 2 2 h(w) + C1 f (w) + C2 g(w) = C3 . © 2004 by Chapman & Hall/CRC . . 2 2 2 g(w) + C1 f (w) + C2 h(w) = C3 . t + C5 where C1 . we obtain a Riccati equation for ξ = ξ(w): 2 2 2 (2) C6 ξw = ξ 2 − C1 f (w) − C2 g(w) − C3 h(w). For exact solutions of equation (2). Solution: (1) To the special case C6 = 0 there corresponds the ﬁrst solution in Item 3◦ . t) is a solution of this equation. C1 z + C4 . C1 y + C3 . 3◦ . 2◦ . . Traveling-wave solution in implicit form: 2 2 2 k1 f (w) + k2 g(w) + k3 h(w) dw − λ2 w = C1 (k1 x + k2 y + k3 z + λt) + C2 . . where C1 . Then the functions w1 = w( C1 x + C2 . C1 x + C2 y + C 3 z + C 4 . . where C1 . C5 are arbitrary constants. . C2 .

see Ibragimov (1994). p2 q = y/t.4.4. z2 = b1 x + b2 y + b3 z + b4 t.4. For C6 ≠ 0. ψ(v) = a2 g(v) + b2 h(v) − c2 . “Two-dimensional” solution (k1 . z. z. y. and k3 are arbitrary constants): ◦ w(x. η = ay + bz + ct. y. g) 6 . where the function v = v(x. z. b. h) ←− (y.4. (5) For exact solutions of equation (5).4. 10◦ +. t) = V (z1 . Φ(U ) = a2 f (U ) − a2 . η) is determined by a differential equation of the form 5. ◦ (4) To the special case C6 = 0 there corresponds the second solution in Item 3 . s) is determined by the differential equation ∂2Θ ∂2Θ ∂2Θ ∂2Θ ∂2Θ ∂2Θ + q 2 2 + r2 2 + 2pq + 2pr + 2rq 2 ∂p ∂q ∂r ∂p∂q ∂p∂r ∂r∂q ∂Θ ∂Θ ∂ ∂Θ ∂ ∂Θ ∂ ∂Θ ∂Θ + 2q + 2r = f (Θ) + g(Θ) + h(Θ) . t) = v(x. + 2p ∂p ∂q ∂r ∂p ∂p ∂q ∂q ∂r ∂r 11◦ . t) = U (ζ. and c are arbitrary constants): w(x. η). where the function u = u(ξ. ξ = k1 x + k2 y + k3 z.6. which can be reduced to a second-order linear equation. ∂ ∂u ∂2u = ϕ(u) . k2 . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 343 which admits the ﬁrst integral 2 2 2 [η 2 f (u) + C1 g(u) + C2 h(u) − C3 ]uη = C6 .8: ∂v ∂ ∂v ∂ f (v) + ψ(v) = 0. s = z/t. t) is determined by a differential equation of the form 3. t). z2 ). 8◦ . ζ = a1 t + a2 x. ρ) is determined by a differential equation of the form 5.8: ∂U ∂ ∂U ∂ Φ(U ) + Ψ(U ) = 0. q. treating u in (4) as the independent variable. where the function Θ = Θ(p. Remark. ∂x ∂x ∂η ∂η which can be reduced to a linear equation. t) = Θ(p. “Three-dimensional” solution: w(x. The solutions speciﬁed in Items 7◦ and 8◦ can be used to obtain other “twodimensional” solutions by means of the cyclic permutations of variables and determining functions as shown in Item 5◦ . y.4. p = x/t. For results of the group analysis of the original equation. z.4. where the function U = U (ζ. ∂t2 ∂ξ ∂ξ which can be reduced to a linear equation. q.6: 2 2 2 ϕ(u) = k1 f (u) + k2 g(u) + k3 h(u). see Polyanin and Zaitsev (2003). s). There are more complicated “two-dimensional” solutions of the form w(x. t) = u(ξ. 9◦ . “Two-dimensional” solution (the an and bn are arbitrary constants): w(x. 2 1 1 2 ∂ζ ∂ζ ∂ρ ∂ρ which can be reduced to a linear equation. “Two-dimensional” solution (a. © 2004 by Chapman & Hall/CRC . 7◦ . z1 = a1 x + a2 y + a3 z + a4 t. we obtain a Riccati equation for η = η(u): 2 2 2 C6 ηu = η 2 f (u) + C1 g(u) + C2 h(u) − C3 . Formula (3) and equation (5) can be used to obtain two other “one-dimensional” solutions by means of the following cyclic permutations of variables and determining functions: (x. f ) (z. Ψ(U ) = b2 g(U ) + b2 h(U ). z. y. ρ = b1 y + b2 z. y. ρ).

Remark. z). © 2004 by Chapman & Hall/CRC .6. HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES + ∂ ∂w f (w) ∂y ∂y + ∂ ∂w f (w) ∂z ∂z – a2 f (w) + b. 2−n 2−m 2−k 2◦ . r r A=2 1−n 1−m 1−k + + . Solution for λ ≠ 0. y. There are “two-dimensional” solutions of the following forms: w = U (ξ. ξ2 = 4 η2 = 4 y 2−m z 2−k x2−n + + . k) ←− (y. ∂2w ∂t2 ∂x2 ∂y 2 ◦ 1 . y 2−m z 2−k 1 x2−n 4 + + − (t + C)2 . ∂x2 ∂y 2 ∂z 2 For this linear equation. where the function U = U (x. 4. and k ≠ 2: w = w(r).344 7. a(2 − n)2 4 ρ2 = 4 y 2−m z 2−k + . f 3 (w) ∂ ∂w ∂2w = f (w) 2 ∂t ∂x ∂x Solution in implicit form: f (w) dw = at + U (x. z) is determined by the Poisson equation ∂ 2U ∂ 2 U ∂ 2 U + + + b = 0. and ν ≠ 0: w = w(r). The above holds true if the constant b in the equation is replaced by an arbitrary function b = b(x. b. ρ). m ≠ 2. n) (z. z). b(2 − m)2 c(2 − k)2 w = W (ζ. 2 2 a(2 − n) b(2 − m) c(2 − k)2 z 2−k 1 y 2−m + − (t + C)2 . w = V (x. y. Other Equations 1. Solution for n ≠ 2. 2 2 2 B a(2 − n) b(2 − m) c(2 − k) 4 where C and B are arbitrary constants (B ≠ 0) and the function w(r) is determined by the ordinary differential equation wrr + A w + Bf (w) = 0. η). r2 = = axn ∂2w + by m ∂2w + cz k ∂2w ∂z 2 + f (w). r2 = = aeλx ∂2w + beµy ∂2w + ceνz ∂2w ∂z 2 + f (w). see Tikhonov and Samarskii (1990) and Polyanin (2002). B aλ2 bµ2 cν 2 4 where B and C1 are arbitrary constants and the function w(r) is determined by the autonomous ordinary differential equation wrr + 6r−1 wr + Bf (w) = 0. µ ≠ 0. m) 2. ζ 2 = 4 The second and third solutions can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x.3. c. t). ∂2w ∂t2 ∂x2 ∂y 2 ◦ 1 . y. b(2 − m)2 c(2 − k)2 4 1 x2−n − (t + C)2 . a. 4 e−λx e−µy e−νz 1 + + − (t + C1 )2 .

b(2 − m)2 cν 2 x2−n e−νz + . aλ2 4 bµ2 cν 2 The second and third solutions can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. wrr + wr + Bf (w) = 0. η 2 = 4 bµ2 cν 2 4 e−µy e−νz e−λx 1 − (t + C)2 . ρ2 = 4 + . r2 = 2 B a(2 − n) bµ2 cν 2 4 where B and C are arbitrary constants and the function w(r) is determined by the ordinary differential equation 2(5 − 3n) 1 w + Bf (w) = 0. There are “two-dimensional” solutions of the following forms: y 2−m e−νz x2−n + + . t). c. wrr + 2−n r r ∂2w = axn ∂2w + beµy ∂2w © 2004 by Chapman & Hall/CRC . ρ1 ). m ≠ 2. ξ2 = 4 a(2 − n)2 b(2 − m)2 cν 2 w = V1 (x. ∂2w ∂t2 = axn ∂2w ∂x2 + by m ∂2w ∂y 2 + ceνz ∂2w ∂z 2 + f (w). w = V3 (z. µ ≠ 0. η2 ). a. η3 ). a(2 − n)2 cν 2 2 w = W1 (ζ1 . η1 ). and ν ≠ 0: x2−n e−µy e−νz 1 4 + + − (t + C)2 . 2 η1 = 4 2 η2 = 4 2 η3 = 4 e−νz 1 y 2−m + − (t + C)2 . w = V (x. r2 = 2 2 B a(2 − n) b(2 − m) cν 2 4 where B and C are arbitrary constants and the function w(r) is determined by the ordinary differential equation 1−n 1−m A A=2 + +1 . 2 a(2 − n) b(2 − m)2 4 1 x2−n − (t + C)2 . b. w = U (ξ.6.4. 2 b(2 − m) cν 2 4 e−νz 1 x2−n + − (t + C)2 . 1◦ . ρ). t). w = U (ξ. µ) 3. ζ 2 = 4 (z. w = w(r). Solution for n ≠ 2. 2 a(2 − n) cν 2 4 y 2−m 1 x2−n + − (t + C)2 . cν 2 4 + ceνz ∂2w ρ2 = 4 1 ρ2 = 4 2 y 2−m e−νz + . ρ2 ). ν) ←− (y. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING ARBITRARY FUNCTIONS 345 2◦ . b(2 − m)2 4 e−νz 1 − (t + C)2 . + f (w). There are “two-dimensional” solutions of the following forms: e−λx e−µy e−νz + + . ρ3 ). η). r 2−n 2−m 2◦ . a(2 − n)2 b(2 − m)2 4. ξ 2 = 4 aλ2 bµ2 cν 2 e−µy e−νz 1 + − (t + C)2 . ζ3 = 4 ρ2 = 4 3 x2−n y 2−m + . ζ1 = 4 2 w = W2 (ζ2 . w = w(r). Solution for n ≠ 2. w = V2 (y. ∂t2 ∂x2 ∂y 2 ∂z 2 1◦ . ζ2 = 4 2 w = W3 (ζ3 . λ) w = W (ζ. and ν ≠ 0: x2−n y 2−m e−νz 1 4 + + − (t + C1 )2 . a(2 − n)2 4 1 y 2−m − (t + C)2 .

w = V3 (z. ξ2 = 4 e−µy e−νz x2−n + + . bµ2 cν 2 2 w = W1 (ζ1 . η1 ). a(2 − n)2 cν 2 x2−n e−µy + . η3 ). ζ2 = 4 2 w = W3 (ζ3 . bµ2 4 e−νz 1 − (t + C)2 . a(2 − n)2 bµ2 © 2004 by Chapman & Hall/CRC .346 HYPERBOLIC EQUATIONS WITH TWO OR THREE SPACE VARIABLES 2◦ . t). There are “two-dimensional” solutions of the following forms: w = U (ξ. ρ1 ). η2 ). a(2 − n)2 bµ2 cν 2 e−µy e−νz 1 2 + − (t + C)2 . ζ1 = 4 2 w = W2 (ζ2 . w = V1 (x. 2 a(2 − n) bµ2 4 1 x2−n − (t + C)2 . η1 = 4 bµ2 cν 2 4 2 η2 = 4 2 η3 = 4 e−νz 1 x2−n + − (t + C)2 . ρ2 ). 2 a(2 − n) cν 2 4 e−µy 1 x2−n + − (t + C)2 . ζ3 = 4 ρ2 = 4 2 ρ2 = 4 3 x2−n e−νz + . w = V2 (y. cν 2 4 ρ2 = 4 1 e−µy e−νz + . ρ3 ). a(2 − n)2 4 e−µy 1 − (t + C)2 .

C1 . C3 .1. ∂x2 ∂y 2 This is a steady heat and mass transfer equation with an nth-order volume reaction in two dimensions. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). Suppose w(x.1.1. This equation arises also in combustion theory and is a special case of equation 5. and C2 are arbitrary constants. Equations with Power-Law Nonlinearities 2 2 5.4. C. and D are arbitrary constants (n ≠ −1).4. Solutions: w(x. C2 . B. s= 1 4 k(1 w(x. 1◦ . Solution (generalizes the second solution of Item 2◦ ): w = w(r). 4◦ . r= (x + C1 )2 + (y + C2 )2 . = kwn . where A.1.1. x sin β + y cos β).Chapter 5 Elliptic Equations with Two Space Variables 5. traveling-wave solutions and solutions with central symmetry are also treated there. 2 B= k(n − 1)2 − A2 . y) = (Ax + By + C) 1−n . 1 1−n − n)2 1 1−n . C. 2(n + 1) . Then the functions n−1 n−1 2 w1 = C1 w( C1 x + C2 .1. where C1 and C2 are arbitrary constants. 2◦ . y) = s (x + C1 )2 + (y + C2 )2 where A. Traveling-wave solution in implicit form (generalizes the ﬁrst solution of Item 2 ◦ ): D+ 2kwn+1 (n + 1)(A2 + B 2 ) −1/2 dw = Ax + By + C. Equations of the Form ∂ w + ∂ w = aw + bw n + cw 2n–1 2 ∂x ∂y 2 The general properties of this type of equation are listed in 5. and the function w(r) is determined by the ordinary differential equation 1 wrr + wr = kwn . and β are arbitrary constants. r © 2004 by Chapman & Hall/CRC . 1. where C1 .1 with f (w) = kwn . C1 y + C3 ). ∂2w + ∂2w w2 = w(x cos β − y sin β. 3◦ . y) is a solution of the equation in question.

if b(n + 1) < 0. 1−n (1 − n)2 6◦ .5◦ .1. = aw + bwn . 1 1−n . z= . 2◦ . y) = − a(n + 1) 1 1−n ∂2w + ∂2w . F. © 2004 by Chapman & Hall/CRC . Traveling-wave solutions for a > 0: 2b sinh2 z w(x. Traveling-wave solutions for a < 0 and b(n + 1) > 0: 2b cos2 z w(x. £ ¤¢ dU = C4 θ. 4 U = kU n . . Multiplicative separable solution in polar coordinates (another representation of the solution of Item 5◦ ): w(x. Polyanin and V.1. ∂2w ∂2w + = awn + bw2n–1 .4. x + C1 where the function U = U (θ) is determined by the autonomous ordinary differential equation Uθθ + Integrating yields the general solution in implicit form: ¡ 4 2k U 2 + C3 U n+1 − n+1 (n − 1)2 where C3 and C4 are arbitrary constants. y) = − a(n + 1) 3. D. Zaitsev (2002).1 with f (w) = aw n + bw2n−1 . y) = (x + C1 ) 1−n u(ξ).4. 2 ξ= y + C2 . x + C1 where the function u(ξ) is determined by the ordinary differential equation (1 + ξ 2 )uξξ − 2(1 + n) 2(1 + n) ξuξ + u − kun = 0. y) = a(n + 1) 2b cosh2 z w(x. y) = 2(n + 1) 2an w(x. (1 − n)2 −1/2 tan θ = y + C2 . Reference: A. z= 1 2 |a| (1 − n)(x sin C1 + y cos C1 ) + C2 . y) = b 1 a(1 − n)2 (x + C1 )2 + (y + C2 )2 − 4 an 1 1−n 1 1−n . 2. Solutions: b(n + 1) a(1 − n)2 (x sin C1 + y cos C1 + C2 )2 − w(x. ∂x2 ∂y 2 This is a special case of equation 5. 2 r= (x + C1 )2 + (y + C2 )2 . 1 1−n 1 2 where C1 and C2 are arbitrary constants.1 with f (w) = aw + bw n . y) = r 1−n U (θ). where C1 and C2 are arbitrary constants. ∂x2 ∂y 2 This is a special case of equation 5. z= 1 2 √ a (1 − n)(x sin C1 + y cos C1 ) + C2 √ a (1 − n)(x sin C1 + y cos C1 ) + C2 if b(n + 1) > 0. Self-similar solution: w(x. 1◦ .

A=− b . x sin β + y cos β). z= √ a (1 − n)(x sin C2 + y cos C2 ). ∂2w ∂2w w2 = w(x cos β − y sin β.5. y) is a solution of the equation in question.4. 2◦ . y) = λ + C1 exp z 1 1−n . C1 y). ∂x2 + 1◦ . y) = (A + B sinh z) 1−n .1. w) 2 ∂x ∂y 2 1. 2 (n + 1)2 a an √ z = a (1 − n)(x sin C1 + y cos C1 ) + C2 . a(n + 1) 1 z= B= |a| (1 − n)(x sin C1 + y cos C1 ) + C2 .7: u ∂2u ∂2u + ∂x2 ∂y 2 + n 1−n ∂u ∂x 2 + ∂u ∂y 2 = a(1 − n)u2 + b(1 − n)u + c(1 − n). where C1 and β are arbitrary constants. y) = (A + B cos z) 1−n . 2◦ . b2 c − 2 an a (n + 1)2 ¥ ¥ ¥ B= 1/2 . The substitution u = w 1−n leads to an equation of the form 5. c b2 − a2 (n + 1)2 an 1/2 . ∂2w ∂x2 + ∂2w ∂y 2 = aw + bwn + cw2n–1 . Suppose w(x. y) = (A + B cosh z) 1−n .2 are roots of the quadratic equation aλ2 − a(n + 1)λ + b = 0. This is a special case of equation 5. 3◦ .4. 2 2 5.1. and C1 and C2 are arbitrary constants. Then the functions n−1 n−1 4 w1 = C1 w( C1 x.1 with f (w) = aw + bw n + cw2n−1 . © 2004 by Chapman & Hall/CRC ¥ . where C1 and C2 are arbitrary constants. 1◦ .2 with f (w) = aw n .1. where C1 and C2 are arbitrary constants (the expressions in square brackets must be nonnegative). A=− b .1. where λ = λ1. ∂2w ∂2w + = aw – a(n + 1)wn + bw2n–1 . Traveling-wave solutions for a > 0: w(x. Traveling-wave solutions for a < 0: w(x. Equations of the Form ∂ w + ∂ w = f (x.1. where the following renaming should be made: b → −a(n + 1) and c → b. a(n + 1) c b2 − . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). a(n + 1) 1 1 z= ¥ √ a (1 − n)(x sin C1 + y cos C1 ) + C2 .4. A=− b .6. = a(x2 + y 2 )wn . Traveling-wave solutions: w(x. See also equation 5. ∂x2 ∂y 2 1◦ . 1/2 B= w(x.1. y.2. 5. ∂y 2 This is a special case of equation 5.

1.4 with f (ξ.1.1: ∂2U ∂2U + = aU n .4.1.12 with f (z. 6.4. + 4.4.4.3 with f (w) = aw n . and q0 (x) = µxk .4.4. The transformation w = U (z. y.8 with a = b = 1. w) = kw n .2 with f (w) = kw n .3. ∂w . ζ = xy leads to a simpler equation of the form 5. + ∂2w ∂y 2 = (a1 x + b1 y + c1 ) ∂w + (a2 x + b2 y + c2 ) ∂w ∂y + kwn . 4. 7. w. ζ). g(x) = c.1. This is a special case of equation 5.5 with f (w) = kw n .1. 5. 2.4.4. = a(x2 + y 2 )k wn . 3. b ∂w ∂ 2 w ∂ 2 w a ∂w + + + = kwn . This is a special case of equation 5.14 with f (u) = kun . ∂x2 ∂y 2 This is a special case of equation 5.2. ∂x2 ∂y 2 This is a special case of equation 5.1. ∂w 2 ∂x ∂y 2 ∂x ∂y 1. 3. q1 (x) = γxm .4 with f (w) = aw n .4. g(x) = h 1 (x) = h0 (x) = p(x) = 0. ∂x2 ∂y 2 ∂y This is a special case of equation 5. ∂2w ∂x2 ∂2w + ∂2w ∂y 2 ∂2w = c(ax + by)k wn . w) = az k wn . + βxn y 2 + γxm y + µxk . ∂x2 ∂y 2 This is a special case of equation 5.1. 2 2 ∂x ∂y x ∂x y ∂y This is a special case of equation 5. ∂x2 + ∂2w ∂2w n + ∂2w + ∂2w . = aeβx wn . 2 2 5.6 with f (x) = b. ∂2w ∂x2 ∂x This is a special case of equation 5. = keax–by wn .2◦ .2. q2 (x) = βxn . + =α ∂2w ∂2w ∂w 2 © 2004 by Chapman & Hall/CRC . = k w + A11 x2 + A12 xy + A22 y 2 + B1 x + B2 y ∂y 2 This is a special case of equation 5.1. ∂z 2 ∂ζ 2 2. Equations of the Form ∂ w + a ∂ w = F x. ¦ ∂2w ∂w 2 ∂2w +a =b + cw + xn .2. and h(x) = sx n . ∂x2 ∂y 2 ∂y This is a special case of equation 5.1. 1 z = 2 (x2 − y 2 ). f (x) = α. ∂2w ∂x2 ∂2w + ∂2w ∂y 2 ∂2w = a(x2 + y 2 )(xy)k wn . w) = cz k wn .10 with f (z.2.4.

2.10 with a = b = 1. f (x) = α.+ bcw2 + kw + .12 with f (w. v) = 0.7 with f (x) = c.3.4. y) ∂w = g(w) ∂x ∂x ∂y ∂y Equations of this form are encountered in stationary problems of heat and mass transfer and combustion theory. y) = A + C1 exp(σx) + C2 exp(−σx) exp y −b . and h(x) = s.2. 5. ∂x2 ∂y 2 ∂y This is a special case of equation 5.4. ∂2w +a ∂2w =c ∂w 2 6.2. ∂x2 ∂y 2 ∂y This is a special case of equation 5. ∂2w + ∂2w = (a1 x + b1 y + c1 ) ∂w k ∂2w ∂2w ∂w 2 ∂2w + ∂2w = awn ∂w 2 + ∂w 2 ∂x2 ∂y 2 ∂x This is a special case of equation 5.1 with f (w) = cw k . Then the equation has the generalized separable solutions √ w(x. g(x) = keµx . Suppose the inequality 2Abc + k + ab = σ 2 > 0 holds.4. 3◦ . Here. ∂x2 ∂y 2 ∂y Let A be a root of the quadratic equation bcA2 + kA + s = 0. and g = g(w) is a source function that deﬁnes the law of heat (substance) release or absorption. ∂w ∂ ∂w ∂ axn + by m = cwk .9 with f (w) = aw n . ∂x2 ∂y 2 ∂x ∂y This is a special case of equation 5. . and g(y) = β. 2◦ . g(x) = k. For more complicated solutions. and h(x) = sxl .4. 1◦ .1.4. ∂x2 ∂y 2 ∂x ∂y This is a special case of equation 5. ∂w 2 ∂2w ∂2w +a = cxn + bcxn w2 + kxm w + xl . ∂w n ∂w m ∂2w ∂2w + =α +β + kw.7 with f (x) = cxn . y): ∂ 2 U ∂ 2U + = 0. u. g(x) = kxm . see equation 5. § ¨ ¨ § 5. 10. then the equation has the generalized separable solutions √ w(x. 9. and h(x) = seνx . © 2004 by Chapman & Hall/CRC .2. +a = ceβx 8. Equations of the Form ∂ f1 (x. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). If 2Abc + k + ab = −σ 2 < 0. + (a2 x + b2 y + c2 ) ∂w ∂y k .2. y) ∂w + ∂ f2 (x. f1 and f2 are the principal thermal diffusivities (diffusion coefﬁcients) dependent on the space coordinates x and y.4.4. The substitution a wn+1 dw U = exp − n+1 leads to the two-dimensional Laplace equation for U = U (x. ∂x ∂x ∂y ∂y This is a special case of equation 5.2.4. 1. 7. where C1 and C2 are arbitrary constants. y) = A + C1 cos(σx) + C2 sin(σx) exp y −b . ∂x2 ∂y 2 For solutions of this linear equation. § + bceβx w2 + keµx w + eνx .7 with f (x) = ceβx .

4. 2◦ .3. 1◦ . Equation (1) admits an exact solution of the form 2(1 + k + A − Ak) w= B(1 − k)2 1 k−1 ξ 1−k 2 with k ≠ 1.4. 3. ∂x ∂x ∂y ∂y This is a special case of equation 5. ξ = bµ2 x2−n + a(2 − n)2 e−µy 4c n 1 w = wm . ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m A w = Bwk . 2 − n ξ ξ abµ2 (2 − n)2 ∂w 1/2 .3. y + m−1 1−m ln C1 . ∂x ∂x ∂y ∂y This is a special case of equation 5. where the function w(ξ) is determined by the ordinary differential equation wξξ + 3. the function w(ξ) is determined by the ordinary differential equation wξξ + where A= (1) 4c 4 − nm . y) is a solution of the equation in question.6 with f (w) = cw m . The substitution ζ = ξ 1−A brings (1) to the Emden–Fowler equation 2A B ζ 1−A wk . Suppose w(x. where C1 is an arbitrary constant.3. Then the function w1 = C1 w C12−n x. 2◦ . where C1 and C2 are arbitrary constants. is also a solution of the equation. Then the function w1 = C1 w C12−n x. B= . (2 − n)(2 − m) ab(2 − n)2 (2 − m)2 3◦ . Functional separable solution for n ≠ 2 and m ≠ 2: w = w(ξ). ∂w ∂ ∂w ∂ axn + beµy = cwm . 3. (2) (1 − A)2 Over 20 exact solutions to equation (2) for various values of k can be found in Polyanin and Zaitsev (2003). wζζ = 2. Here. y) is a solution of the equation in question. C12−m y . For m = 4/n. is also a solution of the equation. 3. Suppose w(x. µ where C1 is an arbitrary constant. the exact solution can be represented in implicit form as © C1 + 2cn2 wk+1 ab(k + 1)(2 − n)4 −1/2 dw = C2 ξ. Functional separable solution for n ≠ 2 and µ ≠ 0: w = w(ξ). ξ ξ 1/2 k−1 k−1 . © 2004 by Chapman & Hall/CRC .8 with f (w) = cw m .1. Below are some exact solutions of equation (1).2.1◦ . ∂ aeβx ∂w + ∂ beµy = cwm .

8 with f (w) = 1 and g(w) = αw + β. Then the function w1 = C1 w x + 1−m 1−m ln C1 .1 with f (w) = kw n . 2 ∂x ∂y (a2 x + b2 y + c2 ) ∂w ∂y = kwn . ξ = bµ2 e−βx + aβ 2 e−µy )1/2 . where the function w(ξ) is determined by the ordinary differential equation wξξ − 1 w = Awm .4. is also a solution of the equation. The equation can be rewritten in the simpler form (ay + c) 5. f1 = f1 (w) and f2 = f2 (w) are the temperature (concentration) dependent principal thermal diffusivities (diffusion coefﬁcients).4. Below are some exact solutions of equation (1).2 with f (w) = kw n . ∂w ∂ ∂w ∂ (ay + c) + (bx + ) ∂x ∂x ∂y ∂y = kwn . Suppose w(x. 4 whose solutions with m = −1 and m = −2 can be found in Polyanin and Zaitsev (2003). w = w(x) and w = w(y).4. 2 3. Equations of the Form ∂ f1 (w) ∂w + ∂ f2 (w) ∂w = g(w) ∂x ∂x ∂y ∂y Equations of this form are encountered in stationary problems of heat and mass transfer and combustion theory. © 2004 by Chapman & Hall/CRC . Functional separable solution for βµ ≠ 0: w = w(ξ). Equation (1) admits a solution of the form w= abmβ 2 µ2 c(1 − m)2 1 m−1 ξ 1−m .5. This is a special case of equation 5. This is a special case of equation 5. Stationary Khokhlov–Zabolotskaya equation (for α = 1 and β = 0). y + ln C1 . nonlinear mechanics. 5. 2◦ . and g = g(w) is a source function that deﬁnes the law of heat (substance) release or absorption. 1. 3. Here.4. Simple solutions dependent on a single space variable.1. 4. are not considered in this subsection. This is a special case of equation 5. The substitution ζ = ξ 2 brings (1) to the Emden–Fowler equation wζζ = 1 Aζ −1 wm .1. and heat and mass transfer theory.4. ξ ξ A= 4c . y) is a solution of the equation in question.4. ∂ ∂x (a1 x + b1 y + c1 ) ∂w ∂x + ∂ ∂y ∂2w ∂2w + (bx + s) 2 = kwn . β µ where C1 is an arbitrary constant.1◦ . abβ 2 µ2 (1) 3◦ . ∂ ∂w ∂2w + (αw + β) ∂x2 ∂y ∂y = 0. It arises in acoustics.2.

Equations (2) and (3) are solved successively (these are linear in the unknowns ψ and χ. 1 y = 1 C1 t2 + C2 t − 3 αC1 w3 − 1 (αC3 + βC1 )w2 − βC3 w + C5 . w(x. B. and λ are arbitrary constants. . α where A. C1 . and C2 are arbitrary constants. y) = Ay − 1 A2 αx2 + C1 x + C2 . y) is a solution of the equation in question. . B1 . χ(x) = x+A α 4(x + A)2 2 54 ϕ(x) = − where A. α(x + A)2 B1 + B2 (x + A)3 . Then the function w1 = 2 β C1 w(C1 x + C3 . Solution in parametric form: x = C1 wt + C2 w + C3 t + C4 . Suppose w(x. is also a solution of the equation. 4◦ . C1 . 3◦ . y) = − α x+B x+B α 1 A(y + λx) + B . respectively). w(x. . Solution in parametric form: x = C1 t2 + C2 wt + C3 t + C4 w − C1 y = 1 C2 t2 + C4 t − C1 t αw2 + 2βw 2 © 2004 by Chapman & Hall/CRC 3 2 1 + C5 . w(x. the third is quadratic in y. where the functions ϕ = ϕ(x). y) = − β + λ2 α β w(x. (1) (2) (3) The nonlinear autonomous equation (1) is independent of the others. and χ = χ(x) are determined by the system of ordinary differential equations ϕxx + 6αϕ2 = 0. 2 2 5◦ . . and the fourth one is a traveling-wave solution. ψ(x) = (x + A)2 2 β αB1 1 1 C1 2 + C2 (x + A)2 − − − αB1 B2 (x + A)3 − αB2 (x + A)8 . χxx + 2αϕχ = −2βϕ − αψ 2 . y) = (Ax + B)y − 12A2 2 C1 β 1 y+A + + C2 (x + B)2 − . Generalized separable solution quadratic in y (generalizes the third solution of Item 2 ◦ ): w(x. y) = (Ax + B) C1 y + C2 − .1◦ . B2 . y) = ϕ(x)y 2 + ψ(x)y + χ(x). System (1)–(3) admits the following ﬁve-parameter family of solutions: 1 . ψ = ψ(x). C2 y + C4 ) + 2 α C2 2 C1 −1 . ψxx + 6αϕψ = 0. 3 αw + βw 3 1 1 − 3 αC2 w − 2 (αC3 + βC2 )w2 − βC3 w + C6 . C4 are arbitrary constants. Solutions: w(x. 2 α (Ax + B)4 + C1 x + C2 . C2 . . its solution can be expressed in terms elliptic integrals. The ﬁrst two solutions are linear in y. 2 C2 where C1 . 2◦ .

Solution (A. ξ = y + B ln( x + A) + C. Solution (A and λ are arbitrary constants): β 1 −2λx e Φ(u) − . w= Its order can be reduced. . α α where the function Φ = Φ(u) is determined by solving the generalized-homogeneous ordinary differential equation 4λ2 Φ − 3λ2 uΦu + λ2 u2 Φuu + (ΦΦu )u = 0. © 2004 by Chapman & Hall/CRC w= y= 2 g − 3 Ag 3 + B2 −1/2 dg. 10◦ . one obtains a Riccati equation for ζ = ζ(w): Cζw = (αw + β)ζ 2 + 1. and k are arbitrary constants): w= β 1 (x + A)2k F (z) − . (x + A)k+1 where the function F = F (z) is determined by solving the generalized-homogeneous ordinary differential equation 2k(2k − 1)F − (k + 1)(3k − 2)zFz + (k + 1)2 z 2 Fzz + (F Fz )z = 0. B2 . y) = which are independent. and C are arbitrary constants): β 1 ( x + A)−2 Ψ(ξ) − . where C is an arbitrary constant. Self-similar solution (A and B are arbitrary constants): w = w(ζ). C1 . 8◦ . C1 x = 3 Af + B1 C2 where B1 . α α z= y+B . ζ= x+A y+B where the function w(ζ) is determined by the ordinary differential equation wζζ + [ζ 2 (αw + β)wζ ]ζ = 0. 7◦ . Solution (generalizes the last solution of Item 2◦ ): β 1 f (x)g(y) − . and C2 are arbitrary constants. (4) w(x. u = (y + A)eλx . (ggy )y = −Ag. α α The functions f (x) and g(y) are determined by the autonomous ordinary differential equations (A is an arbitrary constant) fxx = Af 2 . Its order can be reduced. see Polyanin and Zaitsev (2003). 9◦ . B. B. The general solution to this equation can be expressed in terms of Bessel functions.6◦ . On integrating the equation once and taking w to be the independent variable. α α where the function Ψ = Ψ(ξ) is determined by the autonomous ordinary differential equation 6Ψ − 5BΨξ + B 2 Ψξξ + (ΨΨξ )ξ = 0. Its order can be reduced. Solution (A. Integrating the equations of (4) yields their general solutions in implicit form: −1/2 3 2 df .

Let w(x. which means that the equation is integrable by quadrature. V. ∂v ∂w −(αw + β) ∂x ∂y = . A.11◦ +. y) be any solution of the Khokhlov–Zabolotskaya equation (with α = 1 and β = 0). Kozlov (1995). F. one obtains a linear equation for x = x(w. and x and y. where v = −ϕ1/3 w(x. Zaitsev (2002). −A2 x2 + Bx + C β − . Through appropriate translations in both variables. Gibbons (1989).5. ¤ References for equation 5. y = y(w. Solutions: . v): ∂2x ∂2x + (αw + β) 2 = 0. w(x. η = y + αC1 x2 + αC2 x. as the dependent ones) brings it to the linear system ∂x ∂y = . Zaitsev and A. ∂y ∂x The hodograph transformation x = x(w. x= 1 3 ϕ−2/3 dt. F. p. y) = − Aα cos2 (py + q) α where A. V. D. y) = − Aα sinh2 (py + q) α 2 2 p Ax + Bx + C β − . ∂w2 ∂v 13◦ . y) = © 2004 by Chapman & Hall/CRC 1◦ . F (t. where C1 and C2 are arbitrary constants and the function U (η) is determined by the ﬁrst-order ordinary differential equation 2 2 (αU + β + α2 C2 )Uη + 2αC1 U = 8αC1 η + C3 . and q are arbitrary constants. w(x. ∂x ∂y −(αw + β) ∂w ∂v = . y) = Aα cosh2 (py + q) α 2 2 p Ax + Bx + C β − . one can make the equation homogeneous. ∂v ∂w On eliminating y. u) = 1 9ϕ ∂v + 3ϕtt u + 3ψt . D. v). Polyanin (2001). Solution: 2 w = U (η) − 4αC1 x2 − 4αC1 C2 x. Then the ordinary differential equation utt = F (t. The original equation can be rewritten as the system of equations ∂w ∂v = . w(x.1. y = ϕ−1/3 u − 1 3 ϕ−4/3 ψ dt. has a ﬁrst integral cubic in u t .1: Y. 2. α(Ay + D)2 α 2 2 p Ax + Bx + C β − . v) (w and v treated as the independent variables. Polyanin and V. y) − ϕ−1 (ϕt u + ψ)2 . w(x. ∂u and ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. 12◦ . C. V. Kodama and J. B. ∂2w ∂x2 + ∂ ∂y 1 ∂w αw + β ∂y = 0. u). D.

1 (with swapped variables. Then the functions wm wm n−m−1 n−m−1 2 x + C2 . 4◦ . . y = C 1 t2 + C 2 t − 2 α α where C1 . ∂x ∂x ∂y 2 3. the substitution U = w m+1 leads to an equation of the form 5. = αwn .1. C5 are arbitrary constants.8 with f (w) = 1 and g(w) = (αw + β) −1 . and C2 are arbitrary constants. For m ≠ −1.1. For other solutions with γ = 0.2. For other exact solutions. √ w + β ∂y 1 w + β leads to the equation The substitution U = α ∂U ∂ 2U ∂ U + = 0. Traveling-wave solutions linear in the coordinates: w(x. C2 .1.1: n ∂2U ∂2U + = α(m + 1)U m+1 . Suppose w(x. B. y) = Ax where A and B are arbitrary constants.2◦ . 4. the substitution w = eV leads to an equation of the form 5. ∂x2 ∂y 2 © 2004 by Chapman & Hall/CRC . 5.1. ∂x ∂x ∂y ∂y 1◦ . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). and β are arbitrary constants. ∂x2 ∂y 2 3◦ . The substitution αw + β = eU leads to an equation of the form 5.5. . For m = −1. ∂x ∂x ∂y 2 ∂x2 + ∂y y) and renaming the unknown function. 3◦ . 2◦ . x sin β + y cos β). 3◦ . w1 = C1 w( C1 ∂ ∂w + ∂ ∂w w2 = w(x cos β − y sin β.4. C3 . α2 (A2 α1 + B 2 α2 )w2 + 2(A2 β1 + B 2 β2 )w = γ(Ax + By)2 + C1 (Ax + By) + C2 .2. see equation 5. y) is a solution of this equation. where A.4. x y): ∂U ∂2U ∂ eU + = 0.4. Solution in parametric form: x = C1 wt + C2 w + C3 t + C4 .4.1: ∂2V ∂ 2V + = αenV . ∂x ∂x ∂y ∂y 1◦ . . where C1 . see 5.8 with f (w) = α1 w + β1 and g(w) = α2 w + β2 . Traveling-wave solution in implicit form: γ − A 2 α1 y + B. α ∂w = 0. ∂w ∂ ∂w ∂ (α1 w + β1 ) + (α2 w + β2 ) = γ. this equation ∂2w ∂ Up to the swap of the coordinates (x coincides with a special case of 5. C1 1 1 w − 2 (αC3 − βC1 ) ln |αw + β| + C5 . C 1 y + C3 ). C1 .1. 2◦ . .4.

η = ye(n−m)x . y) = 2 x n−k+1 ξ = α 1 x + α2 y U (z). y) = x n−m G(ξ). and C3 are arbitrary constants. ∂x2 ∂y 2 This is a special case of equation 5. are also solutions of the equation (the plus or minus signs are chosen arbitrarily).6. Integrating the equations of (2) yields their general solutions in implicit form: −1/2 2Ab f n+m+2 + B1 df = C1 x. w1 = C1 w( C1 where C1 . 4◦ .1.8 with f (w) = aw n and g(w) = bwm . There are exact solutions of the following forms: w(x. y) is a solution of this equation. C4 are arbitrary constants. y. see 5. B2 . 2◦ . gm − 3◦ . where k is an arbitrary constant. Then the functions 7. ∂x ∂x ∂y ∂y 1◦ . y) = e2x H(η). ∂w ∂ ∂w ∂ wn + a2 wm = bwk . n m −2 w1 = C1 w( C1 C2 x + C3 .4. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). ∂2w + aw4 ∂2w © 2004 by Chapman & Hall/CRC y. . C2 . y) = x−2k F (z). (1) The functions f (x) and g(y) are determined by the autonomous ordinary differential equations (A is an arbitrary constant) (f n fx )x = Abf m+1 . C1 C2 y + C4 ). a1 k−n−1 k−m−1 2 x + C2 .5. Suppose w(x. 5. self-similar solution. Then the functions 6. y) = f (x)g(y). C1 . 2 −1/2 z = yxmk−nk−1 . Other Equations Involving Arbitrary Parameters 1. Multiplicative separable solution: w(x. There are exact solutions of the following forms: w(x. (g m gy )y = −Aag n+1 . where C1 .4. ∂x ∂x ∂y ∂y 1◦ . For other exact solutions of the original equation. = by n w5 .1 with f (y) = by n .4. n + m + 2 ≠ 0. Suppose w(x. . . w(x. .∂w ∂ ∂w ∂ awn + bwm = 0. C 1 y + C3 ). ξ = y + k ln x. z) is a solution of this equation. (2) which are independent. 2◦ . and C2 are arbitrary constants. fn n+m+2 2Aa g n+m+2 + B2 dg = C2 n+m+2 where B1 . y) = F (ξ). w(x. z= k−m−1 yx n−k+1 traveling-wave solution. w(x. .

Functional separable solution for n ≠ 2 and m ≠ 2: w = w(ξ). 2.4. ∂2w ∂2w + aw4 = beβy w5 .9 with k = s = 0 and f (w) = cw m . B= (2 − n)(2 − m) ab(2 − n)2 (2 − m)2 2◦ .4.1 with f (y) = beβy . the general solution of (1) is written out in implicit form as 3n − 4 C1 + 2c(3n − 4)2 wk+1 ab(k + 1)(2 − n)4 −1/2 dw = C2 ξ. 2 (1 − A) (2) Over 20 exact solutions to equation (2) for various values of k can be found in Polyanin and Zaitsev (2003). 1◦ . 4. abβ 2 µ2 where the function w(ξ) is determined by the ordinary differential equation wξξ + 3 w = Awm .5. Here.5.3. Below are some exact solutions of equation (1). ξ ξ A= (1) © 2004 by Chapman & Hall/CRC .2. aeβx ∂2w ∂2w + beµy = cwm . .5. the function w(ξ) is determined by the ordinary differential equation wξξ + where A= A w = Bwk .2. 3. axn ∂2w ∂2w + beβy = cwm .5 with k = s = 0 and f (w) = cw k . For m = 1 k−1 ξ 1−k . Functional separable solution for βµ ≠ 0: w = w(ξ). ∂x2 ∂y 2 This is a special case of equation 5. 2 ∂x ∂y 2 This is a special case of equation 5. 2. 2 4n − 4 .7 with k = s = 0 and f (w) = cw m . 1◦ .1.5. equation (1) admits an exact solution of the form 2(1 + k + A − Ak) w= B(1 − k)2 2.4. For k ≠ 1. axn + by m = cwk . This is a special case of equation 5. 5. where C1 and C2 are arbitrary constants. 4c . ∂x2 ∂y 2 ∂2w ∂x2 ∂2w ∂y 2 This is a special case of equation 5. ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m 1/2 . The substitution ζ = ξ 1−A brings (1) to the Emden–Fowler equation wζζ = 2A B ζ 1−A wk . ξ = bµ2 e−βx + aβ 2 e−µy )1/2 .4. ξ ξ (1) 4c 3nm − 4n − 4m + 4 .

Traveling-wave solutions for α(1 + σ) < 0: w(x. w w(x. α 1 + 2σ α k1 x + k2 y + C. α 1 + 2σ z= − α k1 x + k2 y + C. α 1 + 2σ z= 2 = αw2 + βw + γ. k2 . β 1+σ .1. 3◦ . Below are some exact solutions of equation (1). Solution: w = w(r).1. β 1+σ . 2 (1 + 2σ)2 α ασ A1 = − B1 = z= w(x. There is a solution of the form ab(2 − m)β 2 µ2 w(ξ) = c(1 − m)2 1 m−1 ξ 1−m . ∂x2 ∂y 2 ∂w 2 ∂w ∂2w ∂2w + +σ + 2 2 ∂x ∂y ∂x ∂y 1◦ . 2◦ . 4 whose solution for m = 3 can be found in Polyanin and Zaitsev (2003).2◦ . The substitution ζ = ξ −2 brings (1) to the Emden–Fowler equation wζζ = 1 Aζ −3 wm . and C are arbitrary constants. The substitution w = eU leads to an equation of the form 5.2. w ∂2w ∂2w + ∂x2 ∂y 2 – ∂w ∂x 2 – ∂w ∂y 2 = αwβ .2. ασ α (1 + 2σ)2 A2 = − B2 = where k1 . where C1 and C2 are arbitrary constants and the function w(r) is determined by the ordinary differential equation 1 2 wwrr + wwr + σ wr = αw2 + βw + γ. y) = A2 + B2 sinh z. 1+σ 2 2 k 1 + k2 γ(1 + σ) β 2 (1 + σ)2 − . and C are arbitrary constants. y) = A1 + B1 cosh z. 2. β 1+σ . 6. α k1 x + k2 y + C. α2 (1 + 2σ)2 ασ A=− B= where k1 . Traveling-wave solutions for α(1 + σ) > 0: 7. 1+σ 2 2 k 1 + k2 β 2 (1 + σ)2 γ(1 + σ) − 2 .1: ∂2U ∂2U + = αe(β−2)U . 2 2. y) = A + B cos z. r= (x + C1 )2 + (y + C2 )2 . r © 2004 by Chapman & Hall/CRC . k2 . 1+σ 2 2 k1 + k2 γ(1 + σ) β 2 (1 + σ)2 − .

= 0. C1 . (x + C1 )3 + C3 y + C 4 . . t) is a solution of the equation in question. y) = − aA2 (y + C1 )7 + 4A(x + C2 )3/2 (y + C1 )5/2 − 7 3a(y + C1 )2 where A. D. a ∂w ∂ 2 w + ∂2w 2 .4◦ . H. and C2 are arbitrary constants. 9. y) = − y + C1 y + C1 3a(y + C1 )2 w(x. . C1 . y) = β 2(1 + 2σ) k1 x + k2 y 2 k1 2 2 γ(1 + σ) β (x + C1 )2 + (y + C2 )2 − . α → 1 α. 8. Ibragimov (1985).10 with f (w) = c1 w + k1 and g(w) = c2 w + k2 . + 2 k2 +C 2 − γ(1 + 2σ) . C6 are arbitrary constants. To this end. where variables and parameters should be renamed as follows: 1 σ → 1 + 2σ. w(x.10 with f (w) = c1 wn and g(w) = c2 wk . apart from the solutions presented in Items 1◦ to 3◦ . where C1 . ∂x ∂x ∂y ∂y This is a special case of equation 5. Solutions for α = 0: w(x. other solutions can be constructed. 4(1 + σ) βσ where k1 . Solutions: w(x. we apply the change of variable w = u2 to the original equation to obtain ∂u ∂u ∂2u ∂2u 1 + + (1 + 2σ) + = 1 αu2 + 2 β. . and γ → 2 β.4. w(x. 2 u 5◦ . C. ∂x ∂x ∂y ∂y This is a special case of equation 5. . . y) = −aC1 y 2 + C2 y + C3 3C1 4 B2 x + C1 y + C 2 (Ax + By + C3 )3/2 . 1◦ . Suppose w(x. y) = 39 3 3a(y + A)2 4 (C1 x + C4 )3/2 . F. © 2004 by Chapman & Hall/CRC ! ! ¤ ¤ ∂w ∂ ∂w ∂ (a1 x + b1 y + c1 w + k1 ) + (a2 x + b2 y + c2 w + k2 ) = 0. Polyanin and V. y) = C1 xy + C2 x + C3 y + C4 . y) = − . For γ = 0. Then the function −3 2 w1 = C1 C2 w(C1 x + C3 . 3a(y + C2 )2 3 2 2 (x + B)3 a 2 C1 (y + A)13 + aC1 (y + A)8 (x + B) + 3C1 (y + A)3 (x + B)2 − . It follows that its solution can be obtained with the formulas given in Items 1◦ and 2◦ . . 2βσ (x + C2 )3 3 + C3 y + C 4 . ◦ 10. . Reference: A. . 2 2 2 ∂x ∂y ∂x ∂y This equation is a special case of the original one. k2 . C4 are arbitrary constants (the ﬁrst solution is degenerate). β → 0. w(x.4. Reference: N.4. y) = −aA3 y 2 − aA2 3 3 aC1 1 (Ay + B)4 + C3 y + C4 . ∂w ∂ ∂w ∂ (a1 x + b1 y + c1 wn ) + (a2 x + b2 y + c2 wk ) = 0.4. y) = (Ay + B)(2C1 x + C2 )3/2 − 3 12A2 3/2 x + C2 (x + C2 )3 9aA2 + 4A − + C3 y + C 4 . Zaitsev (2002). w(x. w(x. C2 y + C4 ) + C5 y + C6 . B. ∂x ∂x2 ∂y 2 This is an equation of steady transonic gas ﬂow. is also a solution of the equation. y) = w(x. w(x.

Svirshchevskii (1995). C6 . 3 3◦ . ψ1 (y) = −2a y y0 (y − t)ψ2 (t)ψ3 (t) dt + B1 y + B2 . y) = η(y)θ(x) − aC1 y 0 (y − t)η 2 (t) dt + C2 y + C3 . 3a where A. A particular solution of the system is given by 45 4 aϕ2 ϕ3 = 18aϕ2 = 0. and the function U = U (z) is determined by the ordinary differential equation aUz Uzz + k 2 z 2 Uzz − 5k(k + 1)zUz + 3(k + 1)(3k + 2)U = 0. + 3C1 θ + C7 dθ = x + C8 . ψ3 + 18aψ3 ψ4 = 0. 2 ψ4 + 18aψ4 = 0. 4◦ . y) = ϕ1 (y) + ϕ2 (y)x3/2 + ϕ3 (y)x3 . R. References for equation 5. and the functions η(y) and θ(x) satisfy the autonomous ordinary differential equations (C4 is an arbitrary constant) ηyy + aC4 η 2 = 0. where k is an arbitrary constant. and C8 are arbitrary constants. Zaitsev (2002). . and C3 are arbitrary constants. Generalized separable solution: w(x. Generalized separable solution: w(x. 2 ψ2 + 2a(2ψ3 + 3ψ2 ψ4 ) = 0. C4 are arbitrary constants and y0 is any number. z = xy k . C7 . where the functions ϕk = ϕk (y) are determined by the autonomous system of ordinary differential equations 9 ϕ1 + 8 aϕ2 = 0. 2 aC 2 (y + A)8 .6. Titov (1988). Self-similar solution: 0. where C5 . © 2004 by Chapman & Hall/CRC $ ¤# .w(x. The general solution of the ﬁrst equation can be written out in implicit form (it is expressed in terms of the Weierstrass function). ψ4 (y) = − 6◦ .1. where the functions ψk = ψk (y) are determined by the autonomous system of ordinary differential equations ψ1 + 2aψ2 ψ3 = 0. 5◦ . Polyanin and V. B2 . D. y) = y −3k−2 U (z). θx θxx = C4 θ + C1 . F. S. C2 . The solutions to equations (1) and (2) can be written out in implicit form: " (1) (2) 2 C5 − 3 aC4 η 3 2 3 2 C4 θ −1/2 −1/3 dη = C6 y. C1 . ψ3 (y) = C1 (y + A)−2 + C2 (y + A)3 . 27 2 2 ψ2 (y) = C3 (y + A)−1 + C4 (y + A)2 − aC1 (y + A)−2 − 2aC1 C2 (y + A)3 − 1 (y + A)−2 .10: S. Generalized separable solution cubic in x: w(x. B1 . . 2 ϕ2 + ϕ3 + where the prime stands for the differentiation with respect to y. . . y) = ψ1 (y) + ψ2 (y)x + ψ3 (y)x2 + ψ4 (y)x3 . S. A. where C1 .

. is also a solution of the equation. and C2 are arbitrary constants. 6◦ . (y + C1 )2n+2 + A(y + C1 )n (x + C2 )3/2 + 16(n + 1)(2n + 1 + a) 9b (y + C1 )2 n2 + (a − 1)n + 5 (a − 3) = 0. Suppose w(x. C4 are arbitrary constants. where C1 . 5◦ . y Integrating yields θ(y) = − AB 3−a A2 B2 y2 − y − y 4−2a + C3 y 1−a + C4 . Then the function −3 2 w1 = C1 C2 w(C1 x + C3 . 2◦ . 1◦ . . y 1 8 2 45 a bϕ2 ϕ3 = 0. t) is a solution of the equation in question. y) = − bC1 y 2 + C2 y 1−a + C3 4(a + 1) 2 (C1 x + C4 )3/2 . 3C1 where C1 . where the functions ϕk = ϕk (y) are determined by the system of ordinary differential equations 9 a ϕ + bϕ2 = 0. 3◦ . 2(a + 1) 3−a 2(2 − a)(3 − a) w(x. . Additive separable solution: % w(x. y) = ϕ1 (y) + ϕ2 (y)x3/2 + ϕ3 (y)x3 . and the function U = U (z) is determined by the ordinary differential equation bUz Uzz + k 2 z 2 Uzz + k(a − 5k − 5)zUz + (3k + 2)(3k + 3 − a)U = 0. . and the n = n1. +b ∂y 2 y ∂y ∂x ∂x2 For b < 0. ϕ2 + ϕ2 + y 4 a ϕ3 + ϕ3 + 18bϕ2 = 0. 4 4◦ . . 3 y ϕ1 + where the prime stands for the differentiation with respect to y. . z = xy k . Generalized separable solution: 3 w(x. y) = − a − 3 (x + C2 )3 9A2 b .2 are roots of the quadratic equation where A.∂w ∂ 2 w ∂ 2 w a ∂w + = 0. Generalized separable solution: w(x. C2 y) + C4 y 1−a + C5 . 11. C1 . this equation describes a transonic gas ﬂow. where A. y) = y −3k−2 U (z). C1 . C5 are arbitrary constants. . y) = (Ay 1−a + B)(2C1 x + C2 )3/2 + 9bC1 θ(y). Generalized separable solutions: w(x. Self-similar solution: where k is an arbitrary constant. and the function θ = θ(y) is determined by the second-order linear ordinary differential equation a θyy + θy + (Ay 1−a + B)2 = 0. . © 2004 by Chapman & Hall/CRC . B. and C2 are arbitrary constants.

1. where the functions ψk = ψk (y) are determined by the system of ordinary differential equations a ψ + 2bψ2 ψ3 = 0. Its solution can be written out in implicit form: 2 3 2 C1 θ + 3C2 θ + C3 −1/3 dθ = x + C4 . 1◦ . Generalized separable solution: w(x. Titov (1988).6.1. F. Here. y y a ξyy + ξy + bC2 η 2 = 0. © 2004 by Chapman & Hall/CRC . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). D. y a ψ3 + ψ3 + 18bψ3 ψ4 = 0.2. y ψ1 + 8◦ . A.4.1. R. S. x sin λ + y cos λ). y) = ξ(y) + η(y)θ(x). C2 . y) is a solution of the equation in question. where C3 and C4 are arbitrary constants. C1 y + C3 ) + 2 ln |C1 |.7◦ . y) = ψ1 (y) + ψ2 (y)x + ψ3 (y)x2 + ψ4 (y)x3 . Zaitsev (2002). y 1 a 2 ψ2 + ψ2 + 2b(2ψ3 + 3ψ2 ψ4 ) = 0. Equations with Exponential Nonlinearities 2 2 5.11: S. the functions ξ(y) and η(y) are determined by the system of ordinary differential equations a η + bC1 η 2 = 0. Polyanin and V. Suppose w(x. ' ¤& References for equation 5. This equation occurs in combustion theory and is a special case of equation 5. w2 = w(x cos λ − y sin λ. ( ( where C1 . y ηyy + where C1 and C2 are arbitrary constants. Then the functions β β w1 = w( C1 x + C2 . 5.2. Generalized separable solution cubic in x: w(x.1 with f (w) = ae βw . and the function θ = θ(x) is determined by the autonomous ordinary differential equation θx θxx = C1 θ + C2 . Svirshchevskii (1995). and λ are arbitrary constants. C3 . Equations of the Form ∂ w + ∂ w = a + beβw + ceγw 2 ∂x ∂y 2 1. ∂2w ∂x2 + ∂2w ∂y 2 = aeβw . y a 2 ψ4 + ψ4 + 18bψ4 = 0. S.

y) = ln β aβ cos2 (Ax + By + C) 8C 2 1 − ln (x + A)2 + (y + B)2 − C . V. w(x. The ﬁrst four solutions are of traveling-wave type and the last one is a radial symmetric solution with center at the point (−A. ◦ 5 . w(x. w(x. N. (2) (3) . Aristov (1999). B. and C are arbitrary constants. ◦ 3 . β aβ[A cosh(kx + C1 ) + B sin(ky + C2 )]2 2k 2 (A2 + B 2 ) 1 . For a = β = 1. and k are arbitrary constants (x and y can be swapped to give another three solutions). ln β aβ sinh2 (Ax + By + C) −2(A2 + B 2 ) 1 if aβ < 0. y) = ln β aβ β where A. and the bar over a symbol denotes the complex conjugate. N. Vekua (1960). The √ solution is bounded at every point inside the circle. F. Polyanin (1996). Kh. (k − r 2 )2 k=5 √ 2 6. cos U . I. and C = k): ) w(r) = ln 8k . r ≤ 1. 4|Φz (z)| where Φ = Φ(z) is an arbitrary analytic (holomorphic) function of the complex variable z = x + iy with nonzero derivative. B. 2k 2k 2 (B 2 − A2 ) 1 ln . A. 1 ¤0 1 ¤0 Reference: S. Sabitov (2001).2◦ . y) = 2(A2 + B 2 ) 1 ln β aβ(Ax + By + C)2 if aβ > 0. 4◦ . C2 . 1 ¤0 References: D. and the second one has a singularity at the circumference ﬁrst r = k. Zaitsev and A. Functional separable solutions: 2 w(x. General solution: w(x. w(x. w r=1 Example. the boundary value problem for the circle r = x2 + y 2 ≤ 1 with the boundary condition = 0 has the following two solutions (see the last solution in Item 2◦ with a = β = 1. y) = 2 ln C1 eky β √ 2aβ cos(kx + C2 ) . y) = ln β aβ cosh2 (Ax + By + C) 2(A2 + B 2 ) 1 if aβ > 0. −B). C1 . 2(A2 + B 2 ) 1 if aβ > 0. y) = − w(x. A = B = 0. y) = w(x. The original equation is related to the linear equation ∂2U ∂2U + =0 ∂x2 ∂y 2 by the B¨ cklund transformation a ∂U 1 ∂w + β = ∂x 2 ∂y ∂U 1 ∂w − 2β = ∂y ∂x © 2004 by Chapman & Hall/CRC 1/2 1 2 aβ 1/2 1 2 aβ (1) exp exp 1 2 βw 1 2 βw sin U . Solutions: w(x. y) = ln β aβ[A sinh(kx + C1 ) + B cos(ky + C2 )]2 where A. D. References: I. y) = − 2 ln β |a|β 2 1 + sign(aβ)Φ(z)Φ(z) . Frank-Kamenetskii (1987).

M. Pukhnachov. Caudrey (1980). β where the functions ϕ(x) and ψ(y) are determined by the ﬁrst-order autonomous ordinary differential equations (ϕx )2 = 2aβϕ3 + C1 ϕ2 + C2 ϕ + C3 . w(x. 2 + C2 C1 2 ∂x2 ∂y 2 The transformation = aeβw – be–βw . V. Kaptsov. in the integration x is treated as a parameter. Rodionov (1999). y) = − b 1 ln − + C1 exp a β a β x sin C2 + y cos C2 b . where C1 . 3◦ .1: √ ∂2u ∂2u + = 2 ab sinh(βu). R.1. ∂x2 ∂y 2 Functional separable solution: + w(x. A. V. and C3 are arbitrary constants. (ψy )2 = 2aβψ 3 − C1 ψ 2 + C2 ψ − C3 − bβ. ∂2w ∂2w where C1 and C2 are arbitrary constants. The function Ψ(x) is determined where k = 1 aβ 2 after substituting this expression for w into equation (3). Grundland and E. 2◦ . C2 . © 2004 by Chapman & Hall/CRC . O. ∂2w + ∂2w aβ 1 + ln 2 2 β C1 + C2 2 2 a2 β 2 + bβ(C1 + C2 ) sin(C1 x + C2 y + C3 ) . Infeld (1992). A. y) = − 2 2 a2 β 2 − bβ(C1 + C2 ) aβ 1 + C3 exp(C1 x + C2 y) + exp(−C1 x − C2 y) . C2 .Suppose there is a (particular) solution U = U (x. y) = − 3. = aeβw + be–2βw . and C3 are arbitrary constants. y) + k. Traveling-wave solution (generalizes the solution of Item 1◦ ): w(x. Finally. 4 ¤3 1/2 Reference: R. Bullough and P. V. K. β β ∂x . ∂x2 ∂y 2 1◦ . Zhdanov (1994). J. K. J. k= b 1 ln 2β a leads to an equation of the form 5. 4 ¤3 References: A. ln − 2 2 2 + C 2 )2 β C1 + C2 4C3 (C1 2 where C1 .3. y) = ∂2w ∂2w 1 ln[ϕ(x) + ψ(y)]. 2 2 ∂x ∂y 4. Then (2) can be treated as a ﬁrst-order ordinary differential equation for w = w(y) with parameter x. = aeβw + be2βw . one obtains separable equations. Miller (Jr. which can 1 be reduced to a linear equation with the help of the change of variable z = exp − 2 βw). Z. y) of the Laplace equation (1). we have 2 ∂U 2 F = dy. and A.) and L. + w(x. w = − F − ln Ψ(x) − k e−F sin U dy . Rubel (1993). Solving these equations for the derivatives. Andreev. V. y) = u(x. Traveling-wave solution: w(x. Traveling-wave solution for bβ > 0: 2.

This is a special case of equation 5. Here.6. ∂x2 ∂y 2 The substitution U = αx + βy + µw leads to an equation of the form 5. y.2.1. ∂x2 ∂y 2 2 2 The substitution u = e−βw leads to a equation with a quadratic nonlinearity of the form 5.3 with f (w) = Aeβw . ∂z 2 ∂ζ 2 4.1.1. ∂2w ∂2w + = a + beβw + ce2βw .2.2.2.1 with f (w) = ceβw . y) ∂w + ∂ f2 (x. w) 2 ∂x ∂y 2 1.4.4.7: u ∂2u ∂2u + ∂x2 ∂y 2 − ∂u ∂x − ∂u ∂y + aβu2 + bβu + cβ = 0.1: 2. w = U (z.1.1: ∂2U ∂2U + = AµeU . 2 2 5.5. The substitution U = αxy + βx + γy + µw leads to an equation of the form 5. ζ). © 2004 by Chapman & Hall/CRC . 5. The transformation ζ = xy leads to a simpler equation of the form 5. ∂2w ∂x2 + ∂2w ∂y 2 = A(x2 + y 2 )eβw . f1 and f2 are the principal thermal diffusivities (diffusion coefﬁcients) dependent on the space coordinates x and y. ∂x2 ∂y 2 ∂2U ∂2U + = AµeU . 1 z = 2 (x2 − y 2 ). ∂2w ∂x2 + ∂2w ∂y 2 = A(x2 + y 2 )k eβw .2. ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y = ceβw .1. ∂2w ∂x2 + ∂2w ∂y 2 = Aeαxy+βx+γy eµw . and g = g(w) is a source function that deﬁnes the law of heat (substance) release or absorption. ∂x2 ∂y 2 3. Equations of the Form ∂ f1 (x. This is a special case of equation 5. Equations of the Form ∂ w + ∂ w = f (x.3. 1. ∂2w ∂2w + = Aeαx+βy eµw . y) ∂w = g(w) ∂x ∂x ∂y ∂y Equations of this form are encountered in stationary problems of heat and mass transfer and combustion theory.3.2.1: ∂2U ∂2U + = AeβU .

For A ≠ 1. Then the function w1 = w C 2−n x. another family of exact solutions follows 3. n−2 © 2004 by Chapman & Hall/CRC . Here. ab(2 − n)2 (2 − m)2 (1) 4 − nm . y) is a solution of the equation in question. β z = yx 2−m .2. C 2−m y + 2 2 2 ln C. 2 2λ2 1 ln β βB sinh2 (λξ + C) −2λ2 1 w(ξ) = ln β βB cosh2 (λξ + C) −8λ2 C1 C2 1 w(ξ) = ln β βB C1 eλξ + C2 e−λξ where λ. ξ ξ B= 4c . if βB < 0. y) = U (z) + n−2 ln x. β 2(1 − A) cn2 4 and B = . which corresponds to m = w(ξ) = w(ξ) = w(ξ) = 2 1 ln β βB(ξ + C)2 2λ2 1 ln β βB cos2 (λξ + C) if βB > 0. w(ξ) = ln − β βB β ab(2 − n)4 where C is an arbitrary constant. there is a solution of the form w(ξ) = − Bβ 1 ln ξ2 . n . the function w = w(ξ) is determined by the ordinary differential equation wξξ + where A= A w = Beβw . if βB > 0. Functional separable solution for n ≠ 2 and m ≠ 2: w = w(ξ). Below are some exact solutions of equation (1).1. Suppose w(x. .1◦ . 2◦ . if βB > 0. (2 − n)(2 − m) 3◦ . ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m 1/2 . There is an exact solution of the form w(x. 4◦ . is also a solution of the equation. For A = 1.3. For A = 0. C. 3. which corresponds to m = n−1 from (1): 8C 2 4c(n − 1)2 1 − ln(ξ 2 + C). B= . and C2 are arbitrary constants. we obtain from (1) several n ab(2 − n)4 more families of exact solutions to the original equation: 3. β where C is an arbitrary constant. C1 .

This is a special case of equation 5. a(2 − n)2 bβ 2 . Functional separable solution: w=− e−βx e−µy 1 ln cλ + λ aβ 2 bµ2 . y − ln C + ln C. y) = U (z) + 3. the function w(ξ) is determined by the ordinary differential equation wξξ − A= 4c . ∂ ∂x axn ∂w ∂x + ∂ ∂y beβy ∂w ∂y z=y− = ceλw . y − 2 2 2 ln C + ln C. y) = U (z) + z=y+ © 2004 by Chapman & Hall/CRC . Functional separable solution for βµ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(ξ).8 with f (w) = ceλw . This is a special case of equation 5. There is an exact solution of the form w(x. Here. Then the function w1 = w C 2−n x. ξ ξ n−2 ln x. 1◦ . ξ ξ β x. β λ where C is an arbitrary constant.3. 2−n 2−n ln x. β 4◦ . Suppose w(x.2. Then the function w1 = w x − 2 2 2 ln C. λ 1/2 . Functional separable solution: w=− cλ(2 − n) x2−n e−βy 1 ln + λ (1 − n) a(2 − n)2 bβ 2 e−βy x2−n + . 1◦ . y) is a solution of the equation in question. 3◦ . Functional separable solution for n ≠ 2 and β ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(ξ). is also a solution of the equation.4. β µ λ where C is an arbitrary constant. ξ2 = 4 where the function w(ξ) is determined by the ordinary differential equation wξξ + A w = ceλw . 2◦ . ∂ ∂x aeβx ∂w ∂x + ∂ ∂y beµy ∂w ∂y = ceλw . 3◦ . y) is a solution of the equation in question.3. Suppose w(x. µ 4◦ . ξ = bµ2 e−βx + aβ 2 e−µy 1 w = Aeλw .6 with f (w) = ceλw . is also a solution of the equation. 2◦ . abβ 2 µ2 β x. There is an exact solution of the form w(x.4. λ A= n .

β β 1 p2 1 . 6 where C1 . ∂x2 ∂y (a2 x + b2 y + c2 ) ∂w ∂y = keλw .4. are not treated in this subsection. β 2 1 w(x. .4.2 with f (w) = keλw .4. D. Traveling-wave solution in implicit form: 2 k1 w + 2 ak2 βw e = C1 (k1 x + k2 y) + C2 . 3◦ . ln(Ay 2 + By + C) + ln β β −aA sinh2 (px + q) where A. 1◦ . k1 . . y) = 1 p2 1 . p. β where C1 . ∂ ∂w ∂2w + aeβw 2 ∂x ∂y ∂y = 0. Suppose w(x. Additive separable solutions: 1 ln(Ay + B) + Cx + D.1 with f (w) = keλw .2. y) is a solution of this equation. B. and g = g(w) is a source function that deﬁnes the law of heat (substance) release or absorption. . ∂ ∂x (a1 x + b1 y + c1 ) ∂w ∂x + ∂ ∂y ∂2w ∂2w + (bx + s) 2 = keλw . f1 = f1 (w) and f2 = f2 (w) are the principal thermal diffusivities (diffusion coefﬁcients) dependent on the temperature (concentration) w. C. y) = ln(−aA2 y 2 + By + C) − ln(−aAx + D). w(x. . Then the functions β w1 = w(C1 x + C3 . and k2 are arbitrary constants. 5. 1. Simple solutions dependent on a single coordinate. y) = ln(Ay 2 + By + C) + ln β β aA cosh2 (px + q) 1 p2 1 . © 2004 by Chapman & Hall/CRC 5 4.4. C4 are arbitrary constants. C2 .4. The equation can be rewritten in the form (ay + c) 5. and q are arbitrary constants. y) = ln(Ay 2 + By + C) + ln β β −aA cos2 (px + q) w(x. y) = w(x. . ∂ (ay + c) ∂w + ∂ (bx + ) ∂w ∂y = keλw . w = w(x) and w = w(y). 2◦ .∂x ∂x ∂y This is a special case of equation 5. Here. w(x. Equations of the Form ∂ f1 (w) ∂w + ∂ f2 (w) ∂w = g(w) ∂x ∂x ∂y ∂y Equations of this form are encountered in stationary problems of heat and mass transfer and combustion theory. C1 C2 y + C4 ) − 2 ln |C2 |. This is a special case of equation 5. are also solutions of the equation.

y) = H(ζ) − x. For other solutions. There are exact solutions of the following forms: 2 w(x. Then the functions ∂x2 aeβw λ−β λ w1 = w( C1 x + C2 . Suppose w(x. This equation admits the ﬁrst integral z 2 + aeβu uz = C. Solution: w = U (ξ) − (β − λ)2 2 2 (β − λ)(β − 2λ) + ξUξ + ξ Uξξ + (aeβU Uξ )ξ = beλU . 2 The substitution Θ(u) = (uz ) leads to the ﬁrst-order linear equation 2 2 2 (k1 + ak2 eβu )Θu + 2ak2 βeβu Θ = 2beλu . y) is a solution of this equation. Treating u as the independent variable. β where k is an arbitrary constant. ξ = yx λ . where k1 and k2 are arbitrary constants. w = u(z). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). z = k1 x + k2 y. z = y+B where the function u(z) is determined by the ordinary differential equation (z 2 uz )z + (aeβu uz )z = 0. Czu = z 2 + aeβu .8 with f (w) = 1 and g(w) = ae βw . we get a Riccati equation for z = z(u). whose solution is expressed in terms of Bessel functions. 7 7 ∂2w + ∂ ∂w where C1 . see equation 5.4. and the function u(z) is determined by the autonomous ordinary differential equation 2 2 k1 uzz + ak2 (eβu uz )z = beλu . w = U (ξ) − β where k is an arbitrary constant and the function U (ξ) is determined by the ordinary differential equation 2(k + 1) + k(k − 1)ξUξ + k 2 ξ 2 Uξξ + (aeβU Uξ )ξ = 0. 2. y) = F (η) − ln |x|. Self-similar solution (A and B are arbitrary constants): x+A . Solution (generalizes the solution of Item 4◦ ): 2(k + 1) ln |x|. 2 λ λ λ2 © 2004 by Chapman & Hall/CRC . Traveling-wave solution: 3◦ . β 2 w(x. C2 . u = u(z). and C3 are arbitrary constants. 5◦ . β−λ 2 ln |x|.4◦ .4. β 6◦ . η = y + k ln |x|. = beλw . λ where the function U (ξ) is determined by the ordinary differential equation 2◦ . ξ = y|x|k . 7◦ . C1 y + C3 ) + 2 ln |C1 |. ∂y ∂y 1◦ . ζ = yex .

Suppose w(x. β−γ 2(β + γ) Aa(β − γ)2 1 ln − (y + C4 )2 . and k2 are arbitrary constants.4. β γ w1 = w( C1 C2 x + C3 . a = ceλw . C1 y + C3 ) + 2 ln |C1 |. C1 . and C2 are arbitrary constants. There are exact solutions of the following forms: 2k ln |x|. the functions ϕ(x) and ψ(y) are determined by the autonomous ordinary differential equations (A is an arbitrary constant) ϕxx + β(ϕx )2 = Abe(γ−β)ϕ . which are independent of each other. 2◦ . Remark. w(x. k1 . ξ = y + k ln |x|. Additive separable solution: w(x. Then the functions eβw eγw λ−β λ−γ w1 = w( C1 x + C2 . C4 are arbitrary constants. β + γ ≠ 0. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). ∂x ∂x ∂y ∂y 1◦ . (2) ψyy + γ(ψy )2 = −Aae(β−γ)ψ .4. B2 . .∂w ∂ ∂w ∂ aeβw + beγw = 0. © 2004 by Chapman & Hall/CRC 8 eγψ − 8 8 −1/2 dϕ = C1 x. Suppose w(x. 4. y. Traveling-wave solution in implicit form: 2 2 ak1 β −1 eβw + bk2 γ −1 eγw = C1 (k1 x + k2 y) + C2 . see 5. where k is an arbitrary constant. y) is a solution of this equation. γ−β 2(β + γ) where C4 and C2 are arbitrary constants. z = y|x|k−1 . y) = G(ξ) + β−γ w(x. and C3 are arbitrary constants. −1/2 dψ = C2 . 4◦ .8 with f (w) = ae βw and g(w) = beγw . C2 . y) = H(η) + 2x. Integrating yields the general solutions to the equations of (2) in implicit form: 8 eβϕ 2Ab (β+γ)ϕ e + B1 β+γ 2Aa (β+γ)ψ e + B2 β+γ where B1 . 8 8 ∂ ∂w +b ∂ ∂w where C1 . . where C1 . Particular solutions to equations (2) are given by ϕ(x) = ψ(y) = Ab(β − γ)2 1 ln (x + C3 )2 . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). For other exact solutions of the original equation. . C2 . 3◦ . . y) = F (z) + β−γ 2 ln |x|. η = ye(β−γ)x . C1 C2 y + C4 ) − 2 ln |C2 |. y) = ϕ(x) + ψ(y). y) is a solution of this equation. 5◦ . ∂x ∂x ∂y ∂y 1◦ . where C1 . Then the functions 3. (1) Here. w(x.

y) = w(x. The ﬁrst solution is degenerate and the others are representable as the sum of functions with different arguments. w(x. = 0. are also solutions of the equation. y) = ln β aA2 sinh2 (Cx + D) 1 C 2 sinh2 (Ay + B) ln . y) = ln β aA2 sinh2 (Cx + D) C 2 sinh2 (Ay + B) 1 . λ−γ 5. 3◦ . Solutions: w(x. ∂2w ∂2w + = aeβw ∂x2 ∂y 2 ∂w ∂x 2 + ∂w ∂y 2 .4. C4 are arbitrary constants. w(x. C. Then the functions β w = w(C1 x + C3 .2.2◦ . y) = ln β a sinh2 (Bx + C) β aA2 (x + C)2 B2 (y + A)2 1 1 cos2 (Ay + B) 1 . . © 2004 by Chapman & Hall/CRC .2. ) − 2 ln |C2 |. 2◦ . y) = ln . y) = Axy + By + Cx + D. ∂x2 ∂y 2 1◦ . w(x. . C1 C2 y + C4 . y) = ln 2 (Bx + C) β a cos β aA2 (x + C)2 w(x. 3. w(x. B. w(x. Other Equations Involving Arbitrary Parameters 1.5. . a > 0. y): ∂ 2 U ∂ 2U + = 0. Self-similar solution: w = w(z). 2. t) is a solution of this equation. a The substitution U = exp − eβw dw leads to the two-dimensional Laplace equation for β U = U (x.4.9 with f (w) = aeβw . ∂x2 ∂y 2 For solutions of this linear equation. There are exact solutions of the following forms: w(x. w(x. w(x. ∂2w ∂x2 + ∂2w ∂y 2 + a ∂w x ∂x + b ∂w y ∂y = ceβw . y) = ln . z = y/x. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). y) = F (z). w) = ceβw . . β aA2 cos2 (Cx + D) where A. and D are arbitrary constants. y) = G(ξ) + z = k1 x + k2 y. 2 ln |x|. β−λ ξ = y|x| β−λ .4 with f (ξ. @ 9 ∂2w + aeβw ∂2w where C1 . β aA2 cos2 (Cx + D) 1 C 2 cos2 (Ay + B) ln . where the function w(z) is determined by the ordinary differential equation (z 2 + aeβw )wzz + 2zwz = 0. This is a special case of equation 5. 1 B2 (y + A)2 1 sinh2 (Ay + B) 1 . y) = C 2 cos2 (Ay + B) 1 .2. This is a special case of equation 5. Suppose w(x.

β where C is an arbitrary constant.4.5 with k = s = 0 and f (w) = ce βw . 1◦ . ξ = y|x| λ . z = k1 x + k2 y. Suppose w(x. Traveling-wave solution: w = u(z). 2◦ . C2 . λ λ2 λ2 5. ξ ξ B= 4c . Solution: z. Here. A A where C1 . Suppose w(x. ∂2w + aeβw ∂2w 2(k + 1) = 0. y) is a solution of this equation. Then the functions λ−β λ w1 = w( C1 y + C2 . ∂x2 ∂y 2 This is a special case of equation 5. ζ = (y + C2 )(x + C1 )k . the function w(ξ) is determined by the ordinary differential equation wξξ + where A= (1) 3nm − 4n − 4m + 4 . is also a solution of the equation. 2 + ak2 eβu where C1 and C2 are arbitrary constants. Functional separable solution for n ≠ 2 and m ≠ 2: w = w(ξ). λ where the function U (ξ) is determined by the ordinary differential equation w = U (ξ) − (β − λ)2 2 2 (β − λ)(β − 2λ) + ξUξ + ξ Uξξ + aeβU Uξξ = beλU . and k are arbitrary constants. (2 − n)(2 − m) © 2004 by Chapman & Hall/CRC . ab(2 − n)2 (2 − m)2 1/2 . Then the function 2 ln C. and the function U = U (ζ) is determined by the ordinary differential equation w = U (ζ) − (k 2 ζ 2 + aeβU )Uζζ + k(k − 1)ζUζ + 4. and C3 are arbitrary constants. Solution: = beλw .5. where k1 and k2 are arbitrary constants. C 2−m y + 2 2 2◦ . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m A w = Beβw . β where C1 . Its solution can be written out in implicit form as A du √ = C1 F (u) 3◦ . ∂x2 ∂y 2 1◦ . β−λ 2 ln |x|. and the function u(z) is determined by the autonomous ordinary differential equation 2 2 (k1 + ak2 eβu )uzz = beλu . w1 = w C 2−n x. β 4◦ . F (u) = 2b 2 k1 eλu du + C2 . C2 . y) is a solution of the equation in question. C1 x + C3 ) + 2 ln |C1 |. axn ∂2w ∂2w + by m = ceβw .2(k + 1) ln |x + C1 |.

another family of exact solutions follows 3. if βB < 0. n . 3.4. we obtain from (1) 3n − 4 ab(2 − n)4 several more families of exact solutions to the original equation: 3.3. which corresponds to m = w(ξ) = w(ξ) = 2λ2 1 ln β βB cos2 (λξ + C) if βB > 0. which corresponds to m = n−1 from (1): 2 8C 4c(n − 1)2 1 − ln(ξ 2 + C).9 with k = s = 0 and f (w) = ce λw . if βB > 0.3◦ . ξ = bµ2 e−βx + aβ 2 e−µy )1/2 . This is a special case of equation 5. β µ λ where C is an arbitrary constant. equation (1) admits an exact solution of the form w(ξ) = − Bβ 1 ln ξ2 . Then the function w1 = w x − 2 2 2 ln C.2. Functional separable solution for βµ ≠ 0: w = w(ξ). β 2(1 − A) c(3n − 4)2 4n − 4 and B = . 2◦ .5. abβ 2 µ2 1 1 ln − Aλξ 2 . axn + beβy = ceλw .1. λ 4 7. For A ≠ 1. B= w(ξ) = ln − β βB β ab(2 − n)4 where C is an arbitrary constant.7 with k = s = 0 and f (w) = ce βw . For A = 0. . 6.5. y − ln C + ln C. © 2004 by Chapman & Hall/CRC . the function w(ξ) is determined by the ordinary differential equation wξξ + which admits the exact solution w=− ∂2w ∂x2 ∂2w ∂y 2 3 w = Aeλw . Suppose w(x.4. Below are some exact solutions of equation (1). 2λ2 1 ln β βB sinh2 (λξ + C) −2λ2 1 w(ξ) = ln β βB cosh2 (λξ + C) where λ and C are arbitrary constants. aeβx ∂2w ∂x2 + beµy ∂2w ∂y 2 = ceλw . This is a special case of equation 5. is also a solution of the equation. 1◦ . y) is a solution of the equation in question. For A = 1. Here. ξ ξ A= 4c .

Functional separable solution: 1 1+z 4 arctanh z = ln . C. J. where C1 and C2 are arbitrary constants. Cheb. and D are arbitrary constants. © 2004 by Chapman & Hall/CRC D 2 arctanh β ξ k+1 tanh . β 2 1−z where the functions f = f (x) and g = g(y) are determined by the ﬁrst-order autonomous ordinary differential equations 2 fx = Af 4 + Bf 2 + C. C. References: R. y) = arctanh f (x)g(y) . Denote k = 2b Traveling-wave solutions: 1 − k sin z 1 arccosh . C.1. ∂x2 ∂y 2 This is a special case of equation 5. . w(x. H. gy = −Cg 4 + (aβ − B)g 2 − A. = a sinh(βw) + b sinh(2βw). Traveling-wave solution in implicit form: D+ 2a cosh(βw) β(A2 + B 2 ) −1/2 dw = Ax + By + C. 4◦ . −C2 ): w = w(ξ). ξ= k−1 2 2bβ(k 2 − 1) (x sin C1 + y cos C1 + C2 ) if |k| > 1. . The original equation is related to (see 5. H. z = 2bβ(1 − k 2 ) (x sin C1 + y cos C1 + C2 ) if |k| < 1. Caudrey (1980). where A. B. Lee (1987). A.1 with f (w) = a sinh(βw).3.4. Equations with Hyperbolic Nonlinearities ∂2w ∂2w + = a sinh(βw). 1. ∂y 2 a . 2◦ . and A. Solution with central symmetry about the point (−C1 . 1◦ . ∂x2 D ∂2w + ∂2w w= where C1 and C2 are arbitrary constants. w= β sin z − k 2.5.3.3. ξ ◦ 3 .3. and C are arbitrary constants. Equations Involving Other Nonlinearities 5. Bullough and P.1. and the function w(ξ) is determined by the ordinary differential equation 1 wξξ + wξ = a sinh(βw). ξ= (x + C1 )2 + (y + C2 )2 .1) ∂2U ∂2U + = a sin(βU ) ∂x2 ∂y 2 by the B¨ cklund transformation a ∂U ∂w + =2 ∂x ∂y ∂U ∂w − =2 ∂y ∂x C ¤B 2 a 1 sin 2 βU cosh β a 1 cos 2 βU sinh β 1 2 βw 1 2 βw . Wing. and Y. B. K.

1 with f (w) = αw ln(βw). 6. ξ= De−2F + © 2004 by Chapman & Hall/CRC E dF . C. . (1) ∂x2 ∂y 2 ∂x ∂y 1◦ . and C are arbitrary constants.9 with f (w) = a coshn (βw).2. 1 U (x. ∂x2 ∂y ∂y This is a special case of equation 5. = k sinh(βw).3. ∂2w ∂2w + = a(x2 + y 2 ) sinh(βw).3.4. ∂x2 ∂y 2 This is a special case of equation 5.4. U (x.4.1. the function F = F (ξ) is deﬁned implicitly by α 1 (F − 2 ) A2 + B 2 where A.4.4.8 with f (w) = 1 and g(w) = a cosh(βw). Equations with Logarithmic Nonlinearities = αw ln(βw). On making the change of variable U = ln(βw). ∂x2 ∂y 2 This is a special case of equation 5. y) = 1 α(x + A)2 + 4 α(y + B)2 + 1. −1/2 Here. ξ = Ax + By + C. = a(x2 + y 2 ) cosh(βw). ∂x2 ∂y 2 This is a special case of equation 5. B. ∂x ∂x ∂y ∂y This is a special case of equation 5.2 with f (w) = a sinh(βw). ∂x2 ∂y 2 This is a special case of equation 5. + ∂2w + ∂2w ∂2w ∂2w = aeβx sinh(λw).2.4.6 with f (w) = k sinh(λw). Equation (1) has exact solutions quadratic in the independent variables: 1.4 with f (w) = a sinh(λw).3. 4.1 with f (w) = k sinh(βw). 2◦ . Equation (1) has a traveling-wave solution: U (x.4.3. 5. 4 √ 1 Aα−4A2 (x+B)(y+C)+( 4 α−A)(y+C)2 + 1 . ∂ ∂w + a cosh(βw) = 0.2 with f (w) = a cosh(βw). 5. .4. axn by m ∂w ∂w ∂ ∂ aeβx + beµy = k sinh(λw).4. B. ∂x2 ∂y 2 ∂x ∂y This is a special case of equation 5.1. y) = F (ξ). 9. and D are arbitrary constants.1. ∂2w ∂ ∂w + ∂ ∂w ∂2w ∂2w ∂w 2 + ∂w 2 8. y) = A(x+B)2 2 ∂2w + ∂2w where A. ∂x ∂x ∂y ∂y This is a special case of equation 5. + = a coshn (βw) 7.1. one obtains an equation with a quadratic nonlinearity: 2 2 ∂U ∂U ∂ 2 U ∂ 2U + + + = αU .

5. V. Rodionov (1999). and the function w(ζ) is determined by the ordinary differential equation 1 wζζ + wζ = αw ln(βw). A.3. Equation (1) has a solution in the form of the sum of functions with different arguments: U (x. It is a special case of equation 5. ψyy = bψ ln |ψ| − Cψ. 4.2. ∂x2 ∂y 2 This is a special case of equation 5.1. y) = ϕ(x)ψ(y). + 1◦ . and D.4. 4◦ . C is an arbitrary constant. 2. and A. Zhurov. V. y) = exp − x + (y + C)2 + 3 + b 4 b 2 where C is an arbitrary constant. Multiplicative separable solution: b a2 1 a . B1 . where A1 .2 with f (w) = α ln(βw). Andreev. V. ∂2w ∂2w + = aw ln w + (bxn + cy k )w. = axw + bw ln |w|. 2◦ . Polyanin. dg. Vyazmin. O.1. Kazenin (1998). ζ H ¤G References: J. ∂x2 ∂y 2 This equation is used for describing some ﬂows of ideal stratiﬁed ﬂuids. ∂x2 ∂y 2 This is a special case of equation 5. A. A. where C1 and C2 are arbitrary constants.8 with f (x) = bxn and g(y) = cy k .4. ζ= (x + C1 )2 + (y + C2 )2 . A. A. H ¤G ∂2w + ∂2w ∂2w ∂2w Reference: V. I.4. Pukhnachov. © 2004 by Chapman & Hall/CRC . = aeβx ln(λw). Here. D.1. w(x. Shercliff (1977). y) = f (x) + g(y).4 with f (w) = a ln(λw). A. A2 . ∂2w ∂2w + = α(x2 + y 2 ) ln(βw).3◦ . 3. Multiplicative separable solution (generalizes the solution of Item 1 ◦ ): w(x.6 with k = a2 = a0 = 0. Kaptsov. where the functions ϕ = ϕ(x) and ψ = ψ(y) are determined by the system of ordinary differential equations ϕxx = bϕ ln |ϕ| + (ax + C)ϕ. the functions f = f (x) and g = g(y) are deﬁned implicitly by F F A1 A2 x= y= 1 B1 e−2f + αf − 2 α 1 B2 e−2g + αg − 2 α −1/2 −1/2 df . ∂x2 ∂y 2 This is a special case of equation 5. K. V. The original equation admits exact solutions of the form w = w(ζ). and B2 are arbitrary constants.

4. w = w(r). w = w(ξ). y) = ϕ(x)ψ(y). Multiplicative separable solution: w(x. This equation is used to describe some steady-state axisymmetric (swirling) ﬂows of ideal ﬂuids. V.4. Rosen (1969). It also occurs in plasma physics. 7.4. ∂2w + ∂2w = a lnn (βw) ∂w 2 © 2004 by Chapman & Hall/CRC Q ¤P References: G. Solution for a1 = a2 = 0: 2◦ . 2 a0 1 b . ∂2w + ∂2w + a ∂w + b ∂w + ∂w 2 = 0. where the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations [(a1 x + b1 )ϕx ]x − kϕ ln(βϕ) + Cϕ = 0.6. Pukhnachov. 1◦ . C is an arbitrary constant. ∂y ∂y This is a special case of equation 5. x ψyy = bψ ln |ψ| − Cψ. ∂2w + ∂ a lnn (βw) ∂w ∂x ∂x ∂y 1◦ . ∂x2 ∂y 2 .4.9 with f (w) = a lnn (βw). w) = cw n ln(βw). V. Traveling-wave solution: 10. [(a2 y + b2 )ψy ]y − kψ ln ψ − Cψ = 0.2. 9. Multiplicative separable solution: w(x. and A. Kaptsov. ∂x2 ∂ (a1 x + b1 ) ∂w + ∂ (a2 y + b2 ) ∂w ∂y = kw ln(βw). where the function w(r) is determined by the ordinary differential equation k+1 wr = a0 w + bw ln |w|. y b1 b2 x + + 2 + 2. Andreev. Solutions for a1 = 0: A= 1 b 8 I b2 + 16a2 . ∂x2 ∂y 2 ∂x ∂y This is a special case of equation 5.8 with f (w) = 1 and g(w) = a ln n (βw). w(x.2. 8. where the functions ϕ = ϕ(x) and ψ = ψ(y) are determined by the system of ordinary differential equations k ϕxx + ϕx = bϕ ln |ϕ| + (a2 x2 + a1 x + a0 + C)ϕ. A. y) = exp Ax2 + 3◦ . 2 2 ∂x ∂y x ∂x Grad–Shafranov equation (with k = −1 and a1 = a0 = 0). k ∂w ∂2w ∂2w + + = (a2 x2 + a1 x + a0 )w + bw ln |w|.4 with f (ξ. ξ= . wrr + r = cwn ln(βw). r = x2 + y 2 . V. K. O. V. y) = ϕ(x)ψ(y). 2◦ . a1 a2 a1 a2 where the function w(ξ) is determined by the ordinary differential equation (ξwξ )ξ = kw ln(βw). x ∂x y ∂y This is a special case of equation 5. C is an arbitrary constant. Rodionov (1999). (y + B)2 + A(k + 1) − + 4 b b 2 where B is an arbitrary constant.

dw = z. w = w(z).4. G= √ 1 + B2 .4. Equations with Trigonometric Nonlinearities 1.1 with f (w) = k ln(βw). ξ ◦ 3 .1 with f (w) = kw ln w and a special case of equation 5. ∂x ∂x ∂y ∂y This is a special case of equation 5. ∂x ∂x ∂y ∂y This is a special case of equation 5. and A. 12. ∂w ∂ ∂w ∂ axn + by m = k ln(βw). = α sin(βw).3. B. Traveling-wave solution: z = Ax + By + C. C.∂w ∂ ∂w ∂ (a1 x + b1 y + c1 ) + (a2 x + b2 y + c2 ) ∂x ∂x ∂y ∂y This is a special case of equation 5.4.1. 5. ξ= (x + C1 )2 + (y + C2 )2 . F = √ 1 + B2 where A and B are arbitrary constants. = kw ln w.6 with f (w) = k ln(λw). cosh G sin A (y + Bx). where C1 and C2 are arbitrary constants and the function w = w(ξ) is determined by the ordinary differential equation 1 wξξ + wξ = α sin(βw).4.3.3. 15. 13.2 with f (w) = kw ln w. ∂x ∂x ∂y ∂y This is a special case of equation 5.3.4. 2◦ .3.4. −C2 ): w = w(ξ). Solution with central symmetry about the point (−C1 .3. 11.9 with f (x) = axn and g(y) = by m . ∂x ∂x ∂y ∂y This is a special case of equation 5. 14. and D are arbitrary constants. ∂x2 ∂y 2 This is a special case of equation 5.6 with f (w) = kw ln w and a special case of equation 5.4.4.9 with f (x) = aeβx and g(y) = beµy . Functional separable solution for α = β = 1: w(x.3. y) = 4 arctan cot A cos A (x − By). where w(z) is deﬁned implicitly by D− 2α cos(βw) β(A2 + B 2 ) −1/2 ∂2w + ∂2w 1◦ . ∂w ∂ ∂w ∂ aeβx + beµy = k ln(λw).1 with f (w) = α sin(βw). © 2004 by Chapman & Hall/CRC cosh F .4. ∂w ∂ ∂w ∂ axn + by m = kw ln w.3. ∂w ∂ ∂w ∂ aeβx + beµy = kw ln w.

∂x2 ∂y 2 This is a special case of equation 5. 3.1: R. K.2 with f (w) = α cos(βw). S ¤R References for equation 5. ∂x2 ∂y 2 4. −i ∂y ∂x 2 k 2 where i2 = −1. Andreev.4.3. Denote k = 2b Traveling-wave solutions: 2. = a sin(βw) + b sin(2βw). O. ∂2w ∂x2 + = α cos(βw). and C are arbitrary constants. Rodionov (1999). ∂2w ∂2w + = α(x2 + y 2 ) sin(βw). A. 6. 1 The substitution βw = βu + 2 π leads to an equation of the form 5. Bullough and P. 5◦ .1. y) = 4 arctan f (x)g(y) . and A. Functional separable solution (generalizes the solution of Item 3 ◦ ): w(x. z= β 2 1 − k2 ξ k+1 2 tan arctan √ . where C1 and C2 are arbitrary constants. Rubel (1993). Miller (Jr. = α(x2 + y 2 ) cos(βw).4.2 with f (w) = α sin(βw).4◦ . V. + ∂2w + ∂2w ∂2w ∂2w 5. ξ= β 2 k2 − 1 ∂2w ∂y 2 2bβ(1 − k 2 ) (x sin C1 + y cos C1 + C2 ) if |k| < 1. ∂x2 T ∂2w + ∂2w w= T w= z k+1 2 coth arctan √ . V. © 2004 by Chapman & Hall/CRC . K. = aeβx sin(λw).4 with f (w) = a sin(λw). β where the functions f = f (x) and g = g(y) are determined by the ﬁrst-order autonomous ordinary differential equations 2 fx = Af 4 + Bf 2 + C. ∂x2 ∂y 2 This is a special case of equation 5. ∂x2 ∂y 2 This is a special case of equation 5. V. J. and A. B. ∂x ∂y 2 k 2 w+w 1 w−w ∂ w ∂w = + k sin − sin .1. V. Pukhnachov. Kaptsov.1. 2bβ(k 2 − 1) (x sin C1 + y cos C1 + C2 ) if |k| > 1. J. A.3.3. gy 2 = Cg 4 + (αβ − B)g 2 + A. ∂y 2 a . Auto-B¨ cklund transformations (α = β = 1): a ∂w w+w 1 w−w ∂w = −i + k sin + sin .3.1: ∂2u ∂2u + = −α sin(βu). Caudrey (1980).4.) and L.

3. Then the functions w1 = w( x + C1 .5. y) is a solution of the equation in question. This is a special case of equation 5.1 with f (w) = k sin(βw).4.1 and 5. Solution with central symmetry about the point (−C1 . This is a special case of equation 5.3. 5. This is a special case of equation 5.6 with f (w) = k sin(λw). are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). where C1 . and D are arbitrary constants. 5.3.4. ∂2w ∂x2 + ∂2w ∂y 2 = a cos(βw) ∂w ∂x 2 + ∂w ∂y 2 .1 (see also Subsection S.2.3. where A. ∂x2 ∂y 2 This is a stationary heat equation with a nonlinear source.8 with f (w) = 1 and g(w) = a cos n (βw). 9. C2 . y + C2 ). 10. 8. For exact solutions of the original equation for some f (w). ∂w ∂ aeβx ∂x ∂x + ∂ ∂w beµy ∂y ∂y = k sin(λw). w) 2 ∂x ∂y 2 1. where C1 and C2 are arbitrary constants and the function w = w(ζ) is determined by the ordinary differential equation 1 wζζ + wζ = f (w). and β are arbitrary constants. Equations Involving Arbitrary Functions 2 2 5. 2◦ . C. ζ 4◦ .4. w2 = w(x cos β − y sin β.1 and equations 5.3.1. see Subsections 5. © 2004 by Chapman & Hall/CRC . Suppose w = w(x.4. 3◦ . y.4.1.9 with f (w) = a cos(βw).1. −C2 ): w = w(ζ).4.1. This is a special case of equation 5. 1◦ .4. ∂ ∂x axn ∂w ∂x + ∂ ∂y by m ∂w ∂y = k sin(βw). and 5. ζ= (x + C1 )2 + (y + C2 )2 . F (w) = f (w) dw.2. Equations of the Form ∂ w + ∂ w = F (x.1. x sin β + y cos β). Example 12). Traveling-wave solution in implicit form: C+ 2 F (w) 2 + B2 A −1/2 U U dw = Ax + By + D. ∂2w ∂2w + = f (w).3.2. B.7.3. ∂2w ∂x2 + ∂ ∂y a cosn (βw) ∂w ∂y = 0.

whose general solution can be written out in implicit form as V V C1 + 2F (w) −1/2 dw = C2 ζ. Polyanin and V. F (w) = f (w) dw. 2 Here. Solution with central symmetry: w = w(r). x sin β + y cos β). where β is an arbitrary constant. ∂y 2 1◦ . © 2004 by Chapman & Hall/CRC ∂2w ∂2w . 4◦ . 2 ζ = xy leads to a simpler equation of the form 5. F. = (x2 + y 2 )k f (w). F (w) = f (w) dw. z = 1 (x2 − y 2 ). ∂x2 ∂y 2 1◦ . w = w(ζ). where C1 and C2 are arbitrary constants. Suppose w = w(x. is also a solution of the equation. D. r= x2 + y 2 . x sin β + y cos β). r r ζ = xy.2. whose general solution can be written out in implicit form as C1 + 2F (w) −1/2 dw = C2 z. 5◦ . Then the function w1 = w(x cos β − y sin β. ∂z 2 ∂ζ 2 X ¤W Reference: A. where the function w = w(r) is determined by the ordinary differential equation wrr + 3◦ . ζ). the function w = w(ζ) is determined by the autonomous ordinary differential equation wζζ = f (w). y) is a solution of this equation. Here. ∂2w ∂2w + = (x2 + y 2 )f (w). is also a solution of the equation.4. where β is an arbitrary constant. The transformation w = U (z. where C1 and C2 are arbitrary constants. the function w = w(z) is determined by the autonomous ordinary differential equation wzz = f (w). Functional separable solution: w = w(z). Zaitsev (2002). y) is a solution of this equation. Self-similar solution: 1 w = r2 f (w). Then the function 3. 2◦ . Suppose w = w(x.1: ∂2U ∂2U + = f (U ).1. z = 1 (x2 − y 2 ). ∂x2 + w1 = w(x cos β − y sin β.

ζ = 1 2 ln(x2 + y 2 ). F. transformation (3) coincides with transformation (2). r r 3◦ . The transformation 1 z = 2 (x2 − y 2 ). 2.4. .1: ∂2W ∂2W + = 4β −2 f (W ). v). D. where the function w = w(r) is determined by the ordinary differential equation wrr + Y Y 1 w = r2k f (w). .1: ∂2w ∂2w + = f (w). 2.1. ∂2w ∂2w + = eβx f (w). Suppose k is an arbitrary constant (k ≠ −1). 2(k + 1)i i2 = −1 (2) leads to equation (1). for any integer k ≠ −1. k+1 ζ= 1 k+1 r sin (k + 1)ϕ . ∂z 2 ∂ζ 2 (1) For arbitrary f = f (w). . 2(k + 1) ζ= (x + iy)k+1 − (x − iy)k+1 . Zaitsev (2002). u = exp 1 2 βx 1 2 βy The transformation cos . x .1. In the general case. z= z= 3 1 3 (x − 3xy ). y for k = −1. . ∂x2 ∂y 2 w = W (u. . ζ = xy 2 2 1 3 (3x y 3 for k = 1. leads to the simpler equation (1). For k = 1. . r= x2 + y 2 . Solution with central symmetry: w = w(r). Polyanin and V. and a solution of the form w = w(z 2 + ζ 2 ). ∂u2 ∂v 2 a ¤` Reference: A. . x2 + y 2 z=− − y ) for k = 2. 4. The transformation z= 1 k+1 r cos (k + 1)ϕ . v = exp 1 2 βx sin 1 2 βy leads to a simpler equation of the form 5. D.2◦ . F. It follows from (2) that z2 + ζ 2 = 1 (x2 + y 2 )k+1 . ζ = arctan x y ζ= 2 for k = −2 x + y2 leads to a simpler equation of the form 5. Suppose k = 1. © 2004 by Chapman & Hall/CRC . k+1 Y Y (3) where x = r cos ϕ and y = r sin ϕ.4. where A and B are arbitrary constants. Zaitsev (2002). this equation admits a traveling-wave solution w = w(Az + Bζ). a ¤` Reference: A. the transformation z= (x + iy)k+1 + (x − iy)k+1 . Polyanin and V. (k + 1)2 4◦ .

∂ξ 2 ∂η 2 6. ∂2w ∂x2 + ∂2w ∂y 2 = f (x. The condition just mentioned means that the function Φ can only have simple zeros. where ε = 1 and F = F (z) is a prescribed analytic function of the complex variable z = x + iy. ψyy − a ln ψ − C ψ = 0. The transformation ξ = x cos σ − y sin σ. y) = − |βF (z)| 1 + ε sign(β)Φ(z)Φ(z) 2 ln . Φ ψ2 = Φ a exp − C 2 z z0 F dz . ∂2w ∂2w + = eax–by f (w). the bar over a symbol denotes the complex conjugate. b β= a 2 + b2 . z0 is an arbitrary point in the complex plane. the holomorphic functions ϕ = ϕ(z) and ψ = ψ(z) are given by ϕ2 = CΦ exp a 2 z z0 F dz . General solution: w(x. © 2004 by Chapman & Hall/CRC . cos σ = a/β. β 4|Φz (z)| where Φ = Φ(z) is an arbitrary analytic (holomorphic) function of the complex variable z = x + iy with nonzero derivative. y) = ln |ϕ(z)|2 + ε|ψ(z)|2 . Kh. Here. ∂2w ∂x2 + ∂2w ∂y 2 = aw ln w + f (x)w.4: ∂2w ∂2w + = eβξ f (w).1. Reference: I. 1◦ . w(x. Multiplicative separable solution: where the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations ϕxx − a ln ϕ + f (x) + C ϕ = 0. sin σ = b/β. Suppose f (x.4. η = x sin σ + y cos σ. b d ¤c 7. and Φ = Φ(z) is an arbitrary holomorphic function satisfying the condition Φ (z∗ ) = 1 aF (z∗ ) at any point z = z∗ 2 where Φ(z∗ ) = 0. and C is an arbitrary constant. Φ where |a| = 1. y)eβw .5. C ≠ 0 is any. y) = ε|F (z)|2. y) = ϕ(x)ψ(y). ∂x2 ∂y 2 √ Let us represent the exponent of e in the form ax − by = β(x cos σ − y sin σ). leads to an equation of the form 5. Another representation of the general solution with β = −2: w(x. Sabitov (2001). 2◦ .

= f (x)w ln w + af (x)y + g(x) w. Zaitsev (2002). ξ = x2 + y 2 1/2 ∂2w + ∂2w . w1 = w(x cos β − y sin β. ζ = xy. ξ 12. ∂x2 ∂y 2 1◦ . is also a solution of the equation. y) = e−ay ϕ(x). U ). w). Polyanin and V. w). 2 . w). x sin β + y cos β). where the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations ϕxx − a ln ϕ + f (x) + C ϕ = 0. ∂x2 ∂y 2 Solution: w = w(ξ). ζ = xy ∂2U ∂2U + = f (ζ. where β is an arbitrary constant. ∂2w ∂2w + = f (ax + by. where the function ϕ(x) is determined by the ordinary differential equation 9. 10. where the function w(ζ) is determined by the ordinary differential equation + wζζ = f (ζ.∂2w ∂2w + = aw ln w + f (x) + g(y) w. y) is a solution of this equation. w). ∂2w ∂2w + = f (x2 + y 2 . y) = ϕ(x)ψ(y). where the function w(ξ) is determined by the ordinary differential equation 1 wξξ + wξ = f (ξ 2 . Solution with central symmetry: w = w(ξ). ∂x2 ∂y 2 Multiplicative separable solution: w(x. D. Self-similar solution: w = w(ζ). ϕxx = f (x)ϕ ln ϕ + g(x) − a2 ϕ. ξ = ax + by. F. leads to the simpler equation 1 z = 2 (x2 − y 2 ). w). and C is an arbitrary constant. Suppose w = w(x. ∂x2 ∂y 2 1◦ . ψyy − a ln ψ + g(y) − C ψ = 0. w). 2◦ . ζ). where the function w(ξ) is determined by the ordinary differential equation (a2 + b2 )wξξ = f (ξ. . ∂z 2 ∂ζ 2 © 2004 by Chapman & Hall/CRC f ¤e Reference: A. Then the function 11. ∂x2 ∂y 2 Multiplicative separable solution: w(x. 8. = (x2 + y 2 )f (xy. The transformation ◦ ∂2w ∂2w w = U (z.

η = y/ b leads to an equation of the form 5. 2◦ . w).13. Functional separable solution: w = w(z). ∂x ∂y ∂y 2 This equation describes steady-state mass transfer with a volume chemical reaction in a translationalshear ﬂuid ﬂow. z = 1 (x2 − y 2 ). where the function w(z) is determined by the ordinary differential equation wzz = f (2z.1. The transformation leads to the simpler equation w = U (z.4. w. y. and the function w(z) is determined by the ordinary differential equation kz + a2 c1 + (k − a1 )c2 wz = a2 + (k − a1 )2 wzz − f (w). . ∂2w ∂x2 + ∂2w ∂y 2 = f w + A11 x2 + A12 xy + A22 y 2 + B1 x + B2 y . a and b are the principal thermal diffusivities (diffusion coefﬁcients) and f = f (w) is a kinetic function that deﬁnes the law of heat (substance) release.2. In the case of an incompressible ﬂuid. U ).4.1.4. ∂z 2 ∂ζ 2 14. ζ).1: ∂2U ∂2U + = f (U ) + 2A11 + 2A22 . © 2004 by Chapman & Hall/CRC ∂w + (a2 x + b2 y + c2 ) ∂w = ∂2w + ∂2w z = a2 x + (k − a1 )y.1: ∂2w ∂2w + = f (w). w). Equations of the Form a ∂ w + b ∂ w = F x. 1 z = 2 (x2 − y 2 ). ∂x2 ∂y 2 1◦ . (a1 x + b1 y + c1 ) ∂x2 w = w(z). Traveling-wave solution: 2. where k is a root of the quadratic equation k 2 − (a1 + b2 )k + a1 b2 − a2 b1 = 0. ∂2w ∂2w + = (x2 + y 2 )f (x2 – y 2 . √ √ The transformation ξ = x/ a. ∂w 2 ∂x ∂y 2 ∂x ∂y 1. ∂w . The substitution U = w + A11 x2 + A12 xy + A22 y 2 + B1 x + B2 y leads to an equation of the form 5. 2 Remark. ∂ξ 2 ∂η 2 – f (w). ∂x2 ∂y 2 2 2 5. 2 ζ = xy ∂ 2U ∂ 2 U + = f (2z. a ∂2w ∂2w +b = f (w). ∂x2 ∂y 2 This equation describes steady-state processes of heat/mass transfer or combustion in anisotropic media. Here. the equation coefﬁcients must satisfy the condition a1 + b2 = 0.

the functions ψ = ψ(x) and χ = χ(x) can be determined successively from equations (3) and (4). ξ = x2 + y 2 1/2 ∂2w ∂2w . 6.∂2w ∂w ∂w ∂2w +b + cx – cy = (bx2 + ay 2 )f (w). w). + g(x)w + h(x). 3. y) = ϕ(x)ψ(y). 2 2 ∂x ∂y ∂x ∂y Solution: w = w(z). which are linear in ψ and χ. where C is an arbitrary constant. It is apparent from the comparison of equations (2) and (3) that (3) has the particular solution ψ = ϕ(x). w). Here. If the functions f and g are proportional. the functions ϕ(x) and ψ(x) are determined by the ordinary differential equations aϕxx = f1 (x)ϕx + kϕ ln ϕ + g1 (x) + C ϕ. (2) (3) (4) χxx = gχ + f ψ 2 + h − 2aϕ. 2003) g ψ(x) = C1 ϕ(x) + C2 ϕ(x) dx . y) = ϕ(x)y 2 + ψ(x)y + χ(x). 4. ψ(x). g. the general solution to (3) is given by (see Polyanin and Zaitsev. Hence. ∂x2 ∂y 2 x ∂x y ∂y Solution with central symmetry: + w = w(ξ). respectively. where the function w(ξ) is determined by the ordinary differential equation wξξ + 5. © 2004 by Chapman & Hall/CRC . to which there corresponds solution (1) linear in y. a ∂w b ∂w + + = f (x2 + y 2 . ψxx = (4f ϕ + g)ψ. then a particular solution to 1 equation (2) is given by ϕ = − 4 g/f = const. (1) ∂2w +a ∂2w = f (x) ∂w 2 where the functions ϕ(x). ξ ∂w ∂y + kw ln w + g1 (x) + g2 (y) w. and h are omitted): ϕxx = 4f ϕ2 + gϕ. ∂x2 ∂y 2 ∂y Generalized separable solution quadratic in y: w(x. a ∂2w ∂x2 +b ∂2w ∂y 2 = f1 (x) ∂w a+b+1 wξ = f (ξ 2 . Whenever a solution ϕ = ϕ(x) of the nonlinear equation (2) is found. a where the function w(z) is determined by the autonomous ordinary differential equation wzz = f (w). bψyy = f2 (y)ψy + kψ ln ψ + g2 (y) − C ψ. ∂x Multiplicative separable solution: + f2 (y) w(x. ϕ2 (x) ϕ 0. Note that equation (2) has the trivial particular solution ϕ(x) ≡ 0. z = xy. and χ(x) are determined by the following system of ordinary differential equations (the arguments of f .

i. p ¤i b < 0. In the special case f = const. and h are proportional. B = 2 . In this case. where the functions ϕ(x) and ψ(x) are determined by the system of ordinary differential equations References: V. © 2004 by Chapman & Hall/CRC . y) = ϕ(x) + ψ(x) cosh y −b . A = 2 . 8. y) = ϕ(x) + ψ(x) A exp y −b + B exp −y −b . The books by Kamke (1977) and Polyanin and Zaitsev (2003) present a large number of exact solutions to the linear equation (5) for various f = f (x).. (5) ψxx = (2bkn + α)f + ab ψ. D. √ 1 1 w(x. h = βf ϕ = k1 . B = − 2 . 2. g. Generalized separable solutions: 7. the general solution of equation (5) is the sum of exponentials (or sine and cosine). y) = ϕ(x) + ψ(x) exp b < 0. F. y) = ϕ(x)y 2 + ψ(x)y + χ(x). h w(x. equation (3) can be rewritten as n = 1. 2◦ . (2) (3) Whenever a solution ϕ = ϕ(x) of equation (2) is found. A = 2 . (6) where the functions ϕ(x) and ψ(x) are determined by the system of ordinary differential equations b > 0. ϕxx = bf ϕ2 + 4ABψ 2 + gϕ + h. y) = ϕ(x) + ψ(x) sinh y −b . ψxx = 2bf ϕψ + gψ + abψ. g = αf . a ∂2w ∂x2 +b ∂2w ∂y 2 = f (x) ∂w ∂y 2 + g(x) ∂w ∂x ∂w + p(x)w + q2 (x)y 2 + q1 (x)y + q0 (x).∂x2 ∂y 2 ∂y ◦ 1 . A. Galaktionov (1995). ϕxx = bf ϕ2 + ψ 2 + gϕ + h. √ y −b . ϕ = k2 . Generalized separable solution (c is an arbitrary constant): √ w(x. the functions ψ = ψ(x) can be determined by solving equation (3). ψxx = (2bf ϕ + g + ab)ψ. (α. (4) then particular solutions of equation (2) are given by where k1 and k2 are roots of the quadratic equation bk 2 + αk + β = 0. which is linear in ψ. ψxx = 2bf ϕψ + gψ + abψ. V. Note two special cases of solution (6) that involve hyperbolic functions. 3◦ . g.e. These are: √ 1 1 w(x. and h are omitted): ϕxx = bf ϕ2 + gϕ + h. y) = ϕ(x) + ψ(x) cos y b + c . Generalized separable solution (generalizes the solutions of Item 1 ◦ ): √ √ w(x. ∂2w +a ∂2w = f (x) ∂w 2 + bf (x)w2 + g(x)w + h(x). Polyanin (1996). If the functions f . (1) where the functions ϕ(x) and ψ(x) are determined by the following system of ordinary differential equations (the arguments of f . β = const). ∂y There is a generalized separable solution quadratic in y: + h1 (x)y + h0 (x) w(x. Zaitsev and A.

where the constants A. where C is an arbitrary constant. ∂x2 ∂y 2 ∂x Solutions are sought in the traveling-wave form w = w(z). w(x. the functions ϕ(x) and ψ(x) are determined by the ordinary differential equations aϕxx − f1 (x) ϕx bψyy − f2 (y) ψy where C is an arbitrary constant. f (w) dw . ∂w ∂x . B. Bwz . ∂2w + ∂2w = (a1 x + b1 y + c1 ) ∂w k − g1 (x)ϕx − kϕ − h1 (x) = C. y) = ϕ(x) + ψ(y). and C are determined by solving the algebraic system of equations a1 Ak + a2 B k = A. + (a2 x + b2 y + c2 ) ∂w ∂y k + f w. © 2004 by Chapman & Hall/CRC . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). The unknown function w(z) is determined by the ordinary differential equation (A2 + B 2 )wzz = z(wz )k + f w. Awz . − g2 (y)ψy − kψ − h2 (y) = −C. y): ∂ 2 U ∂ 2U + = 0. 11. dw . y) = ϕ(x) + ψ(y). ∂w ∂y . a ∂2w ∂x2 +b ∂2w ∂y 2 = f1 (x) ∂w ∂x n + f2 (y) ∂w ∂y m + g1 (x) Additive separable solution: ∂w ∂w + g2 (y) + h1 (x) + h2 (y) + kw. (1) (2) (3) Equations (1) and (2) are ﬁrst solved for A and B and then equation (3) is used to evaluate C. ∂x2 ∂y 2 Additive separable solution: 10. − kψ = −C. c1 Ak + c2 B k = C. the functions ϕ(x) and ψ(x) are determined by the ordinary differential equations − kϕ = C. ∂2w + ∂2w = f (w) U= ∂w ∂x 2 + ∂w ∂y 2 . aϕxx − f (x) ϕx bψyy − g(y) ψy n m Here.∂x2 ∂y 2 The substitution 9. ∂x ∂y w(x. z = Ax + By + C. b1 Ak + b2 B k = B. where F (w) = exp F (w) leads to the two-dimensional Laplace equation for U = U (x. ∂x2 ∂y 2 For solutions of this linear equation. 12. a ∂2w +b ∂2w = f (x) ∂w ∂x n + g(y) ∂w ∂y m + kw. n m Here.

the function w(ξ) is determined by the ordinary differential equation wξξ + where A= (1) 4 4 − nm . Here. which deﬁnes the law of heat (substance) release of absorption. x ϕ ϕ ψy ψyy − f2 y. Here. ∂x ∂x ∂y ∂y 1◦ . ψy − kψ = −C. ∂w ∂y + kw. w.3.∂w ∂2w ∂2w +b = f1 x. f = f (x) and g = g(y) are the principal thermal diffusivities (diffusion coefﬁcients) dependent on coordinates. It is given by q C1 + 2n2 F (w) ab(2 − n)4 −1/2 dw = C2 ξ. the functions ϕ(x) and ψ(x) are determined by the ordinary differential equations aϕxx − f1 x. b ψ ψ = C. Here. 2 2 ∂x ∂y ∂x Additive separable solution: 13. axn by m w = w(ξ). 14. y) = ϕ(x)ψ(y). Heat and Mass Transfer Equations of the Form ∂ f (x) ∂w + ∂ g(y) ∂w = h(w) ∂x ∂x ∂y ∂y Equations of this form describe steady-state heat/mass transfer or combustion processes in inhomogeneous anisotropic media. h = h(w) is the kinetic function (source function). = −C. where C1 and C2 are arbitrary constants. a family of exact solutions to the original equation with arbitrary f = f (w) follows from (1). 1 ∂w 1 ∂w = f1 x. w = w(x) or w = w(y).4. Functional separable solution for n ≠ 2 and m ≠ 2: 1. ∂2w ∂2w where C is an arbitrary constant. 5. ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m A w = Bf (w). (2 − n)(2 − m) ab(2 − n)2 (2 − m)2 For m = 4/n. are not considered in this subsection. = f (w). the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations a ϕ ϕxx − f1 x. bψyy − f2 y. w + f2 y. B= . ξ ξ 1/2 ∂ ∂w + ∂ ∂w . where C is an arbitrary constant. y) = ϕ(x) + ψ(y). a +b Here. w(x. a + f2 y. ∂x2 ∂y 2 w ∂x w ∂y Multiplicative separable solution: w(x. The simple solutions dependent on a single coordinate. © 2004 by Chapman & Hall/CRC . ϕx − kϕ = C. F (w) = f (w) dw.

The substitution ζ = ξ 1−A brings (1) to the generalized Emden–Fowler equation 2A B ζ 1−A f (w). s ¤r Reference: V. Polyanin (1996). ∂ ∂ n ∂w m ∂w a |x| + c + b |y| + = f (w). F (w) = f (w) dw. (1) s ¤r .4. w = w(ξ). 2 ∂x ∂y ∂y Functional separable solution for µ ≠ 0: 1/2 4.3. D.3. Reference: V. ∂y ∂y The substitution ζ = |y| leads to an equation of the form 5. F. ξ = bµ2 (x + C1 )2 + 4ae−µy where C1 is an arbitrary constant and the function w(ξ) is determined by the autonomous ordinary differential equation 1 f (w). w = w(ξ). © 2004 by Chapman & Hall/CRC u t 2. ξ abβ 2 µ2 The substitution ζ = ξ 2 brings (1) to the generalized Emden–Fowler equation wζζ = 1 Aζ −1 f (w). Polyanin (1996). F.1: ∂w ∂ ∂w ∂ aζ n + bη m = f (w).4. a = f (w). a . Zaitsev and A. 5. C2 + abµ2 where C2 and C3 are arbitrary constants.3. Zaitsev and A. η = y + s leads to an equation of the form 5.4. = f (w). ∂w ∂ ∂w ∂ a(x + c)n + b(y + )m = f (w).4. s = const) can be found in Polyanin and Zaitsev (2003).1: ∂w ∂ ∂w ∂ aζ n + bη m = f (w). (2) wζζ = (1 − A)2 A large number of exact solutions to equation (2) for various f = f (w) can be found in Polyanin and Zaitsev (2003). 4 whose solutions with f (w) = (kw + s)−1 and f (w) = (kw + s)−2 (k. ∂ζ ∂ζ ∂η ∂η ∂ ∂w ∂2w + beµy = f (w). ∂ζ ∂ζ ∂η ∂η t 3. ∂x ∂x ∂y ∂y The transformation ζ = |x| + c. wξξ = abµ2 The general solution of this equation with arbitrary kinetic function f = f (w) is deﬁned implicitly by −1/2 2 F (w) dw = C3 ξ.2◦ . ξ = bµ2 e−βx + aβ 2 e−µy where the function w(ξ) is determined by the ordinary differential equation 4 1 A= . ∂x2 beµ|y| 6. η = |y| + s leads to an equation of the form 5. D. 1/2 ∂2w + ∂ ∂w ∂w ∂ ∂w ∂ aeβx + beµy ∂x ∂x ∂y ∂y Functional separable solution for βµ ≠ 0: . ∂x ∂x ∂y ∂y The transformation ζ = x + c. wξξ − wξ = Af (w).

ξ = (a2 b)−1/3 x + (ab2 )−1/3 y + (a2 b)−2/3 c + (ab2 )−2/3 s. 2 (2 − n)2 2−n ξ abµ = kw ln w. y. ∂ ∂x aeβ|x| ∂w ∂x + ∂ ∂y beµ|y| ∂w ∂y = f (w).4. (ay + c) ∂2w ∂x2 + (bx + ) ∂2w ∂y 2 = f (w). Zaitsev (2002). This is a special case of equation 5. and C is an arbitrary constant. w ¤v ∂w ∂y Multiplicative separable solution: Reference: A. D. 10. The transformation ζ = |x|. This equation can be rewritten in the divergence form ∂w ∂ ∂w ∂ (ay + c) + (bx + s) ∂x ∂x ∂y ∂y For ab ≠ 0. Polyanin and V.4. h1 (x) = b. 5.4.4. ∂ ∂x f (x) ∂w ∂x + ∂ ∂y g(y) ∂w ∂y = aw ln w + bw. y. y. w(x. 8. Functional separable solution for n ≠ 2 and µ ≠ 0: w = w(ξ). [g(y)ψy ]y = kψ ln ψ − Cψ.3. F. y) = ϕ(x)ψ(y).7. where the function w(ξ) is determined by the ordinary differential equation wξξ + 9. there is an exact solution of the form w = w(ξ). ξ = bµ2 x2−n + a(2 − n)2 e−µy 1/2 . where the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations [f (x)ϕx ]x = kϕ ln ϕ + Cϕ. = f (w).4. © 2004 by Chapman & Hall/CRC . w) ∂w + ∂ g(x. where the function w(ξ) is determined by the ordinary differential equation ξwξξ = f (w). Equations of the Form ∂ f (x.6.6 with k = a. ∂ ∂x f (x) ∂w ∂x + ∂ ∂y g(y) 4 n 1 wξ = f (w). and h2 (y) = 0. w) ∂x ∂x ∂y ∂y x 1. w) ∂w = h(x. η = |y| leads to an equation of the form 5. ∂ ∂x axn ∂w ∂x + ∂ ∂y beµy ∂w ∂y = f (w).

Generalized separable solution linear in y: w(x. 4. ψxx + 6f ϕψ = 0. Since ψ = ϕ(x) is a particular solution to equation (2). C1 . B. 3. For f = aeλx . f (x)w + g(x) ∂x2 ∂y ∂y 1◦ . ψ = ψ(x). B. where the functions ϕ = ϕ(x). C2 . = f (w). y) = (Ax + B)y − x x0 ∂2w (x − t)(At + B)2 f (t) dt + C1 x + C2 . w = w(z). (1) (2) (3) where the constants A. The nonlinear equation (1) is treated independently from the others. f (y) ∂w = 0. 2◦ . √ ∂y w + a ∂y √ The substitution U = w + a leads to the equation ∂x2 + ∂ ∂U ∂ U ∂x ∂x + ∂ ∂U f (y) ∂y ∂y = 0. y) = ϕ(x)y 2 + ψ(x)y + χ(x). y) = ϕ(y)x + ψ(y). z = Ax + By + C. c1 A2 + c2 B 2 = C. © 2004 by Chapman & Hall/CRC . ϕ2 (x) which has a generalized separable solution of the form U (x. its solution can be expressed in terms of elliptic integrals. This solution is degenerate. and x0 are arbitrary constants. y) = ϕ(y)x2 + ψ(y)x + χ(y). where A. a particular solution to (1) is given 2 by ϕ = − λ e−λx . and C are determined by solving the algebraic system of equations a1 A2 + a2 B 2 = A. and χ = χ(x) are determined by the system of ordinary differential equations ϕxx + 6f ϕ2 = 0. (1) (2) (3) χxx + 2f ϕχ + 2ϕg + f ψ 2 = 0. Equations (2) and (3) are solved successively (these are linear in their respective 6a unknowns). ∂2w dx . U (x. b1 A2 + b2 B 2 = B. The unknown function w(z) is determined by the ordinary differential equation zwzz + (Aa1 + Bb2 )wz = f (w). ∂ ∂w + = 0.∂w ∂ ∂w ∂ (a1 x + b1 y + c1 ) + (a2 x + b2 y + c2 ) ∂x ∂x ∂y ∂y Solutions are sought in the traveling wave form 2. Equations (1) and (2) are ﬁrst solved for A and B and then equation (3) is used to evaluate C. 2003) ψ(x) = C1 ϕ(x) + C2 ϕ(x) where C1 and C2 are arbitrary constants. For f ≡ const. the general solution is expressed as (see Polyanin and Zaitsev. Generalized separable solution quadratic in y: w(x.

Suppose w(x. Then the function w1 = w(C1 x + C2 . ∂ ∂x f (w) ∂w ∂x + ∂ ∂y g(w) ∂w ∂y = 0. y) = ϕ(x)y 2 + ψ(x)y + χ(x). C2 . ∂x2 ∂y 2 1◦ . where A. and C3 are arbitrary constants. 3◦ . y) is a solution of this equation. where C1 . where the function w(z) is determined by the ordinary differential equation [z 2 + f (w)]wzz + 2zwz = 0. and C2 are arbitrary constants. C1 .5. y) = ϕ(x)y + ψ(x). w(x. This is a stationary anisotropic heat (diffusion) equation. ∂w ∂ f (x) ∂x ∂x + ∂ ∂w g(y) ∂y ∂y = kw ln w + h1 (x) + h2 (y) w. Traveling-wave solution in implicit form: A2 f (w) + B 2 g(w) dw = C1 (Ax + By) + C2 . ∂2w ∂2w + f (w) = 0. Self-similar solution: w = w(z). Here. f (w) and g(w) are the principal thermal diffusivities (diffusion coefﬁcients). the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations e−ϕ [f eϕ ϕx ]x − kϕ − h1 (x) = C. 8. z = y/x. where C is an arbitrary constant. 6. There are generalized separable solutions linear and quadratic in y: w(x. ∂ ∂x f (x) ∂w ∂x + ∂ ∂y g(x)w + h(x) ∂w ∂y = 0. C2 . y) is a solution of the equation in question. 2◦ . Then the functions w1 = w(C1 x + C2 . are also solutions of the equation. 7. C1 y + C3 ). 2◦ . y) = exp ϕ(x) + ψ(y) . C1 y + C3 ). where C1 . Degenerate solution: w = C1 xy + C2 x + C3 y + C4 . Suppose w(x. y y Multiplicative separable solution: w(x. and C3 are arbitrary constants. B. © 2004 by Chapman & Hall/CRC . is also a solution of the equation. 1◦ . e−ψ [geψ ψy ]y − kψ − h2 (y) = −C.

y= 1 C1 v 2 + C3 v − 2 F (w) = f (w) dw. y+β where the function w(ζ) is determined by the ordinary differential equation [f (w)wζ ]ζ + [ζ 2 g(w)wζ ]ζ = 0. Solution in parametric form: x = [C1 F (w) + C2 ]v 2 + C3 F (w) + C4 − 2 y= 1 C1 v 3 + C3 v − 2v 3 f (w) g(w)[C1 F (w) + C2 ] dw dw. © 2004 by Chapman & Hall/CRC 1 dϕ d . . . . 7◦ . g(w)[C1 F (w) + C2 ] dw + C5 . . ζ= x+α . 6◦ . y = [C2 sin(λv) − C1 cos(λv)] λ f (w) w where C1 . . . 1 1 H (w) + C4 . y = −[2C1 G(w) + C3 ]v − C2 G(w) + C5 . G(w) = g(w) dw. . y = (C1 eλv − C2 e−λv ) λ f (w) w where C1 . C4 and λ are arbitrary constants. . . dw f (w) dw (3) . Solution in parametric form: x = [C1 F (w) + C2 ]v + C3 F (w) + C4 . and the differential operator Lf is deﬁned as Lf [ϕ] ≡ 8◦ . (1) Integrating (1) and taking w to be the independent variable. Self-similar solution (α and β are arbitrary constants): w = w(ζ). where C1 . 5◦ . Solution in parametric form: x = C1 v 2 + C2 v − f (w)[2C1 G(w) + C3 ] dw + C4 . .3◦ . and the function Z = Z(w) is determined by the ordinary differential equation Lf [Z] − λ2 g(w)Z = 0. A large number of exact solutions to equation (2) for various f = f (w) and g = g(w) can be found in Polyanin and Zaitsev (2003). C4 and λ are arbitrary constants. . 4◦ . g(w)[C1 F (w) + C2 ] dw + C5 . one obtains a Riccati equation for ζ = ζ(w): Cζw = g(w)ζ 2 + f (w). 1 1 Z (w) + C4 . the function H = H(w) is determined by the ordinary differential equation Lf [H] + λ2 g(w)H = 0. Solution in parametric form: x = (C1 eλv + C2 e−λv )H(w) + C3 . (2) where C is an arbitrary constant. C5 are arbitrary constants. . Solution in parametric form: x = [C1 sin(λv) + C2 cos(λv)]Z(w) + C3 .

Solution in parametric form: x= 1 C1 v 3 + C3 v − 2v 3 f (w)[C1 G(w) + C2 ] dw + C5 . Solution in parametric form: x= 1 C1 v 2 + C3 v − 2 f (w)[C1 G(w) + C2 ] dw + C5 . y = C 1 v 2 + C2 v − 10◦ . the function H = H(w) is determined by the ordinary differential equation Lg [H] + λ2 f (w)H = 0. (6) f (w) ∂v ∂w ∂v ∂w Eliminating y yields the following linear equation for x = x(w. x = − [C2 sin(λv) − C1 cos(λv)] λ g(w) w y = [C1 sin(λv) + C2 cos(λv)]Z(w) + C4 . and the differential operator Lg is deﬁned by (3) with f (w) = g(w). ∂y ∂x (4) (5) x = x(w. where C1 . −g(w) = . 13◦ . Solution in parametric form: x = [2C1 F (w) + C3 ]v + C2 F (w) + C5 . brings (4) to the linear system ∂x ∂x ∂y ∂y = . x = − (C1 eλv − C2 e−λv ) λ g(w) w y = (C1 eλv + C2 e−λv )H(w) + C4 . Solution in parametric form: 1 1 H (w) + C3 . C2 . where C1 . 11◦ . v): 1 ∂x ∂2x ∂ + g(w) 2 = 0. C3 . v). ∂x ∂y −g(w) ∂w ∂v = . and the differential operator Lg is deﬁned by (3) with f (w) = g(w).9◦ . y = −[C1 G(w) + C2 ]v 2 − C3 G(w) + C4 + 2 12◦ . ∂w f (w) ∂w ∂v © 2004 by Chapman & Hall/CRC (7) . and λ are arbitrary constants. v). where w. y as the dependent ones. F (w) = f (w) dw. and λ are arbitrary constants. 14◦ . the function Z = Z(w) is determined by the ordinary differential equation Lg [Z] − λ2 f (w)Z = 0. The original equation can be represented as the sum of the equations f (w) The hodograph transformation ∂w ∂v = . C2 . y = −[C1 G(w) + C2 ]v − C3 G(w) + C4 . Solution in parametric form: 1 1 Z (w) + C3 . y = y(w. g(w) f (w)[C1 G(w) + C2 ] dw dw. G(w) = g(w) dw. v are treated as the independent variables and x. g(w)[2C1 F (w) + C3 ] dw + C4 . C3 .

Then y = Φ(w. The function y = y(w. Solutions to equation (7) with odd powers of v: n x= k=0 ψk (w)v 2k+1 . 16◦ . . v) is found in the form w v ∂x 1 ∂x (w. Likewise. . v. g) be a solution to equation (7). f (w) g(w)ψk (w) dw dw. one can ﬁrst construct an exact solution to the linear equation (8) for y = y(w. g). f (w) g(w)ϕk (w) dw dw. . v) is deﬁned by (9) and. (11) where the functions ψk = ψk (w) are determined by the recurrence relations ψn (w) = An F (w) + Bn . f . 1). First. in conjunction with relation (10). ϕk−1 (w) = Ak F (w) + Bk − 2k(2k − 1) where the Ak and Bk are arbitrary constants (k = n. this solution is substituted into the linear system (6). in conjunction with relation (11). We have 1 ∂y ∂2y ∂ + f (w) 2 = 0. Remark 2. F (w) = f (w) dw. 1). gives a solution to the original nonlinear equation in parametric form. v0 ) dη. ∂w g(w) ∂w ∂v (8) The procedure for constructing exact solutions to the original equation consists of the following two stages. © 2004 by Chapman & Hall/CRC . v) and then ﬁnd x = x(w. ∂v where w0 and v0 are any numbers. 15◦ . v) is deﬁned by (9) and. g. f . from which the function y = y(w. −f ) and y = Φ(w. v0 ) dη. v. v). Then the functions x = Ψ(w. (9) y= f (w) ∂w ∂v w0 v0 where w0 and v0 are any numbers.Likewise. v. f ) solves equation (8). (10) where the functions ϕk = ϕk (w) are determined by the recurrence relations ϕn (w) = An F (w) + Bn . Then. ξ) dξ − g(η) (η. v. Let x = Φ(w. Let x = Φ(w. . y = Ψ(w. v) from system (6). Solutions to equation (7) with even powers of v: n x= k=0 ϕk (w)v 2k . v) from (6) in the form x=− v v0 1 ∂y (w. v. Remark 1. −g. −f ) also solve this system. The function y = y(w. f . . we can obtain another linear equation for y = y(w. The thus obtained expressions of (5) deﬁne a solution to the original equation in parametric form. ξ) dξ + g(w) ∂w w f (η) w0 ∂y (η. gives a solution to the original nonlinear equation in parametric form. F (w) = f (w) dw. one ﬁnds an exact solution to the linear equation (7) for x = x(w. ψk−1 (w) = Ak F (w) + Bk − 2k(2k + 1) where the Ak and Bk are arbitrary constants (k = n. . −g. g) be a solution to system of equations (6). . . v.

Traveling-wave solution: 9. see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). Polyanin (2001).5. Taking w to be the independent variable. and the function w(z) is determined by the ordinary differential equation [ϕ(w)wz ]z = [kz + a2 c1 + (k − a1 )c2 ]wz . where the constants A and B are determined by solving the algebraic system of equations The desired function w(ξ) is determined by the ﬁrst-order ordinary differential equation (C is an arbitrary constant): ξ + A2 f (w) + B 2 g(w) wξ = C. Suppose w(x.8: V. 10. ¯ leads to the Laplace equation u= f (w) dw ∂ 2 u ∂ 2u + = 0. 5.17◦ .4. are also solutions of the equation.4. y) is a solution of this equation. ∂ x2 ∂y 2 ¯ For solutions of this linear equation. A. a1 x + b1 y + f (w) ∂w + ∂ a2 x + b2 y + g(w) ξ = Ax + By. we obtain a ﬁrst-order linear equation for ξ = ξ(w): Cξw = ξ + A2 f (w) + B 2 g(w). D. ∂ f (w) ∂w + ∂ ∂w = (a1 x + b1 y + c1 ) ∂w + (a2 x + b2 y + c2 ) ∂w Remark 1. equation coefﬁcients must satisfy the condition a1 + b2 = 0. where k is a root of the quadratic equation k 2 − (a1 + b2 )k + a1 b2 − a2 b1 = 0. In the case of an incompressible ﬂuid. Zaitsev (2002). ∂x ∂x ∂y ∂y ∂x ∂y This equation describes steady-state anisotropic heat/mass transfer in a translational-shear ﬂuid ﬂow. ∂w ∂y = 0. the transformation x = kx. Then the functions ∂2w + aw4 ∂2w w1 = C1 w 2 C1 x + C2 . ϕ(w) = a2 f (w) + (k − a1 )2 g(w). In the special case g(w) = k 2 f (w). ¤ References for equation 5. Zaitsev and A. © 2004 by Chapman & Hall/CRC . Polyanin and V. D. The above remains the same if an arbitrary function. . y . Other Equations 1. = f (y)w5 . g(w) w = w(z). ∂ ∂x ∂x ∂y Solutions are sought in the traveling-wave form w = w(ξ).4. is added to the right-hand side of the original equation. F. b1 A2 + b2 B 2 = B. h(w). F. 2 z = a2 x + (k − a1 )y. where C1 and C2 are arbitrary constants. a1 A2 + a2 B 2 = A. Remark 2. ∂x2 ∂y 2 1◦ .

and D are arbitrary constants. λ. η = ζx −2β−1 (self-similar solution). where A. ∂2w ∂x2 +a ∂ eλw ∂w = f (y)eλw + g(x). a(ψ n ψy )y − f (y)ψ n+1 − Cψ = 0. a(eλψ ψy )y − f (y)eλψ − C = 0. ζ) = g(x)h(ζ) ξ(x. equation (2) has exact solutions with the following structures. bringing it to the form ζ= ∂2ξ ∂2ξ + aξ 4 2 = 0. (multiplicative separable solution). Furthermore. y) = ϕ(x)y 2 + ψ(x)y + χ(x). F. and β are arbitrary constants. ϕxx − g(x) + Ceλϕ = 0. with the change of variable U = eλψ . ∂2w ∂x2 + f3 (x)w + f2 (x)y 2 + f1 (x)y + f0 (x) ∂2w = g2 (x) ∂2w +a ∂ ∂w ∂y ∂y Additive separable solution: where the functions ψ = ψ(y) and ϕ = ϕ(x) are determined by the ordinary differential equations ∂w ∂y + 2 w(x. Reference: V. © 2004 by Chapman & Hall/CRC 3 (x)w 2 (x)y 2 + 1 (x)y + 0 (x). ∂x2 ∂ζ This equation is independent of f explicitly and has a degenerate solution ξ(x. Zaitsev and A. . where k. ζ) = Axζ + Bζ + Cx + D. ζ) = x ϕ(η). ∂y 2 ∂w ∂w + h1 (x)y + h0 (x) + + g1 (x) ∂x ∂y There is a generalized separable solution quadratic in y: 4. = f (y)wn+1 + g(x)w. D. ζ) = ξ(kx + λζ) ξ(x. β ¤ The transformation (2) (traveling-wave solution). (1) w dy . ∂y ∂y Multiplicative separable solution: w = ϕ(x)ψ(y). 2. The second equation can be reduced.2◦ . Let u = u(y) be any nontrivial particular solution of the second-order linear ordinary differential equation auyy − f (y)u = 0. Polyanin (1996). ξ= 2 u u simpliﬁes the original equation considerably. 3. and C is an arbitrary constant. w = ϕ(x) + ψ(y). B. to the linear equation aUyy − λf (y)U − λC = 0. ∂x2 wn where the functions ψ = ψ(y) and ϕ = ϕ(x) are determined by the ordinary differential equations (C is an arbitrary constant) ϕxx − g(x)ϕ + Cϕn+1 = 0. C. for example: ξ(x.

Solution of equation (1) with B = 0 and arbitrary f = f (w) in implicit form: C1 + −1/2 dw = C2 ξ. 4 C1 + 2 F (w) A −1/2 dw = C2 ξ. Polyanin (1996). © 2004 by Chapman & Hall/CRC . there is a functional separable solution of the form 8. 7. there is a functional separable solution of the form w = w(ξ). Zaitsev and A. F (w) = f (w) dw. the function w = w(ξ) is determined by the ordinary differential equation Awξξ + where A = 1 abβ 2 µ2 . aeβx + beµy w = w(ξ). axn ∂2w ∂x2 + by m ∂2w ∂y 2 + kxn–1 f (w) ∂w + y m–1 f (w) ∂w 1/2 . 2 2 ∂x ∂y ∂x ∂y Functional separable solution for βµ ≠ 0: w = w(ξ). ∂2w ∂w ∂w ∂2w + beµy + keβx + eµy = f (w). ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m 5. 4 Solution of equation (1) with B = 0 and arbitrary f = f (w) in implicit form: B w = f (w). where C1 and C2 are arbitrary constants. axn 1/2 . 2 2 ∂x ∂y ∂x ∂y Functional separable solution for n ≠ 2 and m ≠ 2: w = w(ξ). aeβx ξ = b(2 − m)2 x2−n + a(2 − n)2 y 2−m 6. where C1 and C2 are arbitrary constants. D. ∂x2 ∂y 2 ∂x ∂y For βµ ≠ 0. the function w(ξ) is determined by the ordinary differential equation Awξξ + where B = 1 (2 − n)(2 − m) ab(3nm − 4n − 4m + 4) + 2bk(2 − m) + 2as(2 − n) . 4 2 F (w) A B w = f (w). Here. ∂x ∂y For n ≠ 2 and m ≠ 2. ξ ξ (1) A = 1 ab(2 − n)2 (2 − m)2 . ξ = bµ2 e−βx + aβ 2 e−µy 1/2 ∂2w ∂2w . ξ = bµ2 e−βx + aβ 2 e−µy 1/2 . F. ∂w ∂w + keβx f (w) + eµy f (w) = g(w). F (w) = f (w) dw. = g(w). ξ ξ (1) 1 B = 4 βµ(3abβµ − 2bkµ − 2asβ). Here.∂2w ∂w ∂w ∂2w + by m + kxn–1 + y m–1 = f (w). ¤ Reference: V.

where β = (a2 b)−1/3 . the function w(ξ) is determined by the ordinary differential equation ξwξξ = f w. c1 A + c2 B = C. where C1 and C2 are arbitrary constants.∂2w ∂w ∂w ∂2w + beβy + kxn–1 + eβy = f (w). . the function w(ξ) is determined by the ordinary differential equation Awξξ + where A = 1 abβ 2 (2 − n)2 . 12. F (w) = f (w) dw. The desired function w(ξ) is determined by the ordinary differential equation ξwξξ = f w. Bwξ . F. Awξ . (ay + c) ξ = bβ 2 x2−n + a(2 − n)2 e−βy 1/2 ∂2w + beβy ∂2w + kxn–1 f (w) ∂w + eβy f (w) ∂w . there is a functional separable solution of the form 10. Polyanin (1996). ∂x2 ∂y 2 ∂x ∂y For β ≠ 0 and n ≠ 2. ∂w ∂w ∂2w ∂2w + (bx + ) = f w. µwξ . Zaitsev and A. ∂x2 ∂y 2 Solutions are sought in the traveling-wave form w = w(ξ). axn 1/2 . ξ = Ax + By + C. 2 2 ∂x ∂y ∂x ∂y Functional separable solution β ≠ 0 and n ≠ 2: w = w(ξ). ∂w ∂y . 11. ξ = (a2 b)−1/3 x + (ab2 )−1/3 y + (a2 b)−2/3 c + (ab2 )−2/3 s. ¤ Reference: V. Equations (1) and (2) are ﬁrst solved for A and B and then equation (3) is used to evaluate C. µ = (ab2 )−1/3 . and C are determined by solving the algebraic system of equations a1 A2 + a2 B 2 = A. Here. 2 2 2 2 (1) (2) (3) © 2004 by Chapman & Hall/CRC . axn w = w(ξ). Solution of equation (1) with B = 0 and arbitrary f = f (w) in implicit form: C1 + 2 F (w) A −1/2 dw = C2 ξ. b1 A + b2 B = B. ∂x2 ∂y 2 ∂x ∂y Functional separable solution for ab ≠ 0: w = w(ξ). βwξ . 4 B w = f (w). Here. (a1 x + b1 y + c1 ) ∂2w + (a2 x + b2 y + c2 ) ∂2w = f w. ξ = bβ 2 x2−n + a(2 − n)2 e−βy 9. D. ∂w ∂x . where the constants A. = g(w). . B. ξ ξ (1) 1 B = 4 β(2 − n) abβ(4 − 3n) + 2bkβ − 2as(2 − n) .

where C1 and C2 are arbitrary constants. where the prime stands for the differentiation with respect to x. is also a solution of the equation. and φ(x) is determined by the second-order linear ordinary differential equation φxx + f (x)φx + h(x)φ = 0. a1 A2 + a2 B 2 = A. a1 x + b1 y + f (w) ∂2w ∂x2 + a2 x + b2 y + g(w) w = w(ξ). C1 y + C2 + φ(x). w) = ξ + A2 f (w) + B 2 g(w). and C are determined by solving the algebraic system of equations and the function w(ξ) is determined by the ordinary differential equation ϕ(ξ. y) = ϕ1 (x) + ϕ2 (x)y 3/2 + ϕ3 (x)y 3 . A2 b1 + B 2 b2 = B. where the constants A. ∂w ∂x . ∂w ∂w ∂ 2 w ∂2w + f (x) + g(x) + h(x)w = 0. f1 (x) Here. Generalized separable solution: w(x. Traveling-wave solution: ξ = Ax + By. 13. w = w(ξ). A2 a1 + B 2 a2 = A. the functions ϕ(x) and ψ(y) are determined by the ordinary differential equations f1 (x)ϕxx = g1 (x)ϕx + kϕ ln ϕ + h1 (x) + C ϕ. 16. ∂w ∂y . Then the function −3 w1 = C1 w x. and C are determined by solving the algebraic system of equations and the function w(ξ) is determined by the ordinary differential equation ξ + A2 f (w) + B 2 g(w) wξξ = h w. w)wξ ξ = h w. 2 2 ∂x ∂y ∂x ∂y Multiplicative separable solution: w(x. ∂w ∂x . y) is a solution of this equation. Bwξ . f2 (y)ψyy = g2 (y)ψy + kψ ln ψ + h2 (y) − C ψ. ϕ(ξ. where the functions ϕk = ϕk (x) are determined by the system of ordinary differential equations 9 ϕ1 + f (x)ϕ1 + 8 g(x)ϕ2 + h(x)ϕ1 = 0. where C is an arbitrary constant. ∂2w ∂y 2 = h w. y) = ϕ(x)ψ(y). Suppose w(x. B. b1 A2 + b2 B 2 = B. ∂w ∂y . where the constants A. ∂w ∂y = h w. 2 ϕ2 + f (x)ϕ2 + ϕ3 + f (x)ϕ3 + 45 4 g(x)ϕ2 ϕ3 + h(x)ϕ2 = 18g(x)ϕ2 + h(x)ϕ3 = 0. Awξ . 2 ∂x ∂x ∂y ∂y 2 1◦ . 3 0. ∂x Traveling-wave solution: 15. B. Bwξ . © 2004 by Chapman & Hall/CRC . Awξ .∂2w ∂w ∂w ∂2w + f2 (y) = g1 (x) + g2 (y) + kw ln w + h1 (x) + h2 (y) w. 14. 2◦ . ∂ a1 x + b1 y + f (w) ∂w ∂x + ∂ ∂y a2 x + b2 y + g(w) ξ = Ax + By.

y) = ψ1 (x) + ψ2 (x)y + ψ3 (x)y 2 + ψ4 (x)y 3 . = 0. Here. ψ3 + f (x)ψ3 + 18g(x)ψ3 ψ4 + h(x)ϕ3 = 0. The Legendre transformation u(ξ. η) can be found in Polyanin (2002). where the functions ψk = ψk (x) are determined by the system of ordinary differential equations ψ1 + f (x)ψ1 + 2g(x)ψ2 ψ3 + h(x)ϕ1 = 0. Suppose w(x. the functions ξ = ξ(x) and η = η(x) are determined by the system of ordinary differential equations ηxx + f (x)ηx + ag(x)η 2 + h(x)η = 0. Then the function 17. y) is a solution of this equation. ξ= ∂w .3◦ . η) = xξ + yη − w(x. and C4 are arbitrary constants. ξxx + f (x)ξx + bg(x)η 2 + h(x)ξ = 0. ∂x η= ∂w . leads to the linear equation ∂2u ∂ 2u + f (ξ. where C1 . 2◦ . 4◦ . ∂η 2 ∂ξ Exact solutions of this equation for some f (ξ. y) = ξ(x) + η(x)θ(y). is also a solution of the equation. and ξ and η are the new independent variables. where a and b are arbitrary constants. whose solution can be written out in implicit form. C2 . ∂y where u is the new dependent variable. Generalized separable solution: w(x. 2 ψ2 + f (x)ψ2 + 2g(x)(2ψ3 + 3ψ2 ψ4 ) + h(x)ϕ2 = 0. ∂x2 ∂x ∂y ∂y 2 1◦ . Generalized separable solution cubic in y: w(x. η) 2 = 0. © 2004 by Chapman & Hall/CRC . and the function θ = θ(y) is determined by the autonomous ordinary differential equation θy θyy = aθ + b. ∂2w +f −1 w1 = C1 w C1 x + C2 . C 1 y + C 3 + C4 . C4 . y). ∂w ∂w ∂ 2 w . 2 ψ4 + f (x)ψ4 + 18g(x)ψ4 + h(x)ϕ4 = 0.

a(2 − n)2 b(2 − m)2 c(2 − k)2 where the function w(r) is determined by the ordinary differential equation 2 2 2 A wr = swp . w(x. z) = x p−1 F (ρ1 . is also a solution of the equation. y.1. ζ 2 = a(2 − n)2 c(2 − k)2 ξ2 = n−2 w(x. z). y.1.3 with f (w) = sw p . where C1 is an arbitrary constant. ρ1 = yx 2−m . Functional separable solution for n ≠ 2. Equations with Three Space Variables Involving Power-Law Nonlinearities 6.Chapter 6 Elliptic Equations with Three or More Space Variables 6. C12−k z . z) = U (ξ. ζ). There are “two-dimensional” solutions of the following forms: wrr + w(x. z) is a solution of the equation in question. y 2−m x2−n + . n−2 ρ2 = zx 2−k . ∂ axn ∂w + ∂ by m ∂w + ∂ cz k ∂w = wp . p−1 p−1 1. 3◦ . z) = W (y.1.3. m ≠ 2. a(2 − n)2 b(2 − m)2 z 2−k y 2−m + . and k ≠ 2 (generalizes the solution of Item 2 ◦ ): w = w(r). y. w(x. n−2 © 2004 by Chapman & Hall/CRC . Then the function w1 = C1 w C12−n x. z) = V (x. y. 2◦ . Equations of the Form ∂ f (x) ∂w + ∂ g(y) ∂w + ∂ h(z) ∂w = aw p ∂x ∂x ∂y ∂y ∂z ∂z ∂x ∂x ∂y ∂y ∂z ∂z This is a special case of equation 6. k ≠ 2. y. C12−m y. r2 = 4 y 2−m z 2−k x2−n + + . Suppose w(x.1. 1◦ . η). A= + + − 1. η 2 = b(2 − m)2 c(2 − k)2 z 2−k x2−n + . r 2−n 2−m 2−k 4◦ . m ≠ 2. and p ≠ 1: 1 w= s(1 − p) p 1 1 1 + + + 1−p 2−n 2−m 2−k 1 p−1 1 1−p p−1 y 2−m z 2−k x2−n + + a(2 − n)2 b(2 − m)2 c(2 − k)2 . Solution for n ≠ 2. ρ2 ).

3◦ . ¡ 2−n ln x. η 2 = b(2 − m)2 cλ2 e−λz x2−n + .6 with f (w) = sw . y. and p ≠ 1: 1 w= s(p − 1) p 1 + 1−p 2−n 1 p−1 1 1−p e−βy e−λz x2−n + + a(2 − n)2 bβ 2 cλ2 .406 2. w(x. z) is a solution of the equation in question.1. z + ln C1 . y. Then the function w1 = C1 w C12−n x. z) = V (x. Suppose w(x. 3◦ . z + 2◦ .3. Solution for n ≠ 2. z) = x p−1 F (ρ1 . is also a solution of the equation. Then the function w1 = C1 w C12−n x. y. z) = U (ξ.5 with f (w) = sw . y 2−m x2−n + . y + 1−p 1−p ln C1 . λ ≠ 0. ζ). λ where C1 is an arbitrary constant. β ≠ 0. η). z) = W (y. z) is a solution of the equation in question. λ ≠ 0. β ≠ 0. Functional separable solution for n ≠ 2. is also a solution of the equation. ¡ ∂w ∂ ∂w ∂ ∂w ∂ axn + by m + ceλz ∂x ∂x ∂y ∂y ∂z ∂z p This is a special case of equation 6. Functional separable solution for n ≠ 2. λ ∂w ∂ ∂w ∂ ∂w ∂ axn + beβy + ceλz ∂x ∂x ∂y ∂y ∂z ∂z p This is a special case of equation 6. ζ 2 = a(2 − n)2 cλ2 ξ2 = n−2 w(x. a(2 − n)2 bβ 2 cλ2 © 2004 by Chapman & Hall/CRC . β λ where C1 is an arbitrary constant. 1 1−p y 2−m e−λz x2−n + + 2 2 a(2 − n) b(2 − m) cλ2 . 2◦ . y. z). m ≠ 2. r2 = 4 e−βy e−λz x2−n + + . and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). ρ1 = yx 2−m . and p ≠ 1: 1 w= s(p − 1) p 1 1 + + 1−p 2−n 2−m 1 p−1 1−p ln C1 . Solution for n ≠ 2. and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES = wp . C12−m y. Suppose w(x.3. p−1 1◦ . a(2 − n)2 b(2 − m)2 e−λz y 2−m + . 2−n 2−m where the function w(r) is determined by the ordinary differential equation wrr + A w = swp . r2 = 4 y 2−m e−λz x2−n + + . n−2 ρ2 = z + = wp . m ≠ 2. 2 2 a(2 − n) b(2 − m) cλ2 2 2 + − 1. p−1 p−1 1◦ .1. ρ2 ). 3. There are “two-dimensional” solutions of the following forms: w(x. r r A= 4◦ . y. y. w(x.

2 a(2 − n) bβ 2 e−βy e−λz + . w(x. ρ1 = y + β ξ2 = + ∂ ∂w beγy ∂y ∂y + ∂ ∂w ceλz ∂z ∂z ρ2 = z + = wp . 2◦ . y. 2−n 4◦ . y. ∂w ∂ aeβx ∂x ∂x This is a special case of equation 6. Then the function w1 = C1 w x + 1−p 1−p 1−p ln C1 . λ 4. y. z) = V (x. z) = V (x. z). β ≠ 0.1. r r A= n . η). β γ λ where C1 is an arbitrary constant. z) = x p−1 F (ρ1 . y + ln C1 . z) = W (y. Functional separable solution for β ≠ 0. γ ≠ 0. w(x. ρ1 = y − x. 1◦ . ζ 2 = aβ 2 cλ2 β βx F (ρ1 . w(x. r2 = 4 e−βx e−γy e−λz + + . z) = exp p−1 γ w(x. r r 4◦ . η 2 = bγ 2 cλ2 e−βx e−λz + . ¢ 2−n ln x.3. η). ζ). 3◦ . y. ζ 2 = 2 a(2 − n) cλ2 n−2 2−n ln x. w(x. ρ2 ). y. e−βy x2−n + . There are “two-dimensional” solutions of the following forms: e−βx e−γy + . z). z) = W (y. and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). Solution for p ≠ 1. z) is a solution of the equation in question. y. z) = U (ξ. y. aβ 2 bγ 2 cλ2 where the function w(r) is determined by the ordinary differential equation wrr − 1 w = swp . γ ≠ 0.6. Suppose w(x. There are “two-dimensional” solutions of the following forms: w(x.4 with f (w) = sw p . ζ). y. y. λ © 2004 by Chapman & Hall/CRC . and λ ≠ 0: p w= b(1 − p)2 1 p−1 e−βx e−γy e−λz + + aβ 2 bγ 2 cλ2 1 1−p . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 407 where the function w(r) is determined by the ordinary differential equation wrr + A w = swp . w(x. is also a solution of the equation. w(x. aβ 2 bγ 2 e−γy e−λz + .1. z + ln C1 . ξ2 = ρ2 = z − β x. η 2 = bβ 2 cλ2 e−λz x2−n + . ρ2 ). z) = U (ξ.

. −xa1/2 sin β + y cos β. Equations of the Form ∂ f (w) ∂w + ∂ g(w) ∂w + ∂ g(w) ∂w = 0 ∂x ∂x ∂y ∂y ∂z ∂z 1. y. 2◦ . C5 and β are arbitrary constants. z) is a solution of this equation.408 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES 6. y. z). 5◦ . b √ c w(x. Solution: 2 2 w = v(r) − 4bC2 x2 − 4abC1 y 2 . . . . are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). ∂2w ∂x2 +a ∂2w ∂y 2 + ∂ ∂z (bw + c) ∂w ∂z = 0. C1 C2 z + C5 ) + £ £ £ 2 c(1 − C1 ) . z) = (C1 x + C2 )z + C3 x + C4 y + C5 − w(x. one can make the equation homogeneous. ξ = z + bC1 x2 + C2 y. y. z) = C1 |z|1/2 exp aC2 x sin(C2 y + C3 ) − . y. 2 w = u(ξ) − 4bC1 x2 . 1◦ . . . © 2004 by Chapman & Hall/CRC ζ = z + b(B1 x2 + B2 y 2 + B3 xy + B4 x + B5 y). where C1 and C2 are arbitrary constants and the function u(ξ) is determined by the ﬁrst-order ordinary differential equation 2 2 (bu + c + aC2 )uξ + 2bC1 u = 8bC1 ξ + C3 . C1 . With appropriate translations in both variables. z) = C1 |z|1/2 sin( aC2 x + C3 ) exp(C2 y) − . z) = A C 1 x + C2 y + C 3 z + C 4 − 2 2 C1 + aC2 c − . C5 are arbitrary constants (the ﬁrst solution is of the traveling-wave type). Solution: −2 4 1 12 bC1 (C1 x + C2 ) . Then the functions −2 w1 = C1 w( C2 x + C3 . b √ c w(x. . z) = (C1 x + C2 )z + C3 (ax2 − y 2 ) − c w(x. −2 4 1 12 bC1 (C1 x + C2 ) . where C1 and C2 are arbitrary constants and the function v(r) is determined by the ﬁrst-order ordinary differential equation 2 2 (bv + c)vr + 2b(aC2 + C1 )v = 8b(a2 C1 + C2 )r + C3 . With appropriate translations in both variables.1. one can make the equation homogeneous. b 3◦ . r = z + bC1 x2 + bC2 y 2 . Solutions: w(x. 2 bC1 w2 = w(x cos β + ya−1/2 sin β. y. . . which means it is integrable by quadrature.2. where A. z) = |z|1/2 C1 (ax2 − y 2 ) + C2 x + C3 + C4 ) − . Solution (generalizes the solution of Items 3◦ and 4◦ ): w = U (ζ) + A1 x2 + A2 y 2 + A3 xy + A4 x + A5 y. C2 y + C4 . 2 b bC3 w(x. Suppose w(x. y. 4◦ . where C1 . y.

3 with n = 1. and B5 are arbitrary constants. r). A3 = −4bB3 (B1 + aB2 ). B4 . where the functions f and g are determined by the system of differential equations ∂2f ∂2f + = 0. y). 8◦ . ∂x2 ∂η 2 9◦ . and h = h(x. For solutions of these linear equations. y)z + h(x. With appropriate translations in both variables. y. see Tikhonov and Samarskii (1990) and Polyanin (2002).1. 2 2 A2 = −b(B3 + 4aB2 ). which means it is integrable by quadrature. ζ = (ax2 + y 2 )z −2 w(x. and the function U (ζ) is determined by the ﬁrst-order ordinary differential equation 2 2 (bU + c + ab2 B5 + b2 B4 )Uζ + 2b(aB2 + B1 )U + 2(aA2 + A1 )ζ = C1 . η). (2) ∂x2 ∂η 2 Equation (1) is the Laplace equation. b < 0. “two-dimensional” solution. “Two-dimensional” solution (generalizes the last three solutions of Item 2 ◦ ): c w(x.1. z) = H(ζ). g = g(x. 6◦ . Here. “Two-dimensional” generalized separable solution linear in z (generalizes the second and third solutions of Item 2◦ ): w = f (x. “two-dimensional” solution. y). y)z 2 + g(x. y). B3 . η) − . hxx + ahyy + bg 2 + 2bf h + 2cf = 0. the subscripts denote the corresponding partial derivatives. z) = |z|1/2U (x. z) = F (z. “one-dimensional” solution. where the functions f = f (x. gxx + agyy + 6bf g = 0. η)z + g(x. “Two-dimensional” generalized separable solution quadratic in z: w = f (x. one can make the equation homogeneous. y) are determined by the system of differential equations fxx + afyy + 6bf 2 = 0. B2 . η) − . (2) represents a Helmholtz equation. In the special case a = 1.2. η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. z) = x2λ G(ξ. η = a−1/2 y. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 409 where B1 . Remark. There are solutions of the following forms: r = ax2 + y 2 y z c w(x. The substitution u = w + (c/b) leads to a special case of equation 6. where λ is an arbitrary constant. A5 = −2b(B3B4 + 2aB2 B5 ). ξ = . η = λ+1 b x x w(x.6. and c > 0. η = a−1/2 y. η). 7◦ . y. b where the function U = U (x. Given f = f (x. © 2004 by Chapman & Hall/CRC . y. (1) ∂x2 ∂η 2 ∂2g ∂2g + = −bf 2 . A4 = −2b(2B1B4 + aB3 B5 ). 1989). the equation in question describes transonic ﬂows of ideal polytropic gases (Pokhozhaev. y. 10◦ +. the coefﬁcients An are expressed in terms of Bn by 2 2 A1 = −b(4B1 + aB3 ).

z) = z n+1 [C1 (ax2 − y 2 ) + C2 x + C3 y + C4 ]. Then the functions w1 = w( C1 x + C2 . and k3 are arbitrary constants. = 0. . −xa1/2 sin β + y cos β. y. and D are arbitrary constants. z) = A k 1 x + k2 y + k 3 z + B − 2 2 2 k 1 + b1 k 2 + b2 k 3 . B1 . C. . . 1 √ w(x.2. 2◦ . w(x. . C4 are arbitrary constants. y. .3 with f (w) = 1. z) is a solution of this equation. y. where C1 . C1 z + C4 ). . For other solutions. . . 5◦ . 2 where A1 . Suppose w(x. Solutions: w(x. Then the functions ∂x2 ∂y 2 wn n −2 w1 = C1 w( C2 x + C3 . 2 + a k2 a1 k2 2 3 ¤ ¤ ¤ where A. C1 y + C3 . z) = (A1 x + B1 )y + (A2 x + B2 )z − 2 1 12 (a1 A1 2 2 1 1 + a2 A2 )x4 − 3 (a1 A1 B1 + a2 A2 B2 )x3 − 2 (a1 B1 + a2 B2 )x2 + Cx + D. y. . A2 . C4 are arbitrary constants. z) = C1 z n+1 exp( aC2 x) cos(C2 y + C3 ). B. g(w) = a 1 w + b1 . There is a generalized separable solution of the form w(x. and h(w) = a2 w + b2 . 4◦ . . where C1 . Solution linear in y and z: w(x.410 2. 1 1 1 where the function U = U (x. Traveling-wave solution: w(x. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily). y. z) is a solution of this equation. z) = f (x)y 2 + g(x)yz + h(x)z 2 + ϕ(x)y + ψ(x)z + χ(x). where C1 . 2◦ . are also solutions of the equation (the plus or minus signs are chosen arbitrarily). see Tikhonov and Samarskii (1990) and Polyanin (2002). y. y. z). 3. Suppose w(x.3. k2 . η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. . ∂z ∂z 1◦ . 3◦ . z) = z n+1 U (x. B2 . ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES ∂ ∂w ∂ ∂w ∂2w + (a1 w + b1 ) + (a2 w + b2 ) = 0. © 2004 by Chapman & Hall/CRC ¤ ¤ ¤ η = a−1/2 y. ∂x2 ∂η 2 For this linear equation. C1 C2 z + C5 ). y. C5 and β are arbitrary constants. “Two-dimensional” solution (generalizes the solutions of Item 2 ◦ ): w(x. z) = z n+1 [C1 ln(ax2 + y 2 ) + C2 ]. C2 y + C4 . ∂2w +a ∂2w +b ∂ ∂w w2 = w(x cos β + ya−1/2 sin β. 3◦ . 2 ∂x ∂y ∂y ∂z ∂z 1◦ . y. η). . . see equation 6. k1 .

w(x. C4 and λ are arbitrary constants. z) = |z|2/n U (t1 . y. 7◦ . and λ are arbitrary constants. this equation is linear. ρ2 = y exp − . ξ = . y. η). ∂x2 ∂η 2 n2 For n = −1 and n = −2. z) is a solution of this equation. C1 C2 z + C5 ). t2 = y + k2 ln |z|. z) = F (z. y. 1◦ . z) = W (ζ). 2◦ . w(x. y cos β + z a/b sin β. y)θ(z). © 2004 by Chapman & Hall/CRC 1 n+1 . z) = x C1 ln by 2 + az 2 + C2 n+1 . ζ = y/x. w(x.1: ∂ 2 u ∂ 2 u 2b(n + 2) n+1 + + u = 0. t1 = x + k1 ln |z|. ¨ ¨ ¨ w2 = w x. ¦ §¥ Reference: N. where θ = θ(z) is determined by the autonomous ordinary differential equation (θn θz )z = Cθ. √ √ w(x. k2 . ρ1 = x exp − . y. C1 C2 y + C4 . y.1.2. y. . where C1 . z) = x C1 exp λ b y sin λ a z + C2 + C3 where C1 . 5◦ .1. C5 and β are arbitrary constants. y. Suppose w(x. w(x. y. x x w(x. 4. where n should be set equal to zero and then k should be renamed n. η) is determined by a differential equation of the form 5. z) = S(r)z 2/n .6. z) = x C1 (by 2 − ax2 ) + C2 x + C3 y + C4 1 1 n+1 . . ζ = (ax2 + y 2 )z −2 . . ζ). are also solutions of the equation (the plus or minus signs are chosen arbitrarily). ∂ ∂w ∂2w +a wn 2 ∂x ∂y ∂y +b ∂ ∂w wn ∂z ∂z = 0. . .5. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING POWER-LAW NONLINEARITIES 411 4◦ . The solutions of Items 2◦ and 3◦ are special cases of a multiplicative separable solution w = u(x. η = a−1/2 y. . ρ2 ). Remark. see equation 6. Ibragimov (1994). 6◦ . r = ax2 + y 2 . . where the function u = u(x. z y w(x. For other solutions. . “Two-dimensional” solution: w(x. There are “two-dimensional” solutions of the following forms: r = ax2 + y 2 . Then the functions n n −2 w1 = C1 w( C2 x + C3 . η = nλ+1 . y. r). z) = |x|−2/n H(z. z) = x2λ G(ξ. t2 ). nz nz 2z V (ρ1 . y. z) = u(x. η)z 2/n . Degenerate solutions: w(x.1. . y. −y b/a sin β + z cos β . y. w(x. There are solutions of the following forms: w(x. z) = exp − n+1 n+1 n+1 where k1 .

β2 . n2 √ For √ = −2. y. © 2004 by Chapman & Hall/CRC . Suppose w(x. w(x. ξ = b y. where C1 . z) is a solution of the equation in question. see equation 6. C1 C2 z + C5 ). β1 . z) = x2λ G(ξ. y.3. ξ = nλ+1 . ξ= √ b y. η= √ a z. see Tikhonov and Samarskii (1990) and Polyanin (2002). x w(x. where ϕ = ϕ(x) is determined by the autonomous ordinary differential equation ϕzz = Cϕn+1 . ζ). η) 1 n+1 .412 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES 3◦ .1. z) is determined by the differential equation a ∂Θ ∂ Θn ∂y ∂y +b ∂ ∂Θ Θn ∂z ∂z + 2(n + 2) Θ = 0. ρ2 ). z) = |x|−2/n U (z1 .2. w(x. ρ2 = zenx “two-dimensional” solution. 6◦ . 5. to the Laplace equation. and h(w) = bw k . z1 = y + k1 ln |x|. y. z) = x−2/n Θ(y. θ = (by + az )x where k1 . y. C2 . y. and β3 are arbitrary constants. 2 2 −2 “two-dimensional” solution. η= z xnλ+1 “two-dimensional” solution. “Two-dimensional” solution: w(x. 5◦ . . z) = F (x. “one-dimensional” solution. where the function Θ = Θ(y. Ibragimov (1994). g(w) = aw n . y. §© Reference: N. 4◦ . “Two-dimensional” solution (generalizes the solutions of Item 2 ◦ ): w(x. z) = z 2/n H(x. and λ are arbitrary constants. . This is a special case of equation 6. 2◦ . ζ = z/y w(x. C1 C2 y + C4 . with the transformation u = 1/Θ. k2 . . y. Traveling-wave solution in implicit form: 2 β1 w + 2 2 bβ3 k+1 aβ2 n+1 w + w = C1 (β1 x + β2 y + β3 z) + C2 . ∂ξ 2 ∂η 2 For solutions of this linear equation. Then the functions n k −2 w1 = C1 w( C2 x + C3 . n+1 k+1 where C1 . n η = a z. where the function U = U (ξ. There are solutions of the following forms: r = by 2 + az 2 y w(x. Remark. w(x. 1◦ . η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0.5 with k = n. y. η). z) = e−2x V (ρ1 . ρ1 = yenx . ∂2w ∂x2 +a ∂ ∂y wn ∂w ∂y +b ∂ ∂z wk ∂w ∂z = 0. are also solutions of the equation (the plus or minus signs are chosen arbitrarily).3 with f (w) = 1. z). For other solutions. z2 = z + k2 ln |x| “two-dimensional” solution. . C5 are arbitrary constants.2. The solutions of Items 2◦ and 3◦ are special cases of a multiplicative separable solution w = ϕ(x)u(y. z). y. z2 ). z) = W (θ). the equation obtained can be reduced. “two-dimensional” solution. z) = (C1 x + C2 ) U (ξ. r).

Ibragimov (1994). Zaitsev and A.3 with f (w) = seλw . 1 2 which can be reduced to a linear equation.3. y. Suppose w(x.4. “two-dimensional” solution. z) = (y/x) § U (θ). z2 = a2 y + b2 z + c2 x. Equations with Three Space Variables Involving Exponential Nonlinearities 6. y. “two-dimensional” solution. There is a “two-dimensional” solution of the form (generalize the solutions of Items 3 ◦ and 4◦ ): w(x.1. z2 ).3 with f (w) = a1 wn1 .4. a1 w n1 w n2 w n3 ∂ ∂w + a2 ∂ ∂w + a3 ∂ ∂w 6. C12−k y + © 2004 by Chapman & Hall/CRC 2 2 2 . 6. y. z) = x2λ F (ξ. y. ζ = y −k/n z w(x.2. Zaitsev (2002). A. “Two-dimensional” solution (s1 and s2 are arbitrary constants): w(x. 2 2 ψ(v) = s1 + bs2 v k . H(z1 . 6◦ . C12−m y. z) = u(x. x). 1◦ . g(w) = a2 wn2 .2.8: a ∂v ∂ vn ∂y ∂y + ∂ ∂v ψ(v) ∂η ∂η = 0. ∂y ∂y ∂y ∂y ∂z ∂z This is a special case of equation 6. ξ = c1 y + c2 z. η) is determined by a differential equation of the form 5. z) = U (z1 . ∂ ∂u ∂2u + ϕ(u) = 0. F. Polyanin (2001).3. 4◦ . “Two-dimensional” solution (c1 and c2 are arbitrary constants): w(x. w1 = w C12−n x. ξ = nλ+1 . F. η = s1 x + s2 z. θ=x k/n−1 −k/n where λ is an arbitrary constant. z) = e −2x “two-dimensional” solution. where the function u = u(x. ∂ axn ∂w + ∂ by m ∂w + ∂ cz k ∂w = eλw . There are exact solutions of the following forms: z y w(x. y. 2/n z1 = ye . z1 = a1 y + b1 z + c1 x.2. z) = v(y. y.6.4. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 413 3◦ . z2 ). η). Equations of the Form ∂ f (x) ∂w + ∂ g(y) ∂w + ∂ h(z) ∂w = aeλw ∂x ∂x ∂y ∂y ∂z ∂z ∂x ∂x ∂y ∂y ∂z ∂z 1. z) = y 2/n G(ζ.1. is also a solution of the equation. References: N. and h(w) = a3 wn3 .2. λ where C1 is an arbitrary constant. η). η = kλ+1 x x w(x. where the function v = v(y.4. ξ) is determined by a differential equation of the form 5. = 0. 2 ∂x ∂ξ ∂ξ which can be reduced to a linear equation. D. z) is a solution of the equation in question. Then the function 2 ln C1 . 5◦ . ξ). y. D. nx z2 = ze y z kx w(x. y. “one-dimensional” solution. V.8: ϕ(u) = ac2 un + bc2 uk . This is a special case of equation 6. Polyanin and V.

wrr + wr = seσw . Solution for n ≠ 2. ζ 2 = a(2 − n)2 cλ2 n−2 2−n n−2 ln x. w(x. z). w(x. z) = F (ρ1 . y. η). z) = V (x. 2 2 a(2 − n) b(2 − m) c(2 − k)2 where the function w(r) is determined by the ordinary differential equation 2 2 2 B B= + + − 1. and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). 2−n 2−m 2−k 3◦ . y. η 2 = b(2 − m)2 cλ2 e−λz x2−n + . 1◦ . z) = W (y. wrr + wr = seλw . There are “two-dimensional” solutions of the following forms: w(x. r 2−n 2−m 2−k 4◦ . 2 2 axn ∂w + ∂ by m ∂w + ∂ ceλz ∂w = eσw . ρ2 ) + σ λ ξ2 = © 2004 by Chapman & Hall/CRC . η). ρ1 = yx 2−m . y. a(2 − n)2 b(2 − m)2 e−λz y 2−m + . r 2−n 2−m 4◦ . A=1− 1 1 − .3. y 2−m e−λz x2−n + + . w1 = w C12−n x. Functional separable solution for n ≠ 2. A =1− 1 1 1 − − . and k ≠ 2: w=− x2−n y 2−m z 2−k sλ 1 + + ln λ A a(2 − n)2 b(2 − m)2 c(2 − k)2 r2 = 4 . z) = U (ξ. z) = W (y. y. m ≠ 2. Suppose w(x. λ σ where C1 is an arbitrary constant. y 2−m z 2−k x2−n + + . y. y. Solution for n ≠ 2. a(2 − n)2 b(2 − m)2 z 2−k y 2−m + . y. w(x. z) = F (ρ1 . is also a solution of the equation. a(2 − n)2 b(2 − m)2 cλ2 where the function w(r) is determined by the ordinary differential equation 2 2 B B= + − 1. ρ1 = yx 2−m . m ≠ 2. Functional separable solution for n ≠ 2. 2−n 2−m 3◦ . C12−m y. ρ2 ) + λ ξ2 = 2. y. y 2−m x2−n + . ρ2 = z + ln x. and λ ≠ 0: w=− sσ x2−n y 2−m e−λz 1 ln + + 2 2 σ A a(2 − n) b(2 − m) cλ2 r2 = 4 . w(x. z) = V (x.5 with f (w) = seσw . η 2 = b(2 − m)2 c(2 − k)2 z 2−k x2−n + . ∂ ∂x ∂x ∂y ∂y ∂z ∂z This is a special case of equation 6.414 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES 2◦ . ρ2 = zx 2−k . z − 2◦ . y 2−m x2−n + . y. ζ). z) is a solution of the equation in question. ζ 2 = a(2 − n)2 c(2 − k)2 n−2 n−2 n−2 ln x. ζ). and k ≠ 2 (generalizes the solution of Item 2 ◦ ): w = w(r). z) = U (ξ. m ≠ 2. There are “two-dimensional” solutions of the following forms: w(x. w(x. w(x. Then the function 2 2 ln C1 + ln C1 . z).1. m ≠ 2.

1. = eσw . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 415 1◦ . 3◦ . ρ2 = z + 2−n ln x. ζ).2. 3◦ . ζ 2 = 2 a(2 − n) cλ2 2−n n−2 ln x. β ≠ 0. e−βy x2−n + . η). β ≠ 0.3.4 with f (w) = se . w(x. y − 2 2 2 2 ln C1 . is also a solution of the equation. is also a solution of the equation. w(x. γ ≠ 0. 2 a(2 − n) bβ 2 cλ2 where the function w(r) is determined by the ordinary differential equation wrr + A w = seσw . and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). 1◦ .6 with f (w) = se . ρ1 = y + ln x. There are “two-dimensional” solutions of the following forms: w(x. 2◦ . r r A= n . γ ≠ 0. z) is a solution of the equation in question. z) = W (y. z) = F (ρ1 . r2 = 4 e−βx e−γy e−λz + + .3. r2 = 4 e−βy e−λz x2−n + + . y. z) = V (x. aβ 2 bγ 2 cλ2 © 2004 by Chapman & Hall/CRC ∂w ∂ ∂w ∂ ∂w ∂ axn + beβy + ceλz ∂x ∂x ∂y ∂y ∂z ∂z σw This is a special case of equation 6. Functional separable solution for n ≠ 2. y. z − ln C1 + ln C1 . ∂w ∂ ∂w ∂ ∂w ∂ aeβx + beγy + ceλz ∂x ∂x ∂y ∂y ∂z ∂z σw This is a special case of equation 6. ρ2 ) + σ β ξ2 = 4. Then the function w1 = w x − 2 2 2 2 ln C1 . 2−n 4◦ . and λ ≠ 0: w=− sσ(2 − n) x2−n e−βy e−λz 1 ln + + σ 1−n a(2 − n)2 bβ 2 cλ2 . 3. y. = eσw . y. Solution for n ≠ 2. 2 a(2 − n) bβ 2 e−βy e−λz + . η 2 = bβ 2 cλ2 e−λz x2−n + . y. Suppose w(x. Solution for β ≠ 0. β γ λ σ where C1 is an arbitrary constant.6. and λ ≠ 0: w=− e−βx e−γy e−λz 1 ln sσ + + σ aβ 2 bγ 2 cλ2 . y − ln C1 . z) is a solution of the equation in question.1. w(x. y. λ . Suppose w(x. 2◦ . β λ σ where C1 is an arbitrary constant. Functional separable solution for β ≠ 0. z) = U (ξ. z − ln C1 + ln C1 . and λ ≠ 0 (generalizes the solution of Item 2 ◦ ): w = w(r). z). Then the function w1 = w C12−n x.

y. w(x. ρ1 = y − x. √ w(x. y. Solutions: w(x. z) = F (ρ1 . z) = ln b where C1 . C2 z + C5 ) + ln 2 C1 . y. −xa1/2 sin β + y cos β. . ∂z ∂z 1◦ . z) = C1 exp( a C2 x) sin(C2 y + C3 ) + ln(C4 z + C5 ). w(x. z) = W (y. y. Equations of the Form a1 ∂ eλ1 w ∂w + a2 ∂ eλ2 w ∂w ∂x ∂x ∂y ∂y 1.2.2. . see Tikhonov and Samarskii (1990) and Polyanin (2002). z) = U (ξ. b cosh2 (C1 x + C2 y + C3 ) 4aC3 − 2 ln (y + C1 )2 + a(x + C2 )2 + C3 + 2 ln |z + C4 |. z). λ 6. y. © 2004 by Chapman & Hall/CRC . . ρ2 ) + x. z) = U (x. y. ∂2w ∂x2 +a ∂2w ∂y 2 +b ∂ ew ∂w + a3 ∂ eλ2 w ∂w ∂y ∂y = beβw = 0. Suppose w(x. . y. η = a−1/2 y. y. w(x. 2 C2 w2 = w(x cos β + ya−1/2 sin β. z) is a solution of this equation. “Two-dimensional” solution (generalizes the ﬁrst three solutions of Item 2 ◦ ): w(x. z) = C1 exp(C2 y) sin( a C2 x + C3 ) + ln(C4 z + C5 ). ∂x2 ∂η 2 For this linear equation. 2◦ . η). η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. . ξ2 = ρ2 = z − β x. . z). ζ). Then the functions w1 = w(C1 x + C3 . √ w(x. C7 are arbitrary constants. y. z) = C1 (ax2 − y 2 ) + C2 xy + C3 x + C4 y + C5 + ln(C6 z + C7 ).416 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES where the function w(r) is determined by the ordinary differential equation wrr − 1 w = seσw . z) = ln 2 2 (C1 + aC2 )(z + C4 )2 . σ γ w(x. . η 2 = bγ 2 cλ2 e−βx e−λz + . z) = V (x. w(x. r r 4◦ . . There are “two-dimensional” solutions of the following forms: e−βx e−γy + . where C1 and C2 are arbitrary constants and the function U (x. 3◦ . C5 and β are arbitrary constants. η) + ln(C1 z + C2 ). aβ 2 bγ 2 e−γy e−λz + . y. where C1 . y. C1 y + C4 . ζ 2 = aβ 2 cλ2 β β w(x. are also solutions of the equation (the plus or minus signs in w1 are chosen arbitrarily).

. z) = V (ρ1 . where the function V = V (x. y. y. y. k1 . ζ1 = x + k1 ln |z|. z) = V (x. χ = (ax2 + y 2 )z −2 where k. y. w(x. λ 1 w(x. −y b/a sin β + z cos β .3 with f (w) = 1. “one-dimensional”. η = a−1/2 y.2. . Then the functions w1 = w(C1 x + C3 . . “Two-dimensional” solution: w(x. . ξ2 ) − 2k ln |x|.1: ∂2V ∂2V + = −2beV . ∂ ∂w ∂ ∂w +a eλw +b eλw = 0. There are solutions of the following forms: w(x. z) = F (r. z) = C1 x + C2 + © 2004 by Chapman & Hall/CRC 2 1 C1 ln 2 . 2. ξ2 = z|x| k−1 k−1 w(x. C2 y + C4 . z) = C1 x + C2 + b C3 y + ln cos a C3 λz + C4 . g(w) = a. y. + C6 . y. η2 ) + 2k ln |z|. “two-dimensional”. . C2 z + C5 ) + w2 = w x. y. Suppose w(x. λ C2 ∂2w a/b sin β. y. y cos β + z 2◦ . λ 1 w(x. r = ax2 + y 2 ξ1 = yx . y. η2 = y|z| “two-dimensional”. y. y. z) = C1 x + C2 + ln C3 ln(by 2 + az 2 ) + C4 . . k−1 −1 “two-dimensional”. z) = ln λ cos2 (aC1 x + C5 ) √ √ 2 −bC1 z 2 + C2 exp b C3 y cos a C3 z + C4 1 w(x. z) = G(ξ1 . η) + 2 ln |z + C|. . C5 and β are arbitrary constants. Solutions: 1 ln(C3 y + C4 z + C5 ). y. EQUATIONS WITH THREE SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 417 4◦ .6. w(x. z) = ln λ cos2 (bC1 x + C5 ) √ √ 2 2 −aC1 y + C2 exp b C3 y cos a C3 z + C4 1 w(x. y. 6◦ . and k2 are arbitrary constants. z) = W (χ). z) = C1 x + C2 + ln C3 exp b C4 y cos a C4 z + C5 λ √ √ 2 −aC1 y 2 + C2 exp b C3 y cos a C3 z + C4 1 w(x. z). η) is determined by a solvable differential equation of the form 5. ζ2 ) + 2 ln |z|. and h(w) = be w .3. z) = U (ζ1 . y. ρ2 ) + 2z. z) = ln λ sinh2 (bC1 x + C5 ) w(x. y. z) = H(η1 . w(x. z) = C1 x + C2 + ln C3 (by 2 − az 2 ) + C4 yz + C5 . λ √ √ 1 w(x. where C1 . y. ∂x2 ∂η 2 5◦ . see equation 6. w(x. . For other exact solutions. . z) is a solution of this equation.2.2. ρ2 = yez “two-dimensional”. λ √ √ 1 w(x. ∂x2 ∂y ∂y ∂z ∂z 1◦ . η1 = x|z| . z) = ln λ sinh2 (aC1 x + C5 ) √ √ 2 −bC1 z 2 + C2 exp b C3 y cos a C3 z + C4 1 w(x.1. y. ρ1 = xez . ζ2 = y + k2 ln |z| “two-dimensional”.

2. “Two-dimensional” solution (generalizes the ﬁrst ﬁve solutions of Item 2 ◦ ): z y 1 w(x. − ln 2 λ 2C1 ∆V − kλ = 0. and h(w) = beλw . y. ζ = x2 6◦ . z2 ) − λ 2 w(x. y. 1 λ 2 1 − ln − kλ cos2 (C1 x + C2 ) if kλ < 0. 3◦ . y. For other exact solutions. r = by 2 + az 2 “two-dimensional” solution. w(x. z) = W (ζ). a © 2004 by Chapman & Hall/CRC . η). y. w(x. y. y.3 with f (w) = 1. η) is a solution of the Poisson equation ∂2 ∂2 + . λ a b where C1 and C2 are arbitrary constants and the function U = U (ξ. λ y ξ= √ . ρ2 ) − x. τ ) + ln |y|. z1 = y|x|k−1 . b where the function f = f (x) is determined by the autonomous ordinary differential equation (k is an arbitrary constant) fxx + keλf = 0. . τ = . ρ2 = zex λ by 2 + az 2 “one-dimensional” solution. y. w(x. 2k ln |x|. η) − ln |x|. see Tikhonov and Samarskii (1990) and Polyanin (2002). z) = ln λ cosh2 (bC1 x + C5 ) . y. ρ1 = yex . z) = F (x. “Two-dimensional” solution: w(x. ∂ξ 2 ∂η 2 For solutions of this linear equation. z) = U (ξ. . C6 are arbitrary constants. 1 2 λ 2 2C1 1 kλ cosh2 (C1 x + C2 ) if kλ > 0. λ y w(x. z2 = z|x|k−1 “two-dimensional” solution. 2 λ 2C1 f (x) = − 1 ln − kλ sinh2 (C x + C ) if kλ < 0. η) is determined by the Laplace equation ∂ 2 U ∂ 2U + = 0. where C1 . . η). (2) 2 ∂ξ ∂η 2 For solutions of this linear equation.3. η = √ . y.418 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES √ √ 2 aC1 y 2 + C2 exp b C3 y cos a C3 z + C4 1 w(x. There are solutions of the following forms: z 2 “two-dimensional” solution. y. z) = H(z1 . η = z + k2 ln |x| “two-dimensional” solution. z) = G(x. w(x. r). z) = V (ρ1 . . 1 ln V (ξ. z) = C1 x + C2 + ln U (ξ. . g(w) = ae λw . ∆= 5◦ . ξ = y + k1 ln |x|. z) = ln λ cosh2 (aC1 x + C5 ) √ √ 2 bC1 z 2 + C2 exp b C3 y cos a C3 z + C4 1 w(x. ξ = √ . 4◦ . (1) and the function V = V (ξ. z) = f (x) + z η= √ . The general solution of equation (1) is expressed as − 1 ln − 1 kλ(x + C )2 if kλ < 0. see Tikhonov and Samarskii (1990) and Polyanin (2002). λ 2 “two-dimensional” solution. see equation 6.

y. 6◦ . “two-dimensional”. which can be reduced to a linear equation.6. k2 . where the function u = u(x. .4. y. where C1 . y. and h(w) = beλw . Traveling-wave solution in implicit form: 2 2 2 k1 w + ak2 ew + bk3 λ−1 eλw = C1 (k1 x + k2 y + k3 z) + C2 .2. There is a “two-dimensional” solution of the form (generalize the solutions of Items 3 ◦ and 4◦ ): w(x. y.4.8: a ∂v ∂ ev ∂y ∂y + ∂ ∂v ψ(v) ∂η ∂η = 0. . C1 C2 3 z + C5 ) − 2 ln |C2 |. 5◦ . ∂x2 ∂ξ ∂ξ which can be reduced to a linear equation. “two-dimensional”. . see equation 6. “Two-dimensional” solution (c1 and c2 are arbitrary constants): w(x. . z) = F (ξ1 . © 2004 by Chapman & Hall/CRC . Suppose w(x. z2 = a2 x + b2 y + c2 z.3. where C1 . . w(x. ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . C5 are arbitrary constants. k1 . .4. C1 C2 2 y + C4 . z) = G(x. z1 = a1 x + b1 y + c1 z.4. y. a1 ∂ eλ1 w ∂w + a2 ∂ eλ2 w ∂w + a3 ∂ eλ3 w ∂w = 0. Ibragimov (1994). 4◦ . ξ = c1 y + c2 z. η) is determined by a differential equation of the form 5. y. C2 . z) is a solution of this equation. 2 2 ψ(v) = bs2 eλv + s1 . “Two-dimensional” solution (s1 and s2 are arbitrary constants): w(x. For other exact solutions. ρ = |x| λ−1 ζ2 = ze |y| z −λ Reference: N. . η = s1 x + s2 z. η). y. C5 are arbitrary constants. w(x.2 with f (w) = 1. y. § ξ1 = y|x|k−1 . There are exact solutions of the following forms: w(x. z2 ). Suppose w(x. ζ2 ) − 2kx. w(x. ξ2 ) − 2k ln |x|. −λ kx ξ2 = z|x|kλ−1 kλx “two-dimensional”. C1 C2 z + C5 ) − 2 ln |C2 |. Then the functions 4. 7◦ . where C1 . λ λ λ w1 = w( C1 C2 1 x + C3 . λ w1 = w(C1 x + C3 . “one-dimensional”. η = |y| z ζ1 = ye . z) = H(ζ1 . z) = U (z1 . z) is a solution of this equation. 3◦ . is also a solution of the equation. g(w) = ae w . z) = V (ρ) + 2 ln |y/x|. z) = u(x. η) + 2 ln |y|. 1 2 where the function v = v(y. where k is an arbitrary constant.8: ∂ ∂u ∂2u + ϕ(u) = 0. C1 C2 y + C4 .2. z) = v(y. 2 ∂x ∂y ∂y ∂z ∂z 1◦ . ξ). 2◦ . ϕ(u) = ac2 eu + bc2 eλu . EQUATIONS WITH THREE SPACE VARIABLES INVOLVING EXPONENTIAL NONLINEARITIES 419 ∂ ∂w ∂ ∂w ∂2w +a ew +b eλw = 0. and k3 are arbitrary constants. Then the function 3. ξ) is determined by a differential equation of the form 5. y. .

5. Solution: 2 F (w) aA2 + bB 2 + cC 2 dw = C2 θ. 1. a1 ∂ ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . w(x. . r2 = © 2004 by Chapman & Hall/CRC . For other exact solutions. where C1 . and C3 are arbitrary constants. z) = V (ζ1 . w = w(θ). C1 3 z + C4 ) + 2 ln |C1 |. f1 (x). f2 (y). C2 . −1/2 C1 + 2◦ . r w = w(r). g(w) = a2 eλ2 w . β−λ β−λ β−λ w1 = w( C1 1 x + C2 . .3. B. and C2 are arbitrary constants. There is a “two-dimensional” solution of the form β−λ2 2 ln |x|. z) is a solution of this equation.420 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES 2◦ . F (w) = f (w) dw. C1 2 y + C3 . Heat and Mass Transfer Equations of the Form ∂ f (x) ∂w + ∂ f (y) ∂w + ∂ f (z) ∂w = g(w) ∂x 1 ∂x ∂y 2 ∂y ∂z 3 ∂z Equations of this type describe steady-state heat/mass transfer or combustion processes in inhomogeneous anisotropic media. z) = U (η1 . and f3 (z) are the principal thermal diffusivities (diffusion coefﬁcients) dependent on coordinates. η1 = y|x|k(λ2 −λ1 )−1 . and the function w(r) is determined by the ordinary differential equation 2 wrr + wr = f (w). . λ1 − λ 2 x ◦ 3 . C. C1 . a ∂2w ∂x2 ∂y 2 ∂z 2 ◦ 1 . and g = g(w) is the kinetic function. (x + C1 )2 (y + C2 )2 (z + C3 )2 + + .3. and h(w) = a 3 e λ3 w .3 with f (w) = a 1 eλ1 w . ξ = |x| λ1 −λ2 |y| λ1 −λ2 z. see equation 6. are also solutions of the equation (the plus or minus signs are chosen arbitrarily). y. Then the functions ! ! ! eλ1 w ∂w + a2 ∂ eλ2 w ∂w + a3 ∂ eλ3 w ∂w = beβw . . C4 are arbitrary constants. There are “two-dimensional” solutions of the following forms: w(x.2. The function w(θ) is deﬁned implicitly by ! θ = Ax + By + Cz. ζ2 = z exp k(λ3 − λ1 )x . y. η2 ) − 2k ln |x|. y. Here. 4◦ . Three-Dimensional Equations Involving Arbitrary Functions 6. z) = U (ξ) − w(x. β−λ3 6. There is an exact solution of the form λ2 −λ3 λ3 −λ1 y 2 ln . 2◦ . y. which deﬁnes the law of heat (substance) release or absorption. ξ = y|x| λ1 −β . w(x. where A. a b c where C1 . η2 = z|x|k(λ3 −λ1 )−1 . ζ1 = y exp k(λ2 − λ1 )x . η) + λ1 − β η = z|x| λ1 −β . y. ζ2 ) − 2kx. where k is an arbitrary constant.3. Suppose w(x.1. Traveling-wave solution: +b ∂2w +c ∂2w = f (w). z) = U (ξ.

ξ) is determined by the differential equation a ∂ 2 U ∂ 2 U B ∂U + + = f (U ). There are “two-dimensional” solutions of the following forms: w = V (y. Zhurov (1998). η). 4a c(2 − m)2 y 2−n x2 + . z = c z brings the original equation to the form ∆w = f (w). a ∂ ∂w ∂ ∂w + by n + cz m ∂x2 ∂y ∂y ∂z ∂z ◦ 1 .3. The transformation x = a x. The solution speciﬁed in Item 3◦ can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. r r ξ2 = 4 A= 3◦ . η) is determined by the equation 1+ 1 C2 ∂2U ∂2U ∂2U ∂U − 4ξ + 4C 2 (ξ 2 + η) 2 + 2(2C 2 − 1) = f (U ). a) (z. (2 − n)(2 − m) where the function w(r) is determined by the ordinary differential equation wrr + A w = f (w). “Two-dimensional” solution for n ≠ 2 and m ≠ 2: w = U (x. 2◦ . ζ 2 = 4 # §" Reference: A. ∂2w = f (w). ∂ζ 2 ∂ρ2 ρ ∂ρ A + B2 + C 2 √ √ √ ¯ ¯ ¯ 5◦ . Ax By Cz ζ= √ + √ +√ . 2. r2 = 4 y 2−n z 2−m x2 + + . η2 = 4 z 2−m x2 + . z 2−m y 2−n + . a c b ρ2 = Bx Ay √ −√ a b 2 2 2 + Cy Bz √ − √ c b + Az Cx √ −√ c a . ρ). ξ). ζ). Functional separable solution for n ≠ 2 and m ≠ 2: w = w(r). “Two-dimensional” solution: w = V (ζ. c) ←− (y. 4a b(2 − n)2 w = W (z. D. and the function U = U (ξ. b) 4◦ . a c ab where C is an arbitrary constant (C ≠ 0). ρ) is determined by the equation ∂ 2V 1 ∂V 1 ∂2V + + = 2 f (V ). η). 4a b(2 − n)2 c(2 − m)2 2(4 − n − m) . B. Polyanin and A. b(2 − n)2 c(2 − m)2 4 − nm . I. aC b η = (C 2 − 1) z2 x2 xy − 2C √ + C 2 . For n = m = 0. (2 − n)(2 − m) where the function U (x.6. © 2004 by Chapman & Hall/CRC . and C are arbitrary constants and the function V = V (ζ. where A. 2 ∂ξ ∂ξ∂η ∂η ∂η Remark. THREE-DIMENSIONAL EQUATIONS INVOLVING ARBITRARY FUNCTIONS 421 3◦ .3. “Two-dimensional” solution: w = U (ξ.1. ∂x2 ∂ξ 2 ξ ∂ξ B= 4◦ .1. x y ξ= √ +√ . see equation 6. y = b y.

2◦ . Functional separable solution for n ≠ 2. w = w(r).422 3. aβ 2 bγ 2 w = W (z. ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . ζ 2 = 4 % §$ Reference: A. ζ). ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . © 2004 by Chapman & Hall/CRC . D. η). w = V (y. Functional separable solution for n ≠ 2. ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES ∂w ∂ ∂w ∂ ∂w ∂ axn + by m + cz k = f (w). e−βx e−γy e−λz + + . ζ 2 = 4 % §$ Reference: A. a(2 − n)2 b(2 − m)2 cλ2 1 1 + 2−n 2−m where the function w(r) is determined by the ordinary differential equation wrr + A w = f (w). r r A=2 − 1. r2 = 4 y 2−m e−λz x2−n + + . bγ 2 cλ2 e−βx e−λz + . aβ 2 bγ 2 cλ2 where the function w(r) is determined by the ordinary differential equation 1 w = f (w). Functional separable solution for β ≠ 0. ∂x ∂x ∂y ∂y ∂z ∂z 1◦ . There are “two-dimensional” solutions of the following forms: w = U (x. ∂w ∂ ∂w ∂ ∂w ∂ axn + by m + ceλz = f (w). I. aβ 2 cλ2 e−βx e−γy + . r r A=2 − 1. and λ ≠ 0: 5. a(2 − n)2 b(2 − m)2 ∂ ∂w w = W (z. ζ). r2 = 4 y 2−m z 2−k x2−n + + . 2 b(2 − m) c(2 − k)2 z 2−k x2−n + . and k ≠ 2: w = w(r). γ ≠ 0. 2 a(2 − n) c(2 − k)2 y 2−m x2−n + . Zhurov (1998). ξ2 = 4 η2 = 4 z 2−k y 2−m + . Zhurov (1998). D. I. η). Polyanin and A. m ≠ 2. ξ). a(2 − n)2 b(2 − m)2 c(2 − k)2 1 1 1 + + 2−n 2−m 2−k where the function w(r) is determined by the ordinary differential equation wrr + A w = f (w). ξ2 = 4 η2 = 4 e−γy e−λz + . and λ ≠ 0: w = w(r). m ≠ 2. w = V (y. There are “two-dimensional” solutions of the following forms: wrr − w = U (x. r2 = 4 4. ξ). Polyanin and A. r r 2◦ . ∂ aeβx ∂w + ∂ beγy ∂w + ceλz = f (w).

3. Example 2.3. ∂ f1 (x) ∂w + ∂ f2 (y) ∂w + ∂ f3 (z) 6. + (a3 x + b3 y + c3 z + d3 ) ∂z ∂x2 ∂y 2 ∂z 2 This equation describes steady-state mass transfer with a volume chemical reaction in a threedimensional translational-shear ﬂuid ﬂow. bβ 2 cλ2 e−λz x2−n + . w = V (y. 8C 1 ln − β βA − 2 ln(r 2 + C). and λ ≠ 0: w = w(r). ξ2 = 4 η2 = 4 e−βy e−λz + . equation (1) has the one-parameter solution w(r) = where C is an arbitrary constant.3. ζ 2 = 4 6. ξ). = aw ln w + bw. (a1 x + b1 y + c1 z + d1 ) © 2004 by Chapman & Hall/CRC . Functional separable solution for n ≠ 2. equation (1) can be solved by quadrature to obtain & C1 + 2 where C1 and C2 are arbitrary constants. I. ∂x ∂x ∂y ∂y ∂z ∂z This is a special case of equation 6.3. THREE-DIMENSIONAL EQUATIONS INVOLVING ARBITRARY FUNCTIONS 423 2◦ . 7. η). w = W (z. Zhurov (1998). a(2 − n)2 bβ 2 cλ2 where the function w(r) is determined by the ordinary differential equation wrr + n 1 w = f (w). D. 2 b(2 − m) cλ2 e−λz x2−n + .6. ζ 2 = 4 ( §' Reference: A. There are “two-dimensional” solutions of the following forms: w = U (x. r2 = 4 e−βy e−λz x2−n + + . 1. β ≠ 0. ξ2 = 4 η2 = 4 e−λz y 2−m + . ξ).2. w = V (y. For n = 1 and f (w) = Aeβw . a(2 − n)2 cλ2 y 2−m x2−n + .6 with g1 (x) = b and g2 (y) = g3 (z) = 0. f (w) dw dw = C2 r. Heat and Mass Transfer Equations with Complicating Factors ∂w ∂w + (a2 x + b2 y + c2 z + d2 ) ∂x ∂y ∂2w ∂2w ∂2w ∂w = + + – f (w). a(2 − n)2 b(2 − m)2 ceλz ∂w = f (w). β 2◦ . There are “two-dimensional” solutions of the following forms: w = U (x. ζ). For n = 0 and any f = f (w). a(2 − n)2 bβ 2 ∂w w = W (z. ζ). a(2 − n)2 cλ2 e−βy x2−n + . 2−n r r −1/2 (1) Example 1. Polyanin and A. ∂ axn ∂w + ∂ beβy ∂w + ∂ ∂x ∂x ∂y ∂y ∂z ∂z ◦ 1 . η).

Then the functions . B. c3 − k and the constants A. some of the equation coefﬁcients must satisfy the condition a1 + b2 + c3 = 0. ζ = Ax + By + Cz. c1 A2 + c2 B 2 + c3 C 2 = C.424 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES Let k be a root of the cubic equation a1 − k b1 c1 a2 b2 − k c2 a3 b3 = 0. The ﬁrst three equations are ﬁrst solved for A. g(w). © 2004 by Chapman & Hall/CRC ) ) ) ∂ f (w) ∂w + ∂ ∂w + ∂ ∂w 1◦ . ∂y ∂w ∂ (a3 x + b3 y + c3 z + d3 ) + ∂z ∂z Solutions are sought in the form w = w(ζ). g(w) h(w) w1 = w( C1 x + C2 . . where the constants A. ∂ ∂x (a1 x + b1 y + c1 z + d1 ) ∂w ∂x + ∂ ∂y (a2 x + b2 y + c2 z + d2 ) ∂w = f (w). are also solutions of the equation (the plus or minus signs in front of C1 are chosen arbitrarily). Remark. C1 y + C3 . d1 A2 + d2 B 2 + d3 C 2 = D. The resulting expressions are then substituted into the last equation to evaluate D. y. where the function w(ζ) is determined by the ordinary differential equation (kζ + Ad1 + Bd2 + Cd3 )wζ = (A2 + B 2 + C 2 )wζζ − f (w). and C solve the degenerate system of linear algebraic equations (a1 − k)A + a2 B + a3 C = 0. The desired function w(ζ) is determined by the ordinary differential equation ζwζζ + (a1 A + b2 B + c3 C)wζ = f (w). Here. . b1 A2 + b2 B 2 + b3 C 2 = B. Solution: w = w(ζ). B. and C. z) is a solution of the equation in question. b1 A + (b2 − k)B + b3 C = 0. C. Suppose w(x. f (w). can be omitted. . ζ = Ax + By + Cz + D. = 0. C4 are arbitrary constants. . hence. B. ∂x ∂x ∂y ∂y ∂z ∂z This equation describes steady-state heat/mass transfer or combustion processes in inhomogeneous anisotropic media. c1 A + c2 B + (c3 − k)C = 0. C1 z + C4 ). In the case of an incompressible ﬂuid. where C1 . One of the equations follows from the other two and. and h(w) are the principal thermal diffusivities (diffusion coefﬁcients) dependent on the temperature w. 3. and D are determined by solving the algebraic system of equations a1 A2 + a2 B 2 + a3 C 2 = A. 2.

z2 ). 6◦ . 3◦ . see Tikhonov and Samarskii (1990) and Polyanin (2002). z = b z leads to the Laplace equation ∂2v ∂2v ∂2v + = 0. treating w as the independent variable. 7◦ . There are “two-dimensional” solutions of the following forms: w(x. ζ). (5) which can be reduced to a linear equation. © 2004 by Chapman & Hall/CRC . y. For C5 ≠ 0. Relations (1) and equation (2) can be used to obtain two other “one-dimensional” solutions by means of the following cyclic permutations of variables and determining functions: (x. Then. η). f ) (z. Then. g) 4 . θ= 2 2 [θ2 f (w)wθ ]θ + C1 [g(w)wθ ]θ + C2 [h(w)wθ ]θ = 0. see (3). h) ←− (y. k3 . Let g(w) = af (w). z1 = a1 x + a2 y + a3 z. z) = U (x. r = ax2 + y 2 . where the an and bn are arbitrary constants (the ﬁrst solution generalizes the one of Item 3 ◦ ). ξ = y/x. C2 . Let g(w) = af (w) and h(w) = bf (w). Relations (4) and equation (5) can be used to obtain two other “two-dimensional” solutions by means of the cyclic permutations of variables and determining functions. which can be reduced to a second-order linear equation. . the transformation √ √ v = f (w) dw. THREE-DIMENSIONAL EQUATIONS INVOLVING ARBITRARY FUNCTIONS 425 2◦ . y. one obtains a Riccati equation for θ = θ(w): 2 2 C5 θw = θ2 f (w) + C1 g(w) + C2 h(w). . y. y. . + ∂x2 ∂y2 ∂z 2 For solutions of this linear equation. Traveling-wave solution in implicit form: 2 2 2 k1 f (w) + k2 g(w) + k3 h(w) dw = C1 (k1 x + k2 y + k3 z) + C2 . where the function U = U (x. and the function w(θ) is determined by the ordinary differential equation w = w(θ).3. η = z/x. C4 are arbitrary constants. y = a y. k2 . z) = u(r.4.8: + = 0. Solution: which admits the ﬁrst integral 2 2 [θ2 f (w) + C1 g(w) + C2 h(w)]wθ = C5 . z2 = b1 x + b2 y + b3 z. 5◦ . where C1 . see Polyanin and Zaitsev (2003). w(x. there is a “two-dimensional” solution of the form w(x. k1 .4. “Two-dimensional” solution (a and b are arbitrary constants): ◦ (3) (4) w(x. ζ) is determined by a differential equation of the form 5. and λ are arbitrary constants.6. z) = V (z1 . C1 y + C 2 z + C 3 . z) = W (ξ. ∂U ∂ f (U ) ∂x ∂x ∂ ∂U ψ(U ) ∂ζ ∂ζ ζ = ay + bz. z). (1) x + C4 where C1 . (2) For exact solutions of this equation. ψ(U ) = a2 g(U ) + b2 h(U ). .

Solution: w = w(ζ). Other Equations 1. ϕ(w) = A2 f1 (w) + B 2 f2 (w) + C 2 f3 (w). c1 A + c2 B + (c3 − k)C = 0. where the function w(ζ) is determined by the ordinary differential equation [ϕ(w)wζ ]ζ = (kζ + Ad1 + Bd2 + Cd3 )wζ . see 6. Remark 2. where F (w) = exp f (w) dw . Functional separable solution for n ≠ 2. a(2 − n)2 b(2 − m)2 c(2 − k)2 ∂2w + by m ∂2w + cz k ∂2w © 2004 by Chapman & Hall/CRC .3. b1 A + (b2 − k)B + b3 C = 0. ∂x2 ∂y 2 ∂z 2 For solutions of this linear equation. In the case of an incompressible ﬂuid. with an additional term g(w) on the right-hand side. ∂x ∂y ∂z This equation describes steady-state anisotropic heat/mass transfer with a volume chemical reaction in a three-dimensional translational-shear ﬂuid ﬂow. also has a solution of the form (1). F (w) leads to the three-dimensional Laplace equation for U = U (x.426 4. see Tikhonov and Samarskii (1990) and Polyanin (2002). m ≠ 2. Remark. One of the equations follows from the other two and. ∂ ∂x f1 (w) ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES ∂w ∂x + ∂ ∂y f2 (w) ∂w ∂w ∂y + ∂ ∂z f3 (w) ∂w ∂z ∂w + (a3 x + b3 y + c3 z + d3 ) ∂w . some of the equation coefﬁcients must satisfy the condition a1 + b2 + c3 = 0. 2. axn = f (w). r2 = 4 y 2−m z 2−k x2−n + + . A more general equation. Let k be a root of the cubic equation = (a1 x + b1 y + c1 z + d1 ) a1 − k b1 c1 a2 b2 − k c2 a3 b3 = 0. z): ∂2U ∂2U ∂2U + + = 0. dw . B. ∂2w ∂x2 ∂y 2 The substitution + ∂2w + ∂2w ∂z 2 U= = f (w) ∂w ∂x 2 + ∂w ∂y 2 + ∂w ∂z 2 . where g is an arbitrary function.4. Remark 1. ∂z 2 1◦ . y. and k ≠ 2: ∂x2 ∂y 2 w = w(r). with an additional convective term. ζ = Ax + By + Cz.1. can be omitted. c3 − k + (a2 x + b2 y + c2 z + d2 ) and the constants A. For a more complicated equation of the form (v ⋅ ∇)w = ∆w − f (w)|∇w| 2 . and C solve the degenerate system of linear algebraic equations (a1 − k)A + a2 B + a3 C = 0.3. hence. (1) 6.1.

ν) ←− (y. There are “two-dimensional” solutions of the following forms: w = U (x. + + 2−n 2−m 2−k z 2−k y 2−m + . µ) 4. ζ 2 = 4 © 2004 by Chapman & Hall/CRC . THREE-DIMENSIONAL EQUATIONS INVOLVING ARBITRARY FUNCTIONS 427 where the function w(r) is determined by the ordinary differential equation wrr + A w = f (w). r2 = 4 e−λx e−µy e−νz + + . z) = U (x. ρ2 = 4 This solution can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of variables and determining parameters: (x. a. n) (z. a(2 − n)2 b(2 − m)2 cν 2 1−n 1−m + 2−n 2−m where the function w(r) is determined by the ordinary differential equation wrr + A w = f (w). z) = U (x. a. b. Functional separable solution for n ≠ 2. a(2 − n)2 cν 2 y 2−m x2−n + . There is a “two-dimensional” solution of the form e−µy e−νz + . and ν ≠ 0: w = w(r). and ν ≠ 0: w = w(r). ∂x2 ∂y 2 ∂z 2 1◦ . c. m) 3. w(x. c. aλ2 bµ2 cν 2 ∂2w + beµy ∂2w + ceνz ∂2w where the function w(r) is determined by the ordinary differential equation 5 w = f (w). ξ 2 = 4 bµ2 cν 2 wrr + This solution can be used to obtain other “two-dimensional” solutions by means of the following cyclic permutations of the variables and determining parameters: (x. r r A=5−2 1 1 1 . b(2 − m)2 cν 2 e−νz x2−n + . 2 b(2 − m) c(2 − k)2 2◦ .6. λ) (z. ξ). r2 = 4 y 2−m e−νz x2−n + + . ξ). axn ∂2w ∂2w ∂2w + by m + ceνz = f (w). aeλx = f (w). There is a “two-dimensional” solution of the form w(x. a(2 − n)2 b(2 − m)2 w = W (z. ∂x2 ∂y 2 ∂z 2 1◦ . ρ). r r A=2 + 1.3. k) ←− (y. µ ≠ 0. r r 2◦ . η). ξ2 = 4 η2 = 4 e−νz y 2−m + . Functional separable solution for λ ≠ 0. 2◦ . m ≠ 2. b. ζ). y. y. w = V (y.

a(2 − n)2 cν 2 e−µy x2−n + . xn . k=1 ∂2U = ∂x2 k n gk (x1 . . ζ).4. µ ≠ 0. axn ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES ∂2w ∂2w ∂2w + beµy + ceνz = f (w). xn ) ∂w ∂xk . and ν ≠ 0: w = w(r). xn ) k=1 ∂U . ∂x2 ∂y 2 ∂z 2 e−µy e−νz x2−n + + . ζ 2 = 4 6. 2◦ . . . . η). w) 1. The substitution U= leads to the linear equation dw . . bµ2 cν 2 e−νz x2−n + . . ∂2w n ∂x2 k k=1 = f (w) k=1 ∂w ∂xk 2 n + k=1 gk (x1 .4. . a(2 − n)2 bµ2 ∂w ∂z = aw ln w + g1 (x)+g2 (y)+g3 (z) w. . .1. [f3 (z)χz ]z − aχ ln χ − [g3 (z) − C1 − C2 ]χ = 0. 6. ∂xk © 2004 by Chapman & Hall/CRC . w = V (y. Equations with n Independent Variables 6. y. There are “two-dimensional” solutions of the following forms: w = U (x. r2 = 4 where the function w(r) is determined by the ordinary differential equation wrr + 8 − 5n 1 w = f (w). .428 5. where the functions ϕ = ϕ(x). . F (w) n where F (w) = exp f (w) dw . ψ = ψ(y). 2 a(2 − n) bµ2 cν 2 1◦ . ξ). and χ = χ(z) are determined by the ordinary differential equations (C1 and C2 are arbitrary constants) [f1 (x)ϕx ]x − aϕ ln ϕ − [g1 (x) + C1 ]ϕ = 0. ξ2 = 4 η2 = 4 w = W (z. Equations of the Form ∂ f (x ) ∂w + · · · + ∂ f (x ) ∂w ∂x1 1 1 ∂x1 ∂xn n n ∂xn n = g(x1 . 2−n r r e−µy e−νz + . . Functional separable solution for n ≠ 2. [f2 (y)ψy ]y − aψ ln ψ − [g2 (y) + C2 ]ψ = 0. z) = ϕ(x)ψ(y)χ(z). ∂ ∂x f1 (x) ∂w ∂x + ∂ ∂y f2 (y) ∂w ∂y + ∂ ∂z f3 (z) Multiplicative separable solution: w(x.

n 3. Zhurov (1998). k=1 1◦ . 2 − mk Reference: A. I. b k λ2 k B= k=1 2 − 1. Reference: A. I. ak (2 − mk )2 n where the function w(r) is determined by the ordinary differential equation 4 d2 w B dw + = f (w). For f (w) = beβw .6. r2 = A n k=1 e−λk xk . Polyanin and A.4. r2 = A k=1 where the function w(r) is determined by the ordinary differential equation 4 d2 w B dw + = f (w). r2 = A n k=1 x2−mk k . there is an exact solution of the form w=− 1 §0 1 ln bβ β n k=1 e−λk xk ak λ 2 k . there is an exact solution of the form w= p b(1 − p)2 1 p−1 n k=1 e−λk xk ak λ 2 k 1 1−p . 2 dr r dr A Example 1. 2 − mk 2−m 1 1−p there is an exact solution of the form p + 1−p n k=1 1 b(1 − p) 1 2 − mk n k=1 xk k ak (2 − mk )2 . 2 4. D. a k λ2 k where the function w(r) is determined by the ordinary differential equation 4 d2 w 1 dw − = f (w). For f (w) = w= bw p . Polyanin and A. B= k=1 1 p−1 2 − 1. EQUATIONS WITH n INDEPENDENT VARIABLES n 429 2. Example 2. 2 dr r dr A s © 2004 by Chapman & Hall/CRC 2 ∂w ∂ a k x mk k ∂xk ∂xk n + k= +1 ∂ ∂w b k eλk x k ∂xk ∂xk s = f (w). D. k=1 ∂ ∂xk a k eλk x k ∂w ∂xk = f (w). k=1 ∂ ∂xk a k x mk k ∂w ∂xk = f (w). x2−mk k +A ak (2 − mk )2 n k=s+1 e−λk xk . Functional separable solution: w = w(r). For f (w) = beλw . λ 2bλ n B= k=1 2 − 1. Zhurov (1998). Functional separable solution: w = w(r). Functional separable solution: w = w(r). Example 2. 2 − mk . For f (w) = bw p . there is an exact solution of the form w=− 1 §0 1 ln λ n k=1 xk k ak (2 − mk )2 2−m + 1 1−B ln . dr2 r dr A Example 1.

. I.2.1. 5. Cn are related by a single constraint.1). . k = 1. and 5. β 2cβ s B= k=1 2 − 1. 0 ≤ q ≤ s ≤ p ≤ n. . . 2.2. k=1 fk (xk ) ∂2w ∂x2 k n + k=1 gk (xk ) ∂w ∂xk n = aw ln w + w k=1 hk (xk ). 2 − mk 2◦ . xn ) = ϕ1 (x1 )ϕ2 (x2 ) . k=1 ∂ ∂xk fk (xk ) ∂w ∂xk n = aw ln w + w k=1 gk (xk ). Example 2. ϕ2 = ϕ2 (x2 ). For f (w) = ceβw .4. . . xn ) = ϕ1 (x1 )ϕ2 (x2 ) . x2 . 5.430 ELLIPTIC EQUATIONS WITH THREE OR MORE SPACE VARIABLES Example 1. 2 − mk For B1 = B2 = 0 and A1 = A2 = 1. . ϕn (xn ). . ϕn = ϕn (xn ) are determined by the ordinary differential equations dϕk d fk (xk ) − aϕk ln ϕk − gk (xk ) + Ck ϕk = 0. ϕn (xn ).3. . this equation arises in plane problems of heat and mass transfer (see equations 5. For f (w) = cw p . 5. . C1 + · · · + Cn = 0. D. .4. .3. .2. Then we obtain the following equation for w: ∂ 2 w B1 ∂w ∂ 2 w B1 ∂w + + A2 + A1 ∂y 2 y ∂y ∂z 2 z ∂z q B1 = k=1 2 − 1. Cn are related by a single constraint. . . y 2 = A1 ak (2 − mk )2 b k λ2 k k=s+1 k=1 s z 2 = A2 k=q+1 x2−mk k + A2 ak (2 − mk )2 n k=p+1 e−λk xk . 6. Polyanin and A. Multiplicative separable solution: w(x1 . ϕn = ϕn (xn ) are determined by the ordinary differential equations dϕk d 2 ϕk − aϕk ln ϕk − hk (xk ) + Ck ϕk = 0. 4 §3 Reference: A. . . .3. . where p q x2−mk e−λk xk k + A1 . . where the functions ϕ1 = ϕ1 (x1 ). . Zhurov (1998). .1. n. . b k λ2 k = 4f (w). Multiplicative separable solution: w(x1 .1. . there is an exact solution of the form w= 1 c(1 − p) p + 1−p s 1 p−1 1 1−p k=1 1 2 − mk s k=1 e−λk xk xk k + 2 ak (2 − mk ) b k λ2 k k=s+1 n 2−m .1.1. z).1. Other Equations n 1. fk (xk ) 2 + gk (xk ) dxk dxk The arbitrary constants C1 . . 2 − mk s B2 = k=q+1 2 − 1. x2 . . We divide the equation variables into two groups (responsible for both power-law and exponential terms) and look for exact solutions in the form w = w(y. k = 1. .1. 2. 5. n. . . . . . where the functions ϕ1 = ϕ1 (x1 ). © 2004 by Chapman & Hall/CRC . ϕ2 = ϕ2 (x2 ).3. .1.1. dxk dxk The arbitrary constants C1 . . n 5. there is an exact solution of the form w=− 1 ln β s k=1 xk k e−λk xk + 2 ak (2 − mk ) b k λ2 k k=s+1 n 2−m + 1 1−B ln . . .1. C1 + · · · + Cn = 0.

For f (w) = ceβw . Example 2. there is an exact solution of the form w=− 1 ln β n k=1 xk k ak (2 − mk )2 2−m + 1 1−B ln . r2 = A x2−nk k +A ak (2 − nk )2 m1 k=1 k=1 e−λk xk . For f (w) = cw p . there is an exact solution of the form w= 1 2c(1 − p) 1+p +B 1−p 1 p−1 n k=1 xk k ak (2 − mk )2 2−m . B = 2n − 1 − 2 k=1 bk .4. r2 = A n k=1 x2−mk k ak (2 − mk )2 n where the function w(r) is determined by the ordinary differential equation 4 d2 w B dw + = f (w). k=1 ak e λk x k ∂2w ∂x2 k n + k=1 b k eλk x k ∂w ∂xk = f (w). Functional separable solution: w = w(r). β 2cβ m1 4. a k λ2 k n where the function w(r) is determined by the ordinary differential equation 4 d2 w B dw + = f (w). there is an exact solution of the form w=− 1 ln β n k=1 e−λk xk ak λ 2 k m2 + 1 1−B ln . r2 = A n k=1 e−λk xk . For f (w) = cw p . Functional separable solution: w = w(r). k=1 ak xn k k ∂2w ∂x2 k + ak pk xnk –1 k ∂w ∂xk + k=1 m1 k=1 b k eλk x k ∂2w ∂x2 k m2 + b k qk e λ k x k ∂w ∂xk = f (w). 2 dr r dr A B=2 1 − nk + pk −2 2 − nk qk + 2m2 − 1. λk © 2004 by Chapman & Hall/CRC . Example 2. k=1 a k x mk k ∂2w ∂x2 k n + k=1 bk xmk –1 k ∂w ∂xk = f (w). 2 dr r dr A Example 1. For f (w) = ceβw . b k λ2 k m2 k=1 where the function w(r) is determined by the ordinary differential equation 4 d2 w B dw + = f (w). dr2 r dr A B=2 k=1 ak (1 − mk ) + bk − 1. β 2cβ n 3. ak (2 − mk ) 1 1−p Example 1.6. a k λk there is an exact solution of the form 1 2c(1 − p) 1+p +B 1−p 1 p−1 n k=1 w= e−λk xk ak λ 2 k 1 1−p . Functional separable solution: w = w(r). EQUATIONS WITH n INDEPENDENT VARIABLES n 431 2.

Semilinear equations.1. Hunter and R. 1◦ . Then the function w1 = C1 w(C2 x + C2 ϕ(t). See Section S. M. t) is a solution of the equation in question. 1. Dt (wxx ) 2a+1 + Dx −w(wxx ) 2a+1 = 0. J.3 with f (u) = au2 . which can be reduced to the canonical form by standard transformations. Saxton (1991).1. C2 .1.1. where C1 . Calogero (1984). where f (t) and g(z) are arbitrary functions and z is the parameter. Equations Linear in the Mixed Derivative 7. and C3 are arbitrary constants. K. is also a solution of the equation. ∂x∂t ∂x2 This is a special case of equation 7. 3◦ . see Section 3. © 2004 by Chapman & Hall/CRC . 7. General solution in parametric form: w = ft (t) + x = −f (t) + g(z) − at g(z) − at 1−a a 1 a dz. ¡ ¢ 1 1 References: F. and ϕ(t) is an arbitrary function.1. Pavlov (2001).1 for information about semilinear equations.5. ∂2w ∂x∂t =w ∂2w ∂x2 +a ∂w ∂x 2 ∂2w =w ∂2w . are not considered in this chapter. V. For hyperbolic equations that contain mixed derivatives.1. dz. 2. 2◦ .Chapter 7 Equations Involving Mixed Derivatives and Some Other Equations Preliminary remarks. Calogero Equation + a. Suppose w(x.1. Conservation laws: Dt (wx )1/a + Dx −w(wx )1/a = 0. This is a special case of equation 7.3 with f (u) = a. C1 C2 t + C3 ) + ϕt (t).

∂y which is independent of z and can be treated as an ordinary differential equation. we obtain the equation ∂w u ∂w u = =⇒ = . y)) where w = w(z. Introduce the notation u du ∂w . Pavlov (2001). 2◦ . (5). where ψ(t) is an arbitrary function and the function ϕ(t) is deﬁned implicitly by (C is an arbitrary constant) dϕ = t + C. Formulas (4). t) = ϕ(t)x + ψ(t). (5) w= Φ(u) where ψ(y) is an arbitrary function.434 3. The inverse of transformation (2) has the form 1 dz − w dy. is also a solution of the equation. (6) dx = Φ(u) Integrating the ﬁrst relation in (6) yields y z dξ − w(z0 . where C1 and C2 are arbitrary constants and ϕ(t) is an arbitrary function. Then the function −1 w1 = C1 w(C1 x + C1 ϕ(t). dt = dy. v = Φ(u) = exp . τ ) dτ . Integrating yields its solution in implicit form: du = y + ϕ(z). ¤ ¢£ References: F. 1◦ . y in accordance with the rule ∂ ∂ ∂ ∂ ∂ =v . Suppose w(x. (4) f (u) where ϕ(z) is an arbitrary function. Calogero (1984). and u = u(z. With the ﬁrst relations of (1) and (3). (7) x= y0 z0 Φ(u(ξ. dy = dt dz = ∂x ∂t deﬁnes the passage from x. t + C2 ) + ϕt (t). and x0 and y0 are any numbers. (1) u= ∂x f (u) The transformation ∂z ∂z dx + dt (2) dz = v dx + vw dt. t to the new independent variables z. and (7) with y = t deﬁne the general solution of the equation in question in parametric form (z is the parameter). f (ϕ) 3◦ . y) is deﬁned implicitly by (4). Degenerate solution linear in x: w(x. EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂w . = + vw . t) is a solution of this equation. ∂z v ∂z Φ(u) whose general solution is given by u dz + ψ(y). © 2004 by Chapman & Hall/CRC . (3) ∂x ∂z ∂t ∂y ∂z This results in the ﬁrst-order equation ∂u = f (u). M. y) is deﬁned by (5). V. ∂x ∂2w ∂2w +f =w ∂x∂t ∂x2 Calogero equation.

t)x + h(y. C1 4t(x + ϕ) − (y + C2 )2 + ϕt .1. It describes the propagation of a sound beam in a nonlinear medium. t). y. With λ = 0. z = w − ϕt (t). Reissner. w2 = w(x + λy + ϕ(t). . . and C1 and C2 are arbitrary constants. Khokhlov–Zabolotskaya Equation = 0. w(x. M. gyy = −6f g + 2ft . ∂2w –w ∂2w – ∂w 2 – ∂2w 1◦ . Dx = and the function Φ(u) is deﬁned in (1). Conservation law: where Dt = ¦ ¢¥ Dt [Φ(wx )] + Dx [−wΦ(wx )] = 0. t). 1 1 1 (ϕy + ψ)4 + ϕt y 3 + ψt y 2 + χy + θ. Solution in implicit form: tz + x + λy + λ2 t + ϕ(t) = F (z). y + 2λt. 4◦ . y. this relation determines the general y-independent solution of the original equation. t) = 2ϕx + (ϕt − 2ϕ2 )y 2 + ψy + χ. t) + ϕt (t) − λ2 . . 7. Calogero (1984). ψ = ψ(t). see Lin. Suppose w(x. t) = (ϕy + ψ)x − 12ϕ2 6 2 2 w(x. y. . w(x. t and y play the role of the space coordinates and x is a linear combination of time and a coordinate. χ = χ(t). and then substitute w = −∂u/∂x. 3◦ . EQUATIONS LINEAR IN THE MIXED DERIVATIVE 435 4◦ . V. ∂ ∂t . and θ = θ(t) are arbitrary functions.1. t) are determined by the system of differential equations fyy = −6f 2 . To this end.1 with a = b = 1/2) 2uxτ + ux uxx − uyy = 0 can be reduced to the Khokhlov–Zabolotskaya equation. where the functions f = f (y. ∂x∂t ∂x2 ∂x ∂y 2 Two-dimensional Khokhlov–Zabolotskaya equation. Pavlov (2001). ∂ ∂x . © 2004 by Chapman & Hall/CRC . y. where C1 . the prime stands for the differentiation. where ϕ(t) and F (z) are arbitrary functions. Solutions: w(x. hyy = −2f h + gt − g 2 .3. and Tsien (1948). “Two-dimensional” generalized separable solution quadratic in x: w = f (y.1. C1 C2 t + C5 ). t) = C1 x + C2 y + ϕ + ϕt − C2 . C2 y + C4 . y. t). The equation of unsteady transonic gas ﬂows (see 7.2. differentiate the equation with respect to x. and h = h(y.7. t) = t where ϕ = ϕ(t). t) = − x + ϕy + ψ. References: F. 1. y. C5 and λ are arbitrary constants and ϕ = ϕ(t) is an arbitrary function. t)x2 + g(y. g = g(y. t) is a solution of the equation in question. t + C1 w(x. are also solutions of the equation. Then the functions −2 2 −1 2 w1 = C1 C2 w(C1 x + C3 . 2◦ . one should pass to the new variable τ = 2t.

where the function u = u(ξ.5. “Two-dimensional” solution: αt + 4 x.1: Y. pp. 1995. z = x + βy + λt. where A and B are arbitrary constants. . A particular solution of this system is given by C1 (t) ϕ (t) 1 . 8◦ . g= + C2 (t)R3 − t . F. Vinogradov and I. Solution: ∂V ∂V + 2q ∂p ∂q 2 + (36pV + 5) ∂V ∂V + 22qV + 10V 2 = 0. D. we obtain an equation of the form 5. . Ibragimov (1994. Krasil’shchik (1997). r = (At + B)−1 x−1 y 2 . A. A. “Two-dimensional” solution: w = xu(ξ. where β. S. and µ are arbitrary constants. ¨ ¢§ References for equation 7. z) is determined by the differential equation (λ − β 2 ) ∂2U ∂2U ∂2U ∂2U + (µ − 2β) − −U − 2 2 ∂z ∂r∂z ∂r ∂z 2 ∂U ∂z 2 = 0. t). 2 R2 (gt − g 2 ) dy. 7◦ . © 2004 by Chapman & Hall/CRC . . ∂ζ β= αt + 4 . “Two-dimensional” solution: w = U (r. q).436 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS The subscripts y and t denote the corresponding partial derivatives. and the function U = U (r. where the function V = V (p. and the function u = u(r) is determined by the ordinary differential equation r2 (u − Ar + 4)urr + r2 (ur )2 − r(6u − Ar + 6)ur + 6(u + 1)u = 0.1. N. α The last equation has a particular solution of the form v = ζϕ(t).2. 299– 300. “Two-dimensional” solution: w = x−2 V (p. With λ = β 2 and µ = 2β. α where α = α(t) is an arbitrary function and the function v = v(ζ. t) + ξ = yx−1/2 . Gibbons (1989). q) is determined by the differential equation 3p(3V p + 1) ∂2V ∂ 2V ∂2V + (4q 2 V + 1) 2 + 2q(6pV + 1) 2 ∂p ∂q ∂p∂q + 3p 9◦ . . λ. 2 2 R R 2R R2 1 1 C3 (t) 2 + C4 (t)R + (gt − g 2 ) dy − h= R 3 R 3R where ϕ(t). Kodama (1988). H. R = y + ϕ(t). p = tx−3 . ∂ξ ∂t 6◦ . Y. where the function ϕ = ϕ(t) is determined by the Riccati equation αϕt − α2 ϕ2 − (αt + 10)ϕ + βt − β 2 = 0. t) is determined by the differential equation 2ξ − 5ξu ∂u ∂u −4 + 4u2 = 0.1. Kodama and J. ∂p ∂q w = u(r)x2 y −2 . ∂u ∂2u ∂2u + (ξ 2 u + 4) 2 + ξ 2 ∂ξ∂t ∂ξ ∂ξ w = v(ζ. M. r = y + µt. z). pp.1. C1 (t). C4 (t) are arbitrary functions. 447–450). ζ = y 2 + αx. t) is determined by the differential equation α ∂v ∂2v ∂2v − (α2 v + 4ζ) 2 − α2 ∂ζ∂t ∂ζ ∂ζ 2 − (αt + 10) ∂v + βt − β 2 = 0. f =− 5◦ . q = yx−2 . Zaitsev (2002). Polyanin and V.

+b ∂2w = 0. D. λ is an arbitrary constant. t) + ϕ(t). ψ = ψ(t). t)x + χ(y. where C1 . and χ = χ(y. y. where the function ϕ(t) is an arbitrary function. C3 . 2◦ . Zaitsev. χ = χ(t). z = x + λy. t). y. the prime denotes a derivative with respect to t. t) = U (z. 3◦ . EQUATIONS LINEAR IN THE MIXED DERIVATIVE 437 ∂w ∂ ∂2w w +a ∂x∂t ∂x ∂x The transformation 2. © 2004 by Chapman & Hall/CRC . t) = 2ϕx + f dt + C −1 + ϕy + ψ. w2 = w(ξ. Polyanin and V. Solutions: w(x. a − τ = −bt ∂ 2u = 0. τ ). t) is determined by the ﬁrst-order partial differential equation [ψ(t) is an arbitrary function] ∂U ∂U ∂U − f (t)U − [f (t)ϕ(t) + λ2 g(t)] = ψ(t). t). gψyy = −6f ϕψ + 2ϕt . The subscripts y and t denote the corresponding partial derivatives. g ϕ ψ f (ϕy + ψ)4 + t y 3 + t y 2 + χy + θ. and θ = θ(t) are arbitrary functions. y. gχyy = −f (2ϕχ + ψ 2 ) + ψt .1. ∂y 2 w(x. “Two-dimensional” generalized separable solution quadratic in x: w = ϕ(y. Then the functions −2 2 w1 = C1 w(C1 x + C2 . ¢© Reference: A. t) is a solution of the equation in question. t). t) are determined by the system of differential equations gϕyy = −6f ϕ2 . Suppose w(x. η. t) = leads to an equation of the form 7. η = y + 2λ g(t) dt. ∂t ∂z ∂z A complete integral of this equation is sought in the form U = A(t)z + B(t). C2 . y. C is an arbitrary constant. “Two-dimensional” solution: w(x. t). and Moussiaux. 2002). and λ are arbitrary constants and ϕ = ϕ(t) is an arbitrary function. ξ = x + λy + [f (t)ϕ(t) + λ2 g(t)] dt. f = f (t) and g = g(t). ∂y 2 b u(x.7.1: ∂ ∂u ∂2u − u ∂x∂τ ∂x ∂x 3. which allows obtaining the general solution (see Polyanin. where the function ϕ = ϕ(y. ∂x∂t ∂x ∂x ∂y 2 Generalized Khokhlov–Zabolotskaya equation. Zaitsev (2002). F. y. w(x. are also solutions of the equation. y. ψ = ψ(y. t) = −x w(x. ϕt − 2f ϕ2 2 y + ψy + χ. f = f (t) and g = g(t).1. 4◦ . t) = (ϕy + ψ)x − 2 12gϕ 6g 2g where ϕ = ϕ(t). = 0. y. t)x2 + ψ(y. w ∂2w – f (t) ∂ ∂w – g(t) ∂2w 1◦ .2. C1 y + C3 . t) + ϕ(t). and the function U = U (z.

z. © 2004 by Chapman & Hall/CRC . Three-dimensional Khokhlov–Zabolotskaya equation. gyy + gzz = ft − f 2 . where C1 . y cos β + z sin β. z. α2 . ξ = y sin β + z cos β. and h = h(y. Solutions: w(x. z. z. C2 z + C5 . for example. and t denote the corresponding partial derivatives. y. and β are arbitrary constants. λ. µ. where β is an arbitrary constant and the function u = u(x. For solutions of these linear equations. t). z. “Three-dimensional” generalized separable solution linear in x: w = f (y. γ are arbitrary functions of t. z. . β. z.1: ∂2u − ∂x∂t ∂u ∂x 2 −u ∂2u ∂2u − = 0. z. C1 C2 t + C6 ). g = g(y. t). z + 2µt. z. Then the functions −2 2 −1 2 w1 = C1 C2 w(C1 x + C3 . . gyy + gzz = −6f g + 2ft . C2 y + C4 . t3/2 where α1 . t). z. and C is an arbitrary constant. 1◦ . hyy + hzz = −2f h + gt − g 2 . ξ. “Three-dimensional” solution: w = u(x.2. ∂2w ∂x∂t – EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂w ∂x 2 –w ∂2w ∂x2 – ∂2w ∂y 2 – ∂2w ∂z 2 = 0. where the functions f = f (y. −y sin β + z cos β. y + 2λt. 3◦ . 5◦ . t)x + h(y. z. “Three-dimensional” generalized separable solution quadratic in x: w = f (y. t) is determined by the Khokhlov– Zabolotskaya equation of the form 7. y. C6 . t) are determined by the system of differential equations fyy + fzz = −6f 2 . Suppose w(x. ξ. z. are also solutions of the equation.438 4. t) are determined by the differential equations fyy + fzz = 0. t). t). see. Tikhonov and Samarskii (1990) and Polyanin (2002). t) = C 4tx − y 2 − z 2 . t) + ϕt (t) − λ2 − µ2 . t) and g = g(y. t). The subscripts y. where the functions f = f (y. α3 . t)x + g(y. z. 1 w(x. t) is a solution of the equation in question. The ﬁrst equation represents the Laplace equation and the second one is a Poisson equation (for g).1. w3 = w(x. . and ϕ = ϕ(t) is an arbitrary function. z. y. t) = 2α1 x + (α1 − 2α2 − α2 )y 2 + α3 y + α2 z 2 + βz + γ. . t)x2 + g(y. ∂x2 ∂ξ 2 4◦ . 2◦ . w2 = w(x + λy + µz + ϕ(t).

z.2. ∂t∂x ∂x ∂x ∂y 2 ∂z 2 1◦ .7. y. a a where C1 . Then the functions −2 2 −1 2 w1 = C1 C2 w(C1 x + C3 . y = y −b. ζ = . y − 2bλt. t2 xt where the function U = U (ζ) is determined by the ordinary differential equation w(x. C6 . z. Integrating yields (η − 1)eη = C1 p. Solution: 4tx − y 2 z2 V (q). √ C1 ξ + C 2 . H. z. Lychagin (1986). the passage to the new independent variables according to √ √ √ √ ¯ ¯ ¯ x = x −a. where the functions f = f (y. t). z − 2cµt. λ. = 0. q = . t) = α ln(cy 2 + bz 2 ) − (βt + 4abcβ 2 )(cy 2 + bz 2 ) + 4bcβx + γ. © 2004 by Chapman & Hall/CRC 2 +b ∂2w +c ∂2w (1) (2) . y. S. y. I. 2◦ . t) = u(ξ)t−λ . the change of variable ξ(u) = p(u) − 1−λ u. ∂2w +a ∂ ∂w bλ2 + cµ2 1 ϕt (t) − . where α = α(t). Solutions: w(x. Solution: References for equation 7. z = z −a. C2 z + C5 . t) are determined by the differential equations bfyy + cfzz = 0. bgyy + cgzz = −ft − af . M. 4 For λ ≠ 1. are also solutions of the equation. t) − 3◦ . Krasil’shchik. where λ is an arbitrary constant. β = β(t). . . .4. z. 8◦ . t) = αy + βz + 1 x + γ. 2 t z2 where the function V = V (q) is determined by the ordinary differential equation w(x. we have u(ξ) = y2 + z 2 y2 + z 2 U (ζ). the passage to the inverse ξ = ξ(u). and γ = γ(t) are arbitrary functions and C is an arbitrary constant. z. z. V. y.4: A. ¢ 7◦ . and c < 0. and the 4 reduction of order with pu = 1−λ η(p) result in the ﬁrst-order equation pηηp − η + 1 = 0. y. t) is a solution of the equation in question. For λ = 1. EQUATIONS LINEAR IN THE MIXED DERIVATIVE 439 6◦ . t)x + g(y. and V. “Three-dimensional” generalized separable solution linear in x: w = f (y.1. z. b < 0. 1995). Vinogradov. N. C2 y + C4 . and the function u = u(ξ) is determined by the ordinary differential equation [4u + (1 − λ)ξ]uξξ + 4(uξ )2 = 0. Suppose w(x. 4◦ . t) = 2(4V + q 2 − q)Vqq + 8(Vq )2 + (1 − q)Vq + V = 0. t) and g = g(y. ξ = tλ−2 (4xt − y 2 − z 2 ). w leads to the three-dimensional Khokhlov–Zabolotskaya equation 7. Solution: w(x.2. µ. z.1. t) = ζ 2 (ζ 2 U − ζ + 4)Uζζ + ζ 4 (Uζ )2 + ζ(2ζ 2 U − 3ζ + 12)Uζ + 4U = 0. t = t/ −a ¯ 5. y. C1 C2 t + C6 ). w2 = w(x + λy + µz + ϕ(t).1. For a < 0. z. and β are arbitrary constants and ϕ = ϕ(t) is an arbitrary function. . at + C w(x. z. Ibragimov (1994.

and t. 6 . and t. t)x2 + g(y. t). where the function U = U (η. z. t) are determined by the system of differential equations bfyy + cfzz = −6af 2 . bgyy + cgzz = −6af g − 2ft . Remark. z. t). y. z = z |c|. p = txk+1 .3.2 with f (t) = a and g(t) = −b. t) = V (ζ. H. q = yxk/2 . and if bc < 0. Equation of Unsteady Transonic Gas Flows 1. t) is determined by the differential equation aϕ2 (aϕ2 V + 4bcζ) ∂V ∂2V ∂V ∂2V + aϕ3 + aϕ2 aϕ2 − ϕt + 12bc ∂ζ 2 ∂t∂ζ ∂ζ ∂ζ − ϕtt ϕ + 2(ϕt )2 − 12bcϕt + 16b2c2 = 0. ∂2w ∂x∂t +a ∂w ∂ 2 w ∂x ∂x2 –b ∂2w ∂y 2 = 0. V. N. z. ¢ References: P. For solutions of these linear equations. © 2004 by Chapman & Hall/CRC . equation (1) can be reduced to the Laplace equation. Tikhonov and Samarskii (1990) and Polyanin (2002). t) is determined by the differential equation η(aηU + 4bc) ∂ 2U ∂U ∂2U −η + aη 2 ∂η 2 ∂t∂η ∂η 2 − 2(aηU − 2bc) ∂U ∂U + + aU 2 = 0. z. z. t). Likewise. bhyy + chzz = −2af h − gt − ag 2 . t). b. see.1. y. “Two-dimensional” solution: w(x. 5◦ . t) − ϕt − 4bc x. √ ¯ y = y |b|. This remains true if the coefﬁcients a. There are “three-dimensional” solutions of the following forms: ◦ w(x. z. and V = V (ζ. S. where the functions f = f (y. z. z. t) = xU (η. η = (cy 2 + bz 2 )x−1 . b. The above remains true if the coefﬁcients a. Kucharczyk (1967). ξ = cy 2 + bz 2 .1. Remark.3. t)x + h(y. where k is an arbitrary constant. Sukhinin (1978). ∂η ∂t 8◦ . respectively. aϕ ζ = cy 2 + bz 2 + ϕx. g = g(y. 7◦ . z. y. to the ¯ wave equation. z. t) = v(p. r = zxk/2 . If bc > 0. and c are functions of y. Ibragimov (1994).440 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS The subscripts y. w(x. for example. where ϕ = ϕ(t) is an arbitrary function. z. Reissner. t. y. ξ). z. This is a special case of equation 7. t) = u(x. “Two-dimensional” solution: w(x. q. and c are functions of y. “Three-dimensional” generalized separable solution quadratic in x: w = f (y. then by the scaling z. and Tsien (1948). and h = h(y. r)xk+2 . 7. equation (2) can be reduced to a Poisson equation and a nonhomogeneous wave equation. see Lin.√ t denote the corresponding partial derivatives. This is an equation of an unsteady transonic gas ﬂow.

t) are determined by the differential equations bfyy = 18af 2 . bhyy = 6af h + 4ag 2 + 2gt . y. y. z = x + λy. The general solution of equation (1) can be written out in parametric form (Polyanin. λ is an arbitrary constant. ∂z (1) 2 1 U = C1 z + bλ2 C1 − 2 aC1 t + C2 . t) + ϕ(t)y 2 + ψ(t)y + χ(t)x + θ(t). λ is an arbitrary constant. t) is determined by the ﬁrst-order partial differential equation [σ(t) is an arbitrary function]: ∂U ∂U a ∂U + + aχ(t) − bλ2 = 2bϕ(t) − χt (t) z + σ(t). and the function U = U (z. h = h(y. t)x + r(y. µ = µ(t). “Two-dimensional” solution of a more general form: w(x. t) + ϕtt (t)y 2 + 2bϕt (t)x + ψ(t)y + χ(t). g = g(y. t). and r = r(y. and δ = δ(t) are arbitrary functions. t) is a solution of the equation in question. w2 = w(ξ.7. z = x + λy. Then the functions −3 2 −1 2 w1 = C1 C2 w(C1 x + C3 . t) + ϕ(t)y + ψ(t). Solution: − bλ2 ∂U = 0. are also solutions of the equation. and θ = θ(t) are arbitrary functions. w(x. t) is determined by the ﬁrst-order partial differential equation ∂U a ∂U + ∂t 2 ∂z A complete integral of this equation is given by 2 2◦ . t)x2 + h(y. The subscripts y and t denote the corresponding partial derivatives. η = y + 2bλt. γ = γ(t). χ = χ(t). t) = f (y. t)x3 + g(y. t) = 3◦ . where ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. and χ = χ(t) are arbitrary functions. and Moussiaux. 2002): z + bλ2 − as t + fs (s) = 0. EQUATIONS LINEAR IN THE MIXED DERIVATIVE 441 1◦ . U = sz + bλ2 s − 1 as 2 t + f (s). one can integrate the system with respect to y to obtain a solution dependent on six arbitrary functions. Zaitsev. t). bgyy = 18af g + 3ft . + 2b where α = α(t). ξ = x + λy + bλ2 t − 2abϕ(t). t) = U (z. ∂t 2 ∂z ∂z This equation can be fully integrated—a complete integral is sought in the form U = f (t)z + g(t). where the functions f = f (y. © 2004 by Chapman & Hall/CRC 2 .1. “Two-dimensional” solution: w(x. where ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. C2 y + C4 . t). 2 4◦ . where C1 and C2 are arbitrary constants. η. 1 1 (γ + 6aγγt + 4a2 γ 3 )y 4 + (α + 2aαγ)y 3 12b2 tt 6b t 1 [2(γt + 2aγ 2 )x + βt + 2aβγ]y 2 + (αx + δ)y + γx2 + βx + µ. t). ψ = ψ(t). and the function U = U (z. y. Setting f = 0 and g = ϕ(t)y +ψ(t). where f = f (s) is an arbitrary function and s is the parameter. where ϕ = ϕ(t). bryy = 2agh + ht . Suppose w(x. C1 C2 t + C5 ) + C6 yt + C7 y + C8 t + C9 . ψ = ψ(t). y. 5◦ . y. where the Cn and λ are arbitrary constants and ϕ = ϕ(t). β = β(t). t) = U (z. “Two-dimensional” generalized separable solution cubic in x: w(x.

λ is an arbitrary constant. 12g b=− αt + 2f αγ . y. a complete integral is sought in the form U = f (t)z + g(t). t) = v(p. © 2004 by Chapman & Hall/CRC . Suppose w(x. 6g c=− γt + 2f γ 2 . r = yt−1/2 . N. Mamontov (1969). Vorob’ev. µ = µ(t). p = y 2 + γx. −4 2 w1 = C1 w(C1 x + C2 . Generalized separable solution in the form of a polynomial of degree 4 in y: w(x. µ = µ(t). t) + γγtt −2(γt )2 −18bγt −40b2 4 y + 12abγ 3 where γ = γ(t). Semenova (1989). y. t) + C4 yt + C5 y + C6 t + C7 .442 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 6◦ . Polyanin and V. ∂x∂t ∂x ∂x2 ∂y 2 1◦ . are also solutions of the equation. 2g 3◦ . 2 ∂x∂r ∂x ∂x ∂r ∂r 4b + γt p − δ y 2 + µy + λ. and θ = θ(t) are arbitrary functions. β = β(t). . and δ = δ(t) are arbitrary functions. t) is a solution of this equation. λ = λ(t). c = c(t). “Two-dimensional” solution: w(x. t) = a(t)y 4 + b(t)y 3 + [c(t)x + d(t)]y 2 + [α(t)x + β(t)]y + γ(t)x2 + µ(t)x + δ(t). and E. . “Two-dimensional” solution: w(x. E. and the function U = U (z. ∂z This equation can be fully integrated. A. ψ = ψ(t). and χ = χ(t) are arbitrary functions. z = x + λy. Then the functions 2. γ = γ(t).3. ϕ (t) w2 = w(ξ. and δ = δ(t) are arbitrary functions. t) is determined by the ﬁrst-order partial differential equation [σ(t) is an arbitrary function]: ∂U ∂U 1 + f (t) ∂t 2 ∂z 2 + [f (t)χ(t) + λ2 g(t)] ∂U = −[2g(t)ϕ(t) + χt (t)]z + σ(t). . Ignatovich. g d=− µt + 2f γµ . C7 and λ are arbitrary constants and ϕ = ϕ(t). + (γt − 2b) − ∂p2 ∂t∂p ∂p aγ 3 References for equation 7. t) − t y 2 + ψ(t)y + ϕ(t)x + χ(t). = 0. aγ 3 7◦ .1. 2◦ . .1: E. χ = χ(t). y. ψ = ψ(t). 2g(t) ∂2w + f (t) ∂w ∂ 2 w + g(t) ∂2w ξ = x + λy − λ2 g(t) + f (t)ϕ(t) dt. V. where the function v = v(x. t) + ϕ(t)y 2 + ψ(t)y + χ(t)x + θ(t). V. where α = α(t). O. η. t) = U (z. b = b(t). where ϕ = ϕ(t). and d = d(t) are given by a=− ct + 2f γc . t) is determined by the differential equation γt p + aγ 3 ¢ ∂v ∂p ∂2v ∂2v ∂v 2b[p(γt + 4b) − aγ 3 δ] +γ = 0. y. y. “Two-dimensional” solution: w(x. D. r) is determined by the differential equation r ∂v ∂ 2 v ∂2v ∂v ∂2v − 2a + 2b 2 + 2 = 0. F. C1 y + C3 . where C1 . M. Zaitsev (2002). t) = v(x. and the function v = v(p. r)t−1 . η = y − 2λ g(t) dt. and the functions a = a(t).

where the functions ϕ = ϕ(y. t). y. where h = h(t) is an arbitrary function. t). are also solutions of the equation. The subscripts y and t denote the corresponding partial derivatives. t) = u(p. t) are determined by the differential equations gϕyy + 18f ϕ2 = 0. © 2004 by Chapman & Hall/CRC . the constants of integration will be functions of t. “Two-dimensional” generalized separable solution cubic in x: w(x.1. gθyy + 2f ψχ + χt = 0. µ = µ(t). and ϕ = − 3f (y + h)2 5◦ . ∂y ∂x∂y ∂x ∂y 2 w(x. ∂p2 ∂p where the functions a = a(t) and b = b(t) are given by a=− ¢ (bγ)t + 10gb . Zaitsev (2002). and λ = λ(t) are arbitrary functions. 12g b= γt − 4g . χ = χ(y. and θ = θ(y. p. y) is a solution of this equation. Suppose w(x. Polyanin and V. y. C1 y + ϕ(x)) + C2 . t)x3 + ψ(y. EQUATIONS LINEAR IN THE MIXED DERIVATIVE 443 4◦ . ∂w ∂y ∂x∂y ∂x ∂y 2 ∂x ∂y 1. “Two-dimensional” solution: w(x.7. gχyy + 6f ϕχ + 4f ψ 2 + 2ψt = 0. y) = F y + G(x) . F. t) is determined by the differential equation γ ∂u ∂2u + γt p + f γ 3 ∂p∂t ∂p ∂2u ∂u + (γt + 2g) + 2g(bp + c) = 0.1. Equations of the Form ∂w ∂ w – ∂w ∂ w = F x. ∂w ∂ 2 w ∂y ∂x∂y – ∂w ∂ 2 w ∂x ∂y 2 = f (x). D.4. f = f (t) and g = g(t). ψ = ψ(y. t). c = c(t). Here. 397). The ﬁrst equation has the following particular solutions: ϕ = 0 g . t)x2 + χ(y. y. p = y 2 + γ(t)x. t) + a(t)y 4 + [b(t)p + c(t)]y 2 + µ(t)y + λ(t). These equations can be treated as ordinary differential equations for y with parameter t. Zwillinger (1989. General solution: where F (z) and G(x) are arbitrary functions. 2 2 7. where C1 and C2 are arbitrary constants and ϕ(x) is an arbitrary function. Then the functions −1 w1 = C1 w(x. f γ3 Reference: A. and the function u = u(p. ∂w . t). ∂w ∂ 2 w ∂w ∂ 2 w – = 0. gψyy + 18f ϕψ + 3ϕt = 0. ¢ Reference: D. 1◦ . γ = γ(t). 2. t)x + θ(y. t) = ϕ(y.

f (x) dx + C2 . are also solutions of the equation. C1 y + C2 ) + C3 . w2 = w x. y) = y 2 ! f (x) dx + C1 1/2 + ϕ(x). Generalized separable solutions: w(x. η = w. w(x. ∂y ∂x∂y ∂x ∂y 2 ∂y 1◦ . (1) brings the original equation to the ﬁrst-order nonlinear equation U ∂U = f (ξ). y) = C1 y 2 + ϕ(x)y + 1 ϕ2 (x) − 2 4C1 where ϕ(x) is an arbitrary function and C1 and C2 are arbitrary constants. where C1 . w(x. 2F (x) + ψ(w) where ϕ(x) and ψ(w) are arbitrary functions. η) = ∂w . 3◦ . The equation can be rewritten as the relation that the Jacobian of the functions w and v = wy − f (x) dx is equal to zero. ∂w ∂ 2 w – ∂w ∂ 2 w ! f (x) dx. U (ξ. 1 λ f (x) dx + C. we obtain the ﬁrst-order equation 2 ∂w = 2 f (x) dx + ψ(w). Then the functions −1 w1 = C1 w(x. 2◦ . y). ∂y where w = w(x. and C3 are arbitrary constants and ϕ(x) is an arbitrary function. ∂ξ (2) which is independent of η. C2 . The von Mises transformation ξ = x. y) is a solution of this equation. Generalized separable solutions linear and quadratic in y: w(x. On integrating (2) and taking into account the relations of (1). ∂w = f (x) . which means that v is expressible in terms of w: ∂w − ∂y © 2004 by Chapman & Hall/CRC f (x) dx = ϕ(w). (3) ∂y where ψ(w) is an arbitrary function. Suppose w(x. F (x) = 3. Integrating (3) yields the general solution in implicit form: dw √ = y + ϕ(x).444 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2◦ . (1) . 3◦ . y) = ϕ(x)eλy − where ϕ(x) is an arbitrary function and C and λ are arbitrary constants. y + ϕ(x) . It follows that w and v are functionally dependent. y) = y f (x) dx + C + ϕ(x).

y) is a solution of this equation. f ϕψyy + ϕy ψy = gψ + s. Other Equations with Two Independent Variables " 1. ∂y ∂x∂y ∂x ∂y 2 Generalized separable solution linear in x: w = ϕ(y)x + ψ(y).1. and C3 are arbitrary constants and ϕ(x) is an arbitrary function. Integrating yields its general solution in implicit form: ϕ(w) + −1 f (x) dx dw = y + ψ(x). C2 . Equation (1) can be treated as an ordinary differential equation in the independent variable y with parameter x. EQUATIONS LINEAR IN THE MIXED DERIVATIVE 445 where ϕ(w) is an arbitrary function. y + ϕ(x) . This equation can be treated as a ﬁrst-order ordinary differential equation in the independent variable y with parameter x. ∂w ∂ 2 w ∂w ∂ 2 w + f (y) = g(y)w + h(y)x + (y). 2 ∂y ∂x∂y ∂x ∂y ∂y ∂x First integral: ∂w = ϕ(w) − g(y) dy + f (x) dx. – −1 w1 = C1 w(x. 1934). ∂t ∂x2 ∂x ∂t ∂x∂t ∂x ∂t2 Born–Infeld equation (see Born and Infeld. Any solution of the ﬁrst-order equation (1) for any ϕ(w) is a solution of the original equation. ∂w ∂ 2 w ∂w ∂w ∂w ∂ 2 w – = f (x) + g(y) . = f (x)g ∂y ∂x∂y ∂x ∂y 2 ∂y 1◦ . where C1 . First integral: ∂w U dU = ϕ(w) + f (x) dx. ∂w ∂w ∂ 2 w ∂w 2 ∂ 2 w ∂w 2 ∂ 2 w +2 – 1+ = 0. 4. U= .1. ∂y where ϕ(w) is an arbitrary function. Then the functions 5. 7. 1– © 2004 by Chapman & Hall/CRC .7. C1 y + C2 ) + C3 . g(U ) ∂y where ϕ(w) is an arbitrary function. It is used in nonlinear electrodynamics and ﬁeld theory. 2◦ . Suppose w(x. where ψ(x) and ϕ(w) are arbitrary functions. 2.5. where the functions ϕ(y) and ψ(y) are determined by the system of ordinary differential equations f ϕϕyy + (ϕy )2 = gϕ + h. ∂w . ∂w ∂ 2 w ∂w ∂ 2 w w2 = w x. are also solutions of the equation. This equation can be treated as a ﬁrst-order ordinary differential equation in the independent variable y with parameter x.

◦ 3 . we can reduce this system to the second-order linear equation u2 ∂2ξ ∂2ξ ∂ξ ∂ξ ∂2ξ + (1 + 2uv) + v 2 2 + 2u + 2v = 0. v are treated as the independent variables and ξ. s= . Solutions: where ϕ(z1 ) and ψ(z2 ) are arbitrary functions. A. ∂t w = g(x) at t = 0. η = x + t. Integrals of this equation are given by r = C1 and s = C2 . ∂η ∂ξ ∂u ∂v ∂u − (1 + 2uv) + u2 = 0. 2◦ . w(x. ∂ξ v= ∂w . M. The hyperbolicity condition 1 + [fx (x)]2 − g 2 (x) > 0 is assumed to hold. Solution in parametric form: t= x= w= $ ¢# 1 2 β α 1 + [fζ (ζ)]2 1 + [fζ (ζ)]2 − g 2 (ζ) β α dζ. t) = ϕ(x + t). r= 2v 2u © 2004 by Chapman & Hall/CRC (1) Looking for solutions in the hyperbolic domain. u= ∂w . t) = ψ(x − t).446 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS w(x. Barbashov and N. we write out the equation of characteristics (2) . v2 ∂ξ ∂η ∂η The hodograph transformation (where u. Cauchy problem with initial conditions: w = f (x) at t = 0. 2 ∂u ∂u∂v ∂v ∂u ∂v u2 dv 2 − (1 + 2uv) du dv + v 2 du2 = 0. v2 ∂v ∂v ∂u On eliminating η. f (α) + f (β) 1 + 2 2 g(ζ) dζ 1 + [fζ (ζ)]2 − g 2 (ζ) . ∂η the Born–Infeld equation can be rewritten as the equivalent system of equations ∂u ∂v − = 0. Reference: B. where √ √ 1 + 4uv − 1 1 + 4uv − 1 . Chernikov (1966). By the introduction of the new variables ξ = x − t. 1◦ . η as the dependent ones) leads to the linear system ∂ξ ∂η − = 0. ∂v ∂u ∂ξ ∂ξ ∂η + (1 + 2uv) + u2 = 0. α+β 1 − 2 2 fζ (ζ)g(ζ) dζ 1 + [fζ (ζ)]2 − g 2 (ζ) β α .

w = Re f3 (z). ∂s ∂s Eliminating η yields the simple equation ∂2ξ = 0. The Euler transformation ∂u . Whitham (1974). y) + u(ξ. and C3 are arbitrary constants. z can be taken as one of the functions fk (z).7. EQUATIONS LINEAR IN THE MIXED DERIVATIVE 447 Passing in (1) to the new variables of (2). ∂ζ t= ∂u ∂τ ∂2u ∂2u ∂2u − 2ζτ − 1 + ζ2 = 0. It describes. ∂x τ= ∂w .1. ∂ξ∂η ∂η w(x.3) ∂u ∂ 2u = F (η. ∂x η= ∂w . C2 . τ ) = xζ + tτ . y) + u(ξ. ξ) . Hilbert (1989). leads to the linear equation 1 − τ2 3. 2 ∂η ∂ξ∂η ∂ξ 2 2◦ . 1◦ . ∂ξ leads to the linear equation (for details.2. 2◦ . x= y=η © 2004 by Chapman & Hall/CRC . The Legendre transformation w(x. ∂x∂y ∂x ∂y ∂x2 1◦ . leads to the linear equation ∂2u ∂2u ∂2u + 2bξη + c + ξ2 = 0. B. ◦ 4 . & ¢% Reference: G. for example. where the fk (z) are arbitrary analytic functions of the complex variable z = α + iβ with derivatives constrained by [f1 (z)]2 + [f2 (z)]2 + [f3 (z)]2 = 0. ∂r∂s whose solution is the sum of two arbitrary functions with different arguments. ξ = ϕ(r) + ψ(s). Then the function 4. For example. ∂y ∂x2 ∂x ∂y ∂x∂y ∂x ∂y 2 Equation of minimal surfaces (with a = b = c = 1). η) = xξ + yη. ∂w ∂w ∂ 2 w ∂2w = F y. . ∂ξ y= ∂u ∂η Reference: R. ∂τ 2 ∂ζ∂τ ∂ζ 2 + c+ ∂w 2 = 0. see Subsection S. y = Re f2 (z). t) + u(ζ. r2 ∂r ∂r (3) ∂ξ 2 ∂η +s = 0. where C1 . General solution in parametric form for a = b = c = 1: a + η2 x = Re f1 (z). The Legendre transformation w(x. ∂t x= ∂u . −1 w1 = C1 w(C1 x + C2 . Courant and D. a+ ∂w 2 ζ= ∂w . Suppose w(x. The function η is found from system (3). y) is a solution of this equation. is also a solution of the equation. we obtain ∂ξ ∂η + = 0. η) = xξ. the shape of a foam ﬁlm bounded by a given contour. y) + C3 . & ¢% ∂2w – 2b ∂w ∂w ∂ 2 w ∂2w ξ= ∂w . ∂y x= ∂u .

where σ(t) is an arbitrary function. leads to the linear equation f 2 − ξ2 ( ¢' ξ= ∂w . and χ = χ(t) are arbitrary functions. and C1 and C2 are arbitrary constants. y − 2bλt. where C1 . A complete integral of this equation has the form ϕ(t) dt + C1 . λ is an arbitrary constant. w1 = w(ξ. t) = U (z. Reference: A. are also solutions of the equation. y) + U (ξ. f2 – EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂w ∂w ∂ 2 w ∂w 2 ∂ 2 w ∂w 2 ∂ 2 w 2 2 –2 = 0. The Legendre transformation w(x. is also a solution of the equation. t) is determined by the ﬁrst-order partial differential equation ∂U +F ∂t ∂U ∂z + bλ2 ∂U + 2bϕ(t)z = σ(t). where the functions A(t) and B(t) are given by A(t) = −2b ( ¢' F (u) = f (u) du. Courant and D. f = f (wx + wy ). y. C1 t + C4 ) + C5 yt + C6 y + C7 t + C8 . ∂ξ y= ∂U ∂η ∂2U ∂2U ∂2U + 2ξη + f 2 − η2 = 0. C8 and λ are arbitrary constants and ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. Suppose w(x. η) = xξ + yη.448 5. ∂z U = A(t)z + B(t). y. D. where C1 . t) + ϕ(t)y 2 + ψ(t)y + χ(t). Then the function 2. where ϕ = ϕ(t). Hilbert (1989). Then the functions 1. y. η. t) + ϕ(t)y + ψ(t). . . “Two-dimensional” solution: w(x. Suppose w(x. ∂y x= ∂U . ξ = x + λy − λ2 g(t) dt + C1 . 2◦ . ∂η 2 ∂ξ∂η ∂ξ 2 f = f (ξ 2 + η 2 ). and λ are arbitrary constants and ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. t) is a solution of this equation. −1 w1 = C1 w(C1 x + C2 . Zaitsev (2002). . + f2 – ∂x ∂x2 ∂x ∂y ∂x∂y ∂y ∂y 2 This equation describes a two-dimensional steady-state isentropic ﬂow of a compressible gas. z = x + λy. ∂w ∂ 2 w ∂2w ∂2w + f (t)Φ + g(t) = 0. C1 y + C3 . C2 . © 2004 by Chapman & Hall/CRC . t) + ϕ(t)y + ψ(t). t) is a solution of this equation.1. ∂x∂t ∂x ∂x2 ∂y 2 1◦ . Other Equations with Three Independent Variables = 0. F. ∂x∂t ∂x ∂x2 ∂y 2 1◦ . . ∂x η= ∂w . ∂2w +f ∂w ∂2w +b ∂2w w2 = w(x + λy − bλ2 t. Polyanin and V. 7. B(t) = σ(t) − F (A(t)) − bλ2 A(t) dt + C2 . Reference: R. w is the velocity potential and f is the sound speed. η = y − 2λ g(t) dt + C2 . ψ = ψ(t). and the function U = U (z.6.

Equations Quadratic in the Highest Derivatives 2 2 7. C5 are arbitrary constants. y) is a solution of this equation. ∂t ∂z ∂z where σ(t) is an arbitrary function. where C1 . Generalized separable solutions: x (x − t)f (t) y dt + C3 xy + C4 x + C5 y + C6 . ∂2w ∂2w w(x. y) is a solution of the equation in question. Generalized separable solution: w(x. . . C4 are arbitrary constants. t) is determined by the ﬁrst-order partial differential equation ∂U ∂U ∂U + f (t)Ψ + λ2 g(t) + 2g(t)ϕ(t)z = σ(t). Then the function −k−2 2 w1 = C1 w(x. y) = ϕ(x)ψ(y). C4 are arbitrary constants and the function ϕ = ϕ(x) is determined by the ordinary differential equation k(k + 2)ϕϕxx = 4f (x). where ϕ = ϕ(t). C4 are arbitrary constants. C6 are arbitrary constants. . Suppose w(x. Additive separable solution: w(x. y) 2 ∂y 2 ∂x Suppose w(x. y) = (C1 x + C2 )y k+2 + (k + 1)(k + 2) 0 (C1 t + C2 ) where C1 . Ψ(u) = Φ(u) du. y. 2◦ . is also a solution of the equation. = f (x)y k . . and χ = χ(t) are arbitrary functions. Equations of the Form ∂ w ∂ w = F (x. = f (x)g(y).7. © 2004 by Chapman & Hall/CRC .2. . λ is an arbitrary constant. t) = U (z. . . . where C1 . Multiplicative separable solution: w(x. k(k + 1) 0 (C1 t + C2 ) x (x − t)f (t) 1 dt + C3 xy + C4 x + C5 y + C6 . “Two-dimensional” solution: w(x. −1 ψψyy = C1 g(y).2. . Then the function w1 = w(x. y) + C1 xy + C2 x + C3 y + C4 . where the functions ϕ = ϕ(x) and ψ = ψ(y) are determined by the ordinary differential equations (C1 is an arbitrary constant) ϕϕxx = C1 f (x). . . 2. y) = C1 x 0 k+2 ∂2w ∂2w (x − t)f (t) dt + C2 x + 1 C1 y 0 (y − τ )g(τ ) dτ + C3 y + C4 . and the function U = U (z. y) = (C1 x + C2 )y k+1 + 3◦ . where C1 . ∂x2 ∂y 2 1◦ . . . . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 449 2◦ . . This equation can be fully integrated—a complete integral is sought in the form U = A(t)z + B(t).1. ψ = ψ(t). C1 y) + C2 xy + C3 x + C4 y + C5 . y) = ϕ(x)y 2 + C1 xy + C2 x + C3 y + C4 . 2◦ . t) + ϕ(t)y 2 + ψ(t)y + χ(t). . . . 7. w(x.2. ∂x2 ∂y 2 1◦ . 1. where C1 . z = x + λy. . is also a solution of the equation. .

. Then the function w1 = C1 w x. ∂2w ∂2w ∂x2 ∂y 2 = f (x)e2λy + g(x)eλy . y) is a solution of this equation. y) = z 0 (z − t) f (t) dt + C1 xy + C2 x + C3 y + C4 . is also a solution of the equation. ∂x2 ∂y 2 w(x. . and the function ϕ = ϕ(x) is determined by the ordinary differential equation λ2 ϕϕxx = 4f (x).450 3. ∂2w ∂2w ∂x2 ∂y 2 = f (x)eλy . . where C1 . C4 are arbitrary constants. where C1 . 4. k(k + 1)ϕχxx = h(x). . k(k + 1)ϕψxx = g(x). y − 2 ln |C1 | + C2 xy + C3 x + C4 y + C5 . Suppose w(x. Generalized separable solution: w(x. . EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂2w ∂2w = f (ax + by). . 6. ψ = ψ(x). Generalized separable solution: w(x. ∂2w ∂2w = f (x)y 2k + g(x)y k + h(x)y k–1 . . . y) = ϕ(x)eλy/2 + C1 xy + C2 x + C3 y + C4 . C1 t + C 2 where C1 . C6 are arbitrary constants. where the functions ϕ = ϕ(x). 2◦ . and χ = χ(x) are determined by the system of ordinary differential equations k(k + 1)ϕϕxx = f (x). © 2004 by Chapman & Hall/CRC . . ∂x2 ∂y 2 1 ab ) Solutions: w(x. y) = ϕ(x)eλy + 1 λ2 x x0 (x − t) g(t) dt + C1 xy + C2 x + C3 y + C4 . λ where C1 . . ϕ(t) where C1 . C4 are arbitrary constants. . . . . y) = (C1 x + C2 )eλy + 1 λ2 x x0 (x − t)f (t) dt + C3 xy + C4 x + C5 y + C6 . . . Generalized separable solution: 1◦ . . and x0 is any number such that the integrand does not have a singularity at x = x0 . C5 are arbitrary constants. C4 are arbitrary constants and the function ϕ = ϕ(x) is determined by the ordinary differential equation λ2 ϕϕxx = f (x). . y) = ϕ(x)y k+1 + ψ(x)y + χ(x). Generalized separable solution: w(x. z = ax + by. 5. 3◦ . . .

. The Martin transformation u(ξ. S(ξ) .2. where C1 . The equation of state is assumed to have the form V = V p. ∂x∂y ∂x2 ∂y 2 where F (x. Then the function e w1 = w(x. p the pressure. y) is a solution of the homogeneous Monge–Amp` re equation. where S = S(ξ) is a prescribed entropy proﬁle. 1◦ . ¯ ¯ ¯ where α and β are arbitrary constants. © 2004 by Chapman & Hall/CRC . 3◦ . y). N. = 0. Ibragimov (1994). and V the speciﬁc volume. and k are arbitrary constants. V. L. C10 are arbitrary constants. 2◦ .2. Yanenko (1983). takes the Monge–Amp` re equation to an equation e of the same form. Rozhdestvenskii and N. the system of equations of the one-dimensional gas dynamics with plane waves is as follows: ∂V ∂u ∂u ∂p + = 0. where the an . The transformation x = a 1 x + b1 y + c 1 . ξ the Lagrangian coordinate. is also a solution of the equation. y = y(1 + αx + βy)−1 . The Monge–Amp` re equation is encountered in differential geometry. 1 + C 1 x + C2 y 1 + C 1 x + C2 y ∂2w 2 – ∂2w ∂2w where C1 . 1◦ . t) = ∂w (x. y) + C1 x + C2 y + C3 . cn . are also solutions of the equation. In Lagrangian coordinates. C5 x + C6 y + C7 ) + C8 x + C9 y + C10 . . ∂x t= ∂w (x. and C3 are arbitrary constants. C2 . ¯ y = a 2 x + b2 y + c 2 . B. Martin (1953). ∂t ∂ξ ∂t ∂ξ where t is time. The transformation ¯ x = x(1 + αx + βy)−1 . ◦ 4 . ∂x∂y ∂x2 ∂y 2 Homogeneous Monge–Amp` re equation. Suppose w(x. S(ξ) . ¯ w = kw + a3 x + b3 y + c3 . t) ` reduces the equations of the one-dimensional gas dynamics to the nonhomogeneous Monge–Amp ere equation 2 ∂2w ∂2w ∂2w − = F (x. y) 2 ∂y 2 ∂x Preliminary remarks. gas dynamics. H.7. y). EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 451 ` 7. − = 0. N. y) is a solution of the Monge–Amp` re equation. ∂y x = ξ. e 1.2. and meteorole ogy. y) = − ∂V p. N. . takes the Monge–Amp` re equation to an equation of the same e form. w = w(1 + αx + βy)−1 . y = p(ξ. y). Khabirov (1990). u the velocity. ∂p 1 ¢0 References: M. ¯ ¯ F = k 2 (a1 b2 − a2 b1 )−2 F . . Monge–Ampere equation ∂2w ∂x∂y 2 2 2 – ∂ w ∂ w = F (x. 1 ¢0 References: S. bn . Suppose w(x. Then the functions e w1 = C1 w(C2 x + C3 y + C4 . w2 = (1 + C1 x + C2 y)w y x . . F = F (1 + αx + βy)4 .

w −x −y = 0. w= . y) = 2 2 2 1 x + C1 2 C2 2 x + C3 y + C 4 x + C5 . − ax ∂x ∂y ∂w ∂w − ay.2. ∂2w ∂x∂y 2 – ∂2w ∂2w ∂x2 ∂y 2 = A. Khabirov (1990). S. y) = (C1 x + C2 y + C3 )ϕ where C1 . 4◦ . 2). . y) = ϕ(C1 x + C2 y) + C3 x + C4 y + C5 . b. y) = w(x. 4C1 C2 C2 y 2 + C3 y + 3 + C4 y + C 5 x + C6 . H. 1◦ . y) = 4 ¢3 (C1 x + C2 y + C3 )k+1 + C7 x + C8 y + C 9 .452 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2◦ . N. where the a. C9 are arbitrary constants and ϕ = ϕ(z) is an arbitrary function. General solution in parametric form: w = tx + ϕ(t)y + ψ(t). and ϕ = ϕ(t) and ψ = ψ(t) are arbitrary functions. 2a 2a 4a where β and λ are the parameters. 4C2 w(x. + ax ∂x ∂y = 0. = 0.2. y) = (C1 x + C2 y + C3 )k + C4 x + C5 y + C6 . Φ1 3◦ . where t is the parameter. Solutions involving arbitrary constants: w(x. V. . w(x. w(x. 2◦ . x + ϕ (t)y + ψ (t) = 0. y) = C1 y 2 + C2 xy + w(x. (C4 x + C5 y + C6 )k C1 (x + a)2 + C2 (x + a)(y + b) + C3 (y + b)2 + C5 x + C6 y + C7 . where the Φn (u. .1: E. First integrals: ∂w ∂w = 0. 5◦ . First integrals for A = a2 > 0: Φ1 Φ2 ∂w ∂w + ay. C4 x + C5 y + C 6 C1 x + C2 y + C 3 + C7 x + C8 y + C 9 . ∂x ∂y ∂w ∂w ∂w . y) = (C1 x + C2 y)ϕ y x + C3 x + C4 y + C 5 . Goursat (1933). Φ2 ∂x ∂x ∂y where Φ1 (u. x= © 2004 by Chapman & Hall/CRC . and the Cn are arbitrary constants. v) are arbitrary functions of two arguments (n = 1. ϕ = ϕ(β) and ψ = ψ(λ) are arbitrary functions. v) and Φ2 (u. . . General solution in parametric form for A = a2 > 0: ψ (λ) − ϕ (β) (β + λ)[ψ (λ) − ϕ (β)] + 2ϕ(β) − 2ψ(λ) β−λ . w(x. Ibragimov (1994). Solutions involving one arbitrary function: w(x. 2. References for equation 7. z) are arbitrary functions of two arguments. y= .

where a. Another ﬁve solutions can be obtained: (a) from the solution of equation 7. Polyanin and V. y) = 7 ¢6 5 ∂2w 2 – ∂2w ∂2w −1 w1 = C1 w(x. Martin (1953). Reference: A. where β is an arbitrary constant. b. . 4C1 2C1 0 x C2 1 C2 y 2 + C3 y + 3 − (x − t)(t + C1 )f (t) dt + C4 y + C5 x + C6 .2. . . = f (x). N. w(x. b. y) = y where ϕ(x) is an arbitrary function. . . Zaitsev (2002). and s are arbitrary constants. where C1 . (d) from the solution of equation 7. y) = C1 y 2 + C2 xy + 4C1 2 C2 A 1 C2 y 2 + C3 y + 3 − (x3 + 3C1 x2 ) + C4 y + C5 x + C6 .18 with α = 0 and f (u) = A. k. ∂x∂y ∂x2 ∂y 2 1◦ .22 with α = 0 and f (u) = A. F. N. y) = x + C1 4C2 12C2 √ 2 A 2 (C1 x − C2 y 2 + C3 )3/2 + C4 x + C5 y + C6 . 2◦ .2. References: M. .24 with α = 0 and f (u) = A. Then the functions 3. Solutions: √ A x(C1 x + C2 y) + ϕ(C1 x + C2 y) + C3 x + C4 y. 4C2 2C2 0 where C1 .2. w(x.2.2. y) = 3C1 C2 where C1 . and ξ and η are the new dependent variables. y) = C2 1 (C 2 − A)x2 + C3 y + C4 x + C5 .2.2. . y). EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 453 3◦ . C2 x C1 y + C 3 ) + C4 x + C5 y + C 6 . © 2004 by Chapman & Hall/CRC . ◦ 3 . D. . (b) from the solution of equation 7. . L.2. leads to an equation of the similar form ∂2u ∂ξ∂η 7 ¢6 2 − ∂ 2u ∂ 2 u 1 = . Yanenko (1983). Generalized separable solutions quadratic in y: w(x. ∂x η= ∂w .7. c. (c) from the solution of equation 7. and c are arbitrary constants. The Legendre transformation u = xξ + yη − w(x. y) is a solution of this equation. C6 are arbitrary constants and ϕ = ϕ(z) is an arbitrary function. C6 are arbitrary constants.20 with f (u) = A. Rozhdestvenskii and N.21 with f (u) = A. η) is the new independent variable. 1 x + C1 x 2 1 C2 2 x − (x − t)f (t) dt + C3 y + C4 x + C5 . ∂ξ 2 ∂η 2 A Reference: E. . w(x. . where a.2. C6 are arbitrary constants. Goursat (1933). y) = C1 y 2 + C2 xy + w(x. (e) from the solution of equation 7. ξ= ∂w . Suppose w(x.2. w(x. ∂y 5 5 where u = u(ξ.2. . where β is an arbitrary constant. 4◦ . Generalized separable solutions for f (x) > 0: w(x. B. are also solutions of the equation. 7 ¢6 5 f (x) dx + ϕ(x) + C1 y. where β is an arbitrary constant.

C5 are arbitrary constants. V. C1 y) + C2 x + C3 y + C4 . . C1 x + C2 2C1 [ϕ(x)]3 2C1 [ϕ(x)]2 [ψ (t)]2 1 x (x − t) t dt + C5 x + C6 . See solution of equation 7. w(x. w(x. 9 ¢8 Reference: S. y) is a solution of this equation. . w(x. . and ϕ(x) and ψ(y) are arbitrary where Dx = ∂x ∂y functions. ◦ 5 . Polyanin and V. y) = ϕ(x)y 2 + ψ(x)y + χ(x).2. . F. w(x. 9 ¢8 where Reference: A. 5. y) = (C1 x + C2 )2 6 a where C1 . . 2◦ . C5 are arbitrary constants. Generalized separable solutions cubic in y: x 1 (x − t)f (t) dt + C2 x + C3 y + C4 . g = y f (x) dx + ϕ(x) + ψ(y).2. .24 with f (u) = A and α = k/2. Generalized separable solution quadratic in y: w(x. 3◦ . Solutions for f (x) = Aeλx : √ 2 A λx/2 e sin(C1 x + C2 y + C3 ) + C4 x + C5 y + C6 . Solutions that can be obtained by the formulas of Items 1◦ and 2◦ are omitted. C4 are arbitrary constants. Suppose w(x. = f (x)y. . Conservation law: Dx y(wx wyy − wy wxy + gy ) − g − wx wy + Dy y(wy wxx − wx wxy + gx ) + (wx )2 = 0. χ(x) = 2 a ϕ(t) 4◦ . ϕ(x) = 5◦ . A ∂2w 2 – ∂2w ∂2w where C1 . . Generalized separable solution quadratic in y: 1 w(x. y) = C2 λ √ 2 A λx/2 e sinh(C1 x + C2 y + C3 ) + C4 x + C5 y + C6 . ψ(x) = C3 ϕ(x) + C4 + − . Khabirov (1990). .2.2. ∂ ∂ .2. 5.2. Solutions with f (x) = Axk can be obtained: (a) from the solution of equation 7. F (x) = 2C1 where C1 . . © 2004 by Chapman & Hall/CRC @ @ @ x a (x − t)F 2 (t) dt + C2 x + C3 y + C4 . y) = C1 y 3 − 6C1 a 1 x y3 − (x − t)(C1 t + C2 )2 f (t) dt + C3 x + C4 y + C5 .454 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 4◦ . where β is an arbitrary constant. (b) from the solution of equation 7. are also solutions of the equation. where β is an arbitrary constant. Zaitsev (2002). Then the functions −3 2 w1 = C1 w(x.18 with f (u) = A and α = k/2. . y) = C2 λ √ 2 −A λx/2 e cosh(C1 x + C2 y + C3 ) + C4 x + C5 y + C6 . Dy = . where β is an arbitrary constant. w(x.7 in Item 3◦ with k = 1. ϕ(x) f (x) dx 1 f (x) dx 1 .22 with α = λ and f (u) = A. y) = C2 λ Another solution can be obtained from the solution of equation 7.2. y) = C1 y 2 − y F (x) dx + 2C1 1 f (x) dx + C5 .1.2. . D. 4. ∂x∂y ∂x2 ∂y 2 1◦ . Let us consider some speciﬁc functions f = f (x).2. .

C1 y) + C2 x + C3 y + C4 . Generalized separable solutions in the form of polynomials of degree 4 in y: w(x.2. . = f (x)y 2 + g(x)y + h(x). is also a solution of the equation. B ∂2w 2 – ∂2w ∂2w where C1 . . – w(x. D ¢C Reference: A. F. D. y) is a solution of this equation. 2 ϕψxx = 2ϕx ψx − 1 g(x). C4 are arbitrary constants. C5 are arbitrary constants. . Generalized separable solutions: w(x. See solution of equation 7. 1 y4 − (C1 x + C2 )3 12 where C1 . Then the functions −k−2 2 w1 = C1 w(x.2.7 in Item 3◦ with k = 2. ∂x∂y ∂x2 ∂y 2 1◦ . . . (x − t)(C1 t + C2 )3 f (t) dt + C3 x + C4 y + C5 . . . . y) = C1 y 4 − w(x. Then the function 5. y) = 1 12C1 x a x a (x − t)f (t) dt + C2 x + C3 y + C4 . y) = x 1 C1 y k+2 − (x − t)f (t) dt + C2 x + C3 y + C4 . . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 455 = f (x)y 2 . ∂x∂y ∂x2 ∂y 2 1◦ . Suppose w(x. Zaitsev (2002). where the function ϕ = ϕ(x) is determined by the ordinary differential equation ϕϕxx = 2(ϕx )2 − 1 f (x). . ∂x∂y ∂x2 ∂y 2 Generalized separable solution quadratic in y: 6. where the functions ϕ = ϕ(x). 2 3◦ . (k + 1)(k + 2) C1 a x 1 y k+2 − (x − t)(C1 t + C2 )k+1 f (t) dt + C3 x + C4 y + C5 . Suppose w(x. 2 ∂2w 2 ∂2w ∂2w 7. w(x. . and χ = χ(x) are determined by the system of ordinary differential equations ϕϕxx = 2(ϕx )2 − 1 f (x). 2◦ . . Polyanin and V.2.7. are also solutions of the equation. C5 are arbitrary constants. © 2004 by Chapman & Hall/CRC . C4 are arbitrary constants. 4◦ . . 2◦ . 2 1 ϕχxx = 1 (ψx )2 − 2 h(x). – −2 w1 = C1 w(x. . = f (x)y k . y) = k+1 (C1 x + C2 ) (k + 1)(k + 2) a where C1 . y) is a solution of this equation. Generalized separable solution quadratic in y: w(x. y) = ϕ(x)y 2 + C1 ϕ2 (x) dx + C2 y + 1 2 C 2 1 x a (x − t)ϕ3 (t) dt + C3 x + C4 . ψ = ψ(x). . B ∂2w 2 ∂2w ∂2w where C1 . y) = ϕ(x)y 2 + ψ(x)y + χ(x). . C1 y) + C2 x + C3 y + C4 .

Let us consider the case where f is a power-law function of x. w(x. w(x. ∂x∂y ∂x2 ∂y 2 Generalized separable solution: 8. f (x) = Ax n . Another exact solution for f (x) = Axn can be obtained from the solution of equation 7. (n + 1)(n + 2) C1 (k + 1)(k + 2) Ay k+n+3 C1 xn+2 − + C2 y + C 3 x + C4 . Polyanin and V. n2 (n + 2) whose exact solutions for various values of k and n can be found in the books by Polyanin and Zaitsev (1995. y) = ψ(y)x n+2 2 . 4◦ . y) = k+1 (k + 1)(k + 2)x C1 (k + n + 2)(k + n + 3) x A (x − t)tn (C1 t + C2 )k+1 dt + C3 y + C4 x. F. y) = Ay k+2 C1 xn+2 − + C2 y + C 3 x + C4 . The substitution ψ = U −n/2 brings it to the Emden–Fowler equation Uyy = 8A y k U n+1 . = f (x)y 2k+2 + g(x)y k . y) = ϕ(x)y k+2 − 1 (k + 1)(k + 2) x a ∂2w 2 ∂2w ∂2w (x − t) g(t) dt + C1 x + C2 y + C3 . Zaitsev (2002). 2003). C4 are arbitrary constants. Polyanin and V. Zaitsev (2002). y) = (C1 y + C2 )−n−1 xn+2 − (n + 1)(n + 2) a where C1 . . w(x. y) = ϕ(x)y k+2 2 . Multiplicative separable solution: w(x. w(x.2. F. where α and β are arbitrary constants. D. y) = (C1 x + C2 )−k−1 y k+2 − (k + 1)(k + 2) a y A (y − t)tk (C1 t + C2 )n+1 dt + C3 y + C4 x. where the function ψ = ψ(y) is determined by the ordinary differential equation n(n + 2)ψψyy − (n + 2)2 (ψy )2 + 4Ay k = 0. . © 2004 by Chapman & Hall/CRC . and a solution of the same type: w(x. in more detail. Solutions: w(x. . where the function ϕ = ϕ(x) is determined by the ordinary differential equation k(k + 2)ϕϕxx − (k + 2)2 (ϕx )2 + 4f (x) = 0. D. .456 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 3◦ . see Item 3◦ with f (x) = Axn .18 with f (u) = Auk and n = 2α + kβ. F ¢E Reference: A. There are also a multiplicative separable solution. – w(x.2. y) = n+1 (n + 1)(n + 2)y C1 (k + n + 2)(k + n + 3) Axk+n+3 C1 y k+2 − + C2 y + C 3 x + C4 . F ¢E Reference: A. ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation (k + 1)(k + 2)ϕϕxx − (k + 2)2 (ϕx )2 + f (x) = 0.

. Suppose w(x. C4 and β are arbitrary constants. . 9. ∂2w ∂x∂y 1◦ . y) = z = ax + by ∂2U ∂2U ∂2U = + b−2 f (z). ∂x2 ∂y 2 x f (z) dz + ϕ(z) + C1 x + C2 y. . . . ∂x∂z ∂x2 ∂z 2 Here. 2 λ a ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation w(x. Zaitsev (2002). ∂x∂y ∂x2 ∂y 2 1◦ . I ¢H 1 2 λy . Zaitsev (2002). 11. C4 are arbitrary constants. y) = C1 a x G ∂2w 2 – ∂2w ∂2w (x − t)f (t) dt + C2 x − − 1 λy e + C3 y + C 4 . . – ∂2w 2 ∂2w ∂2w . are also solutions of the equation. Generalized separable solutions: w(x. F. Solutions: 2 – ∂2w ∂2w = f (ax + by).2. y) = ϕ(x) exp where the function ϕ = ϕ(x) is determined by the ordinary differential equation ϕϕxx − (ϕx )2 + 4λ−2 f (x) = 0. ∂x∂y ∂x2 ∂y 2 Generalized separable solution: 10. The transformation w = U (x. leads to an equation of the form 7. . w(x. . y − ln |C1 | + C2 x + C3 y + C4 . D. 2◦ . Reference: A. y) is a solution of this equation. Polyanin and V. C 1 λ2 x 1 (x − t)e−βt f (t) dt + C2 x + C3 y + C4 . . Reference: A.7. D. . y) = C1 e βx+λy 3◦ .2. x and z play the role of y and x in 7. respectively.3: 2 G w(x. z). C4 are arbitrary constants. x g(t) 1 (x − t) dt + C1 x + C2 y + C3 . a and b are any numbers. y) = ϕ(x)eλy − ϕϕxx − (ϕx )2 + λ−2 f (x) = 0. Polyanin and V. y) = C1 (x − t)f (t) dt − C1 b a where C1 . z = ax + by. Then the functions 2 w1 = C1 w x. w(x. Multiplicative separable solution: w(x. 2 C1 λ a where C1 .2. 12.2.2. . ∂x∂y ∂x2 ∂y 2 Generalized separable solutions: y x 1 (y − ξ)g(ξ) dξ Axy + C2 x + C3 y + C4 . © 2004 by Chapman & Hall/CRC G I ¢H = f (x)g(y) + A2 . b where C1 and C2 are arbitrary constants and ϕ(z) is an arbitrary function. . 2◦ . F. ∂2w 2 – ∂2w ∂2w = f (x)e2λy + g(x)eλy . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 457 = f (x)eλy .3. λ where C1 .

16. Suppose w(x. respectively. 2 ∂y 2 ∂x∂y ∂x x x This is a special case of equation 7. © 2004 by Chapman & Hall/CRC P 1◦ .18 with α = −2 and β = −1.2. . ∂x2 ∂z 2 Here. w = U (x.2.2.2.2.2. Then the functions . w = U (x. y) is a solution of the equation in question.2. ∂x2 ∂z 2 Here.7: = ∂2U ∂2U + b−2 xk f (z).9. z). ∂2U ∂x∂z 2 z = ax + by leads to an equation of the form 7.2. ∂2w 2 ∂x∂y ∂x2 ∂y 2 The transformation – ∂2w ∂2w = x2k+2 f (ax + by) + xk g(ax + by). C1 y) + C2 x + C3 y + C4 . 14.2. z).9: = ∂2U ∂2U + b−2 eλx f (z). x and z play the role of y and x in 7. w = U (x. 1 y ∂2w 2 ∂2w ∂2w – = 4f .2. 15. ∂2w 2 ∂x∂y ∂x2 ∂y 2 The transformation – ∂2w ∂2w = eλx f (ax + by). respectively.2. ∂x2 ∂z 2 Here.2. are also solutions of the equation. 2 z = ax + by leads to an equation of the form 7.2.458 13. ∂2w 2 ∂x∂y ∂x2 ∂y 2 The transformation – ∂2w ∂2w = e2λx f (ax + by) + eλx g(ax + by). x and z play the role of y and x in 7.8: ∂2U ∂x∂z = ∂ 2U ∂ 2 U + b−2 xk+2 f (z) + b−2 xk g(z). w = U (x.8. where C1 . . z). w1 = w(C1 x.2. z).2. .7. respectively. ∂2U ∂x∂z 2 z = ax + by leads to an equation of the form 7. respectively.10. x and z play the role of y and x in 7. .10: ∂2U ∂x∂z = ∂2U ∂2U + b−2 e2λx f (z) + b−2 eλx g(z).2.2. C4 are arbitrary constants. EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂2w 2 ∂x∂y ∂x2 ∂y 2 The transformation – ∂2w ∂2w = xk f (ax + by).2. 2 z = ax + by leads to an equation of the form 7. ∂x2 ∂z 2 Here. x and z play the role of y and x in 7. 17.

4◦ . Zaitsev (2002). Yanenko (1983). . Dy = . D. are also solutions of the equation. and F (z) = ∂x ∂y f (z) dz. . S ¢R S ¢R References: M. z= y . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 459 2◦ . . ∂2w ∂x∂y 2 – ∂2w ∂2w ∂x2 ∂y 2 = f (ax – by 2 ). ∂ ∂ . Then the functions w1 = w(x + 2bC1 y + abC1 . S ¢R Reference: A. z= y . are also solutions of the equation. N. F (z) = 1 a2 b f (z) dz. where C1 . C4 are arbitrary constants. where the function u = u(z) is determined by the ordinary differential equation [β(β + 1)zuz + (α − β)(β − α − 1)u]uzz + (α + 1)2 (uz )2 − f (z) = 0. . y) is a solution of this equation. 1◦ . y + aC1 ) + C2 x + C3 y + C4 . . 19. 3◦ . Rozhdestvenskii and N. where C1 . 2◦ . Martin (1953). . C4 are arbitrary constants. y) = xα−β+1 u(z) z = xβ y. V. B. Solutions: Q w(x.2. N. . y) = F (z) + C1 dz + C2 x + C3 y + C4 .7. Khabirov (1990). x + Dy − ∂w ∂ 2 w + x−3 F ∂x ∂x∂y y x = 0. . z = ax − by 2 . Suppose w(x. . Conservation law: Dx where Dx = ∂w ∂ 2 w ∂x ∂y 2 y x f (z) dz + C. . Self-similar solution: w(x. ∂2w ∂x∂y 2 – ∂2w ∂2w ∂x2 ∂y 2 = x2α f (xβ y). Then the functions β−α−1 −β w(C1 x. Integral: Q w−x ∂w ∂w −y ∂x ∂y f (z) dz = C. Khabirov (1990). . L. 18. C1 y) + C2 x + C3 y + C4 . x where C is an arbitrary constant. Solutions: Q w = xϕ where ϕ(z) is an arbitrary function. 1◦ . Polyanin and V. C4 are arbitrary constants. w1 = C1 Q where C1 . Suppose w(x. Reference: S. S ¢R Reference: S. © 2004 by Chapman & Hall/CRC . 2◦ . . F. y) is a solution of this equation. V.

The substitution V (z) = (uz )2 leads to a ﬁrst-order linear equation. C4 are arbitrary constants. D.460 20. z = eβx y. where the function u = u(z) is determined by the ordinary differential equation 2 (4ac − b2 )z + as 2 + ck 2 − bks uz uzz + (4ac − b2 )(uz )2 + f (z) = 0. 22. z = ax2 + bxy + cy 2 + kx + sy. ∂x2 ∂y 2 α−2β 2β w(x − 2 ln C1 . z F (z) = b2 1 − 4ac f (z) dz + C2 . y) = u(z). F. where the function u = u(z) is determined by the ordinary differential equation Integrating yields T T u(z) = F (z) dz + C1 . ∂2w ∂x∂y 2 – ∂2w ∂2w = eαx f (eβx y). . 2x Solution: where the function ϕ = ϕ(x) is determined by the ordinary differential equation x2 ϕϕxx − x2 (ϕx )2 + 2xϕϕx − ϕ2 + 4k −2 x4 f (x) = 0. Suppose w(x. Zaitsev (2002). 2◦ . are also solutions of the equation. where C1 and C2 are arbitrary constants. ∂2w ∂x∂y 2 – ∂2w ∂2w = eky/x f (x). F. ∂x2 ∂y 2 w(x. © 2004 by Chapman & Hall/CRC W 21. Zaitsev (2002). Solution for b2 ≠ 4ac: w(x. 2(4ac − b2 )zuz uzz + (4ac − b2 )(uz )2 + f (z) = 0. C1 y) + C2 x + C3 y + C4 . y) = u(z) z = ax2 + bxy + cy 2 . . Polyanin and V. ∂2w 2 – ∂2w ∂2w ∂x2 ∂y 2 = f (ax2 + bxy + cy 2 + kx + y). V ¢U Reference: A. 2 where the function U = U (z) is determined by the ordinary differential equation β 2 zUz Uzz − µ2 U Uzz + (β + µ)2 (Uz )2 − f (z) = 0. y) = exp ky ϕ(x). y) = eµx U (z). . EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS ∂2w ∂x∂y 2 – ∂2w ∂2w ∂x2 ∂y 2 = f (ax2 + bxy + cy 2 ). Generalized self-similar solution: w(x. ∂x∂y Solution: w(x. V ¢U Reference: A. 23. Then the functions where C1 . µ = 1 α − β. Polyanin and V. . . y) is a solution of this equation. w1 = C1 1◦ . D.

2. and C3 are arbitrary constants. ∂2w 2 ∂ 2 w 2 – ∂ 2 w ∂ 2 w = F x. z = xβ ey/x . Y ¢X Reference: S. is also a solution of the equation. – w1 = C1 w(x. ∂x∂y ∂x2 ∂y 2 1◦ . Then the function 2. Suppose w(x. its general solution can be expressed via the particular solution of the ﬁrst equation). V. w(x. 2ϕψxx + f (x)ψ − 4ϕx ψx = 0. where the function u = u(z) is determined by the ordinary differential equation z 2 βzuz + (α + 2)(α + 1)u uzz + z Y ¢X β − (α + 1)2 zuz + (α + 2)(α + 1)u uz + f (z) = 0. C2 . For exact solutions of the nonhomogeneous Monge–Amp` re equation for some speciﬁc F = F (x. and the function ϕ = ϕ(z) is determined by the ordinary differential equation 2 (2βϕz + α2 ϕ)ϕzz − α2 ϕz + f (z) = 0. ∂x∂y ∂x2 ∂y 2 1◦ . = f (x)w2 .2. and χ(x) are determined by the system of ordinary differential equations 2ϕϕxx + f (x)ϕ − 4(ϕx )2 = 0. EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 461 ∂x∂y Solution: 24. w = y exp(αy −1 )ϕ(z) + C1 y + C2 x + C3 . C2 .3. Generalized separable solution quadratic in y: w = ϕ(x)y 2 + ψ(x)y + χ(x). ∂2w 2 – ∂2w ∂2w ∂x2 ∂y 2 = y –4 exp(2αy –1 )f (xy –1 + βy –2 ). ∂w ∂x∂y ∂x2 ∂y 2 ∂x ∂y = f (x)w + g(x). where ϕ(x). ∂w . y) e (without functional arbitrariness). and C3 are arbitrary constants. – ∂2w ∂2w 2◦ . Khabirov (1990). Reference: S. w. e 7. ∂x∂y Solution: 25. where C1 and C2 are arbitrary constants. © 2004 by Chapman & Hall/CRC ∂2w 2 ∂2w ∂2w . where C1 . Then the function w1 = w(x. Equations of the Form 1. ψ(x). 2ϕχxx + f (x)χ + g(x) − (ψx )2 = 0. where C1 . Khabirov (1990). ∂2w 2 – ∂2w ∂2w ∂x2 ∂y 2 = x2α f (xβ ey/x ). y) is a solution of this equation. The Cauchy problem for the Monge–Amp` re equation is discussed in Courant and Hilbert (1989). is also a solution of the equation. see Khabirov (1990) and Ibragimov (1994). Note that the second equation is linear in ψ and has a particular solution ψ = ϕ (hence.7. Suppose w(x. y) = xα+2 u(z). y. V. z = xy −1 + βy −2 . y) is a solution of this equation. y + C2 x + C3 ). y + C1 x + C2 ).

λ n+2 ∂2w 2 – ∂2w ∂2w where C is an arbitrary constant. = f (x)y n wk . and s are arbitrary constants and the function u = u(z) is determined by the ordinary differential equation 2 (4ac − b2 )z + as 2 + ck 2 − bks uz uzz + (4ac − b2 )(uz )2 + f (u) = 0. y) is a solution of this equation. ∂2w 2 ∂2w ∂2w – = f (w). y) is a solution of this equation. 2−k where the function U (x) is determined by the ordinary differential equation U Uxx − (Ux )2 + (k − 2)2 λ−2 f (x)U k = 0. Suppose w(x. w1 = C n+2 w(x. and any three of the four constants A1 . where C is an arbitrary constant. b. 4. Multiplicative separable solution: w(x. 2◦ . ∂2w 2 ∂2w ∂2w – = f (x)eλy wk . c. y) = y 2−k U (x). y + k−2 ln C . B1 . y) is a solution of this equation. z = ax2 + bxy + cy 2 + kx + sy. is also a solution of the equation. 2◦ . y) = exp λy U (x). where λ is an arbitrary constant and the function u = u(x) is determined by the ordinary differential equation uuxx − (ux )2 + λ−2 f (x)u2 = 0.462 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2◦ . where the function U (x) is determined by the ordinary differential equation (n + 2)(n + k)U Uxx − (n + 2)2 (Ux )2 + (k − 2)2 f (x)U k = 0. Then the function 3. ∂x∂y ∂x2 ∂y 2 1◦ . Multiplicative separable solution with k ≠ 2 and λ ≠ 0: w(x. C2 . A2 . where C1 . and B2 are arbitrary. Multiplicative separable solution with n ≠ −2 and k ≠ 2: w(x. Then the function w1 = Cw x. y) = u(z). 5. Suppose w(x. C k−2 y). |A2 B1 − A1 B2 | = 1. 2◦ . is also a solution of the equation. ∂x∂y ∂x2 ∂y 2 1◦ . where a. Suppose w(x. © 2004 by Chapman & Hall/CRC . y) = eλy u(x). Functional separable solution: w(x. is also a solution of the equation. k. Then the function w1 = w(A1 x + B1 y + C1 . ∂x∂y ∂x2 ∂y 2 1◦ . A2 x + B2 y + C2 ).

EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 463 a λy u(x). w(x. ∂x∂y Solution: 6. √ 1 w = 4 ( a x + C)2 y + ϕ(x). C7 are arbitrary constants. where ϕ(x) is an arbitrary function and C is an arbitrary constant. ∂x∂y ∂x2 ∂y 2 ∂y This equation is used in meteorology for describing wind ﬁelds in near-equatorial regions. . . is also a solution of the equation. a ¢` −1 1 w = 4 aC2 (x + C1 )2 coth(C2 y + C3 ). 2 0 F (t) and C1 . . C2 y + C4 x + C5 ) + C6 x + C7 . 2 C1 x + C2 6C1 [G (t)]2 − aG(t) 1 x (x − t) t dt + C5 x + C6 . There are exact solutions of the following forms: w(x. . λ = . 2 w = 1 a(x + C1 )2 (y + C2 ). w(x. 4 −1 1 w = 4 aC2 (x + C1 )2 tanh(C2 y + C3 ). and λ are arbitrary constants. 2C3 ξ = ye−kx . ∂2w 2 = ∂2w ∂2w ∂x2 ∂y 2 + f (x) exp ay x wk . η = y + k ln |x|. . R. Solutions: w = C1 eλy − 1 aλ−1 x2 + C2 x + C3 .2. −2 −1 w1 = C1 C2 w(C1 x + C3 . Suppose w(x. Solutions: −1 1 w = 4 aC2 (x + C1 )2 tan(C2 y + C3 ). Generalized separable solution quadratic in y: w = F (x)y 2 + G(x)y + H(x). C1 x + C2 G(x) = − C3 a (C1 x + C2 )2 + + C4 . where k is an arbitrary constant. w = ϕ(x).7. . 5◦ . Rozendorn (1984). © 2004 by Chapman & Hall/CRC a 2 x + C4 x + C5 . C3 . x 2−k where the function u = u(x) is determined by the ordinary differential equation w(x. Generalized separable solution: w = C1 exp(C2 x + C3 y) − 6◦ . 7. . where ϕ(x) is an arbitrary function and C1 . y) = x2 W (η). The ﬁrst solution is a solution in additive separable form and the other four are multiplicative separable solutions. . Reference: E. y) is a solution of the equation in question. z = y|x|−k . where F (x) = H(x) = 1 . C2 . y) = |x|k+2 U (z). 3◦ . 4◦ . y) = exp x2 uuxx − (xux − u)2 + λ−2 x4 f (x)uk = 0. C6 are arbitrary constants. ∂w ∂2w 2 ∂2w ∂2w = +a . 1◦ . Then the function where C1 . y) = ekx V (ξ). . 2◦ .

. Solutions: w = ϕ(x). the last solution is an additive separable solution. y) = ϕ(y)x + ψ(y). Suppose w(x. y) is a solution of this equation. For C 2 = 0. . . Its general solution can be written out in parametric form as x = −2C1 dt + C3 . . 9. where the function z(x) is determined by the autonomous ordinary differential equation 2C 1 zxx + f (zx ) = 0. ∂x∂y ∂y 1◦ . . b ∂2w 2 ∂2w ∂2w where C1 . ∂w +f . C1 y + C2 x + C3 ) + C4 x + C5 . w= 1 4 f (x) dx + C 2 y + ϕ(x). . Then the function = ∂x2 ∂y 2 + f (x) −1 w1 = C1 w(x. C5 are arbitrary constants. C6 are arbitrary constants. y) = C1 y 2 + C2 y + C3 + z(x). is also a solution of the equation. 1 C3 x 0 w = C1 exp(C2 x + C3 y) − (x − t)f (t) dt + C4 x + C5 . Generalized separable solution linear in x: w(x. ∂2w ∂2w ∂w where C1 . Generalized separable solution quadratic in y: w = ϕ(x)y 2 + ψ(x)y + χ(x). 3◦ . C6 are arbitrary constants. .464 8. are also solutions of the equation. f (t) z = −2C1 t dt + C4 . 3◦ . where ϕ(x) is an arbitrary function and C is an arbitrary constant. f (t) b © 2004 by Chapman & Hall/CRC . Then the functions = −1 w1 = C1 w(C1 x + C2 y + C3 . . . . dx dx + 2 (x) ϕ C1 x + C2 (x − t) [ψt (t)]2 − f (t)ψ(t) dt + C5 x + C6 . 2◦ . ϕ(t) and C1 . C1 x + C2 1 2 x 0 ψ(x) = − ϕ2 (x) f (x) C3 + C4 . Additive separable solution: w(x. y + C4 ) + C5 y + C6 . where ψ(y) is an arbitrary function and the function ϕ(y) is deﬁned implicitly by dϕ √ = y + C. . . ∂2w EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2 . 2◦ . y) is a solution of this equation. ∂x∂y ∂x2 ∂y 2 ∂x 1◦ . where ϕ(x) = χ(x) = 1 . Suppose w(x. f (ϕ) where C is an arbitrary constant.

. 2 ∂η 2 ∂ξ∂η ∂ξ F (ξ. Generalized separable solution involving an exponential of y: w(x. y). y) = ϕ(x) + C1 y + C2 . where C1 . y) = ϕ(y) + C1 x + C2 . and λ are arbitrary constants and the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕϕxx − (ϕx )2 = 0.3: u = xξ + yη − w(x. = f (x) ∂2w ∂2w where C1 . leads to an equation of the form 7. 3◦ . where C1 and C2 are arbitrary constants and ϕ = ϕ(z) is an arbitrary function. y) = ϕ(x)y k + C1 x + C2 y + C3 where the function ϕ = ϕ(x) is determined by the ordinary differential equation (k − 1)f (x)ϕϕxx − k(ϕx )2 = 0. Degenerate solutions involving arbitrary functions: w(x. The Legendre transformation ∂w ∂w . w(x. ∂2w 2 2 − ∂2u ∂2u 1 = . ∂2w 2 ∂2w ∂x∂y 2 2 2 = f (x. y) is a solution of this equation. y) = ϕ(x)eλy + C1 x + C2 y + C3 .4. ∂x ∂y ξ= ∂w ∂w . 2 ∂η 2 ∂ξ f (ξ) ∂x∂y The Legendre transformation = ∂2w ∂2w ∂x2 ∂y 2 +F ∂w ∂w . Suppose w(x. 2◦ . . y). η) is the new dependent variable and ξ and η are the new independent variables.2. ∂x ∂y where u = u(ξ. C2 y + C3 ) + C4 x + C5 y + C6 . is also a solution of the equation. 4◦ . C6 are arbitrary constants. © 2004 by Chapman & Hall/CRC . ∂x ∂y where u = u(ξ. u = xξ + yη − w(x. . . C3 .2. Equations of the Form 1.2. C2 . . see Subsection 7. 7. leads to the simpler equation 2 ∂2u ∂2u 1 ∂2u − = .7. ∂x∂y ∂x2 ∂y 2 1◦ .2. EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 465 4◦ . Then the function w1 = C1 w(x. w(x. Generalized separable solution quadratic in y: [ϕ (t)]2 C2 x (x − t) t dt + C3 x + C4 . y) ∂ w ∂ w + g(x. y) 2 ∂y 2 ∂x . η) is the new dependent variable and ξ and η are the new independent variables.2. ξ= ∂2u ∂ξ∂η 10. η= . η= . Generalized separable solution involving an arbitrary power of y: w(x.2. η) For exact solutions of this equation. y) = ϕ(x)y 2 + [C1 ϕ(x) + C2 ]y + 1 2 0 f (t)ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕϕxx − 2(ϕx )2 = 0. 5◦ .

Then the functions = f (x) −2 w1 = C1 w(x. 4◦ . see equation 7.466 2. ∂x∂y ∂x2 ∂y 2 1◦ . . . ∂2w ∂2w ∂2w 2 = f (x) + g(x)y. ∂x∂y ∂x2 ∂y 2 1◦ . 2◦ . f (t)ϕ(t) c g(x) dx + ϕ(x) + C1 y. . f (t) (x − t) g(t) dt + C2 x + C3 . are also solutions of the equation. y) is a solution of this equation. . © 2004 by Chapman & Hall/CRC . . For an exact solution quadratic in y. C4 are arbitrary constants. . C1 y + C2 ) + C3 x + C4 y + C5 . C1 y) + C2 x + C3 y + C4 . 3◦ . Then the function −3 2 w1 = C1 w(x. . C4 are arbitrary constants. 3◦ . Suppose w(x. is also a solution of the equation. where C1 . See the solution of equation 7. 2◦ .6 in Item 3◦ with k = 1. A more general solution is given by w(x. Generalized separable solution quadratic in y: w(x. C5 are arbitrary constants. . which has a particular solution ϕ = C6 . Suppose w(x. . are also solutions of the equation. ∂2w EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2 + g(x). f (t)ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation 2f (x)ϕϕxx − 3(ϕx )2 = 0.2. 4. . Suppose w(x. ∂x∂y ∂x2 ∂y 2 1◦ . Generalized separable solution cubic in y: w(x.5 with g 2 = g0 = 0. y) is a solution of this equation. C1 y) + C2 x + C3 y + C4 . c ∂2w 2 ∂2w ∂2w where C1 . + g(x)y 2 . Then the functions −1 w1 = C1 w(x. . y) = ϕ(x)y 3 + C1 y − 1 6 x a 1 6C1 x a (x − t) g(t) dt + C3 x + C4 . C4 are arbitrary constants. . c = f (x) ∂2w ∂2w where C1 . y) = C1 y 3 + C2 y − where C1 . C4 are arbitrary constants and the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕϕxx − 2(ϕx )2 = 0. .4. y) = y where ϕ(x) is an arbitrary function. . y) = ϕ(x)y 2 + [C1 ϕ(x) + C2 ]y + 1 2 x 0 2 C1 [ϕt (t)]2 − g(t) dt + C3 x + C4 . (x − t) where C1 . . 3. Generalized separable solution linear in y: w(x. . . . .2.4. y) is a solution of this equation. .

f (t) where C1 . y) = ψ(x)y k+2 − 1 (k + 1)(k + 2) x a (x − t) g(t) dt + C1 x + C2 y + C3 . f (t)ψ(t) where the function ψ = ψ(x) is determined by the ordinary differential equation (k + 1)f (x)ψψxx − (k + 2)(ψx )2 = 0. C4 are arbitrary constants. . + g2 (x)y 2 + g1 (x)y + g0 (x). . 3◦ . . . y) = C1 y 4 + C2 y − where C1 . Multiplicative separable solution: w(x. see equation 7. . . 4◦ . 3◦ . f (t)ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation 3f (x)ϕϕxx − 4(ϕx )2 = 0.4. C4 are arbitrary constants. + g(x)y k . 2 ∂2w 2 ∂2w ∂2w 6. .2. ∂x∂y ∂x2 ∂y 2 1◦ . 2 1 f (x)ϕχxx = 1 (ψx )2 − 2 g0 (x). Additive separable solution: w(x. .6 in Item 3◦ with k = 2. 2◦ . © 2004 by Chapman & Hall/CRC . C1 y) + C2 x + C3 y + C4 . and χ = χ(x) are determined by the system of ordinary differential equations 1 f (x)ϕϕxx = 2(ϕx )2 − 2 g2 (x). Generalized separable solution: w(x. ψ = ψ(x). . .2. EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 467 2◦ . are also solutions of the equation. For an exact solution quadratic in y. See the solution of equation 7. . y) = ϕ(x)y 2 + ψ(x)y + χ(x). f (t) (x − t) g(t) dt + C2 x + C3 . C4 are arbitrary constants. ∂x∂y ∂x2 ∂y 2 Generalized separable solution quadratic in y: = f (x) w(x. 5. y) = 1 C1 y k+2 + C2 y − (k + 1)(k + 2) C1 x a (x − t) g(t) dt + C3 x + C4 .7. . where the functions ϕ = ϕ(x). d ∂2w 2 = f (x) ∂2w ∂2w where C1 . y) is a solution of this equation. where the function ϕ = ϕ(x) is determined by the ordinary differential equation k(k + 2)f (x)ϕϕxx − (k + 2)2 (ϕx )2 + 4g(x) = 0. y) = ϕ(x)y 4 + C1 y − 1 12 x a 1 12C1 x a (x − t) g(t) dt + C3 x + C4 .2.5 with g 1 = g0 = 0. A more general solution is given by w(x. Then the functions −k−2 2 w1 = C1 w(x.4. y) = ϕ(x)y k+2 2 . f (x)ϕψxx = 2ϕx ψx − 1 g1 (x). 4◦ . Generalized separable solution involving y to the power of four: w(x. Suppose w(x.

. . © 2004 by Chapman & Hall/CRC . y) = ϕ(x)eλy − x a (x − t) h(t) dt + C1 x + C2 y + C3 . y) = ϕ(x) exp 1 2 λy 1 C 1 λ2 x a (x − t) g(t) dt + C3 x + C4 . Generalized separable solution: w(x. Then the functions w1 = C1 w x. are also solutions of the equation. Suppose w(x. λ 1◦ . Additive separable solution: w(x. ∂2w ∂x∂y 2 = f (x) ∂2w ∂2w + g(x)eλy . f (t)ψ(t) where the function ψ = ψ(x) is determined by the ordinary differential equation f (x)ψψxx − (ψx )2 = 0. C4 are arbitrary constants. 8. . y) is a solution of this equation. ∂x2 ∂y 2 2 ln |C1 | + C2 x + C3 y + C4 . where the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕϕxx − (ϕx )2 + 4λ−2 g(x) = 0. . . . f (t) . 3◦ . ∂2w ∂x∂y EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2 = f (x) ∂2w ∂2w ∂x2 ∂y 2 + g(x)y 2k+2 + h(x)y k . Generalized separable solution: w(x. y) = ϕ(x)y k+2 − 1 (k + 1)(k + 2) x a (x − t) h(t) dt + C1 x + C2 y + C3 . 4◦ . . f (t)ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation (k + 1)(k + 2)f (x)ϕϕxx − (k + 2)2 (ϕx )2 + g(x) = 0. C4 are arbitrary constants. y − where C1 . Multiplicative separable solution: w(x.468 7. 2◦ . ∂x2 ∂y 2 1 λ2 Generalized separable solution: w(x. . y) = C1 eλy + C2 y − where C1 . 9. f (t)ϕ(t) where the function ϕ = ϕ(x) is determined by the ordinary differential equation f (x)ϕϕxx − (ϕx )2 + λ−2 g(x) = 0. y) = ψ(x)eλy − 1 λ2 x a (x − t) g(t) dt + C1 x + C2 y + C3 . ∂2w ∂x∂y 2 = f (x) ∂2w ∂2w + g(x)e2λy + h(x)eλy .

which is easy to integrate. .2. . C4 are arbitrary constants. C5 are arbitrary constants and the function ϕ(z) is determined by the ordinary differential equation (abϕzz + C2 )2 = f (z)(a2ϕzz + 2C1 )(b2 ϕzz + 2C3 ) + g(z). ∂x2 ∂y 2 z = ax + by. Additive separable solution for f1 g1 ≠ 0: w(x. C2 . ∂2w 2 = ∂2w ∂2w ∂x2 ∂y 2 + f (x) ∂2w ∂y 2 . where the functions ϕ = ϕ(x) and ψ = ψ(y) are determined by the ordinary differential equations f1 (x)ϕϕxx − C4 (ϕx )2 = 0. and C3 are arbitrary constants. . y) = ϕ(x) + C1 y + C2 . to this end. ∂2w ∂x∂y Solution: 2 = f (ax + by) ∂2w ∂2w + g(ax + by). y) − leads to an equation of the form 7. x a w = U (x. w(x. .1: ∂2U ∂x∂y 2 (x − t)f (t) dt = ∂2U ∂2U . the equation should ﬁrst be solved for ϕ zz . . y) = C1 a x ∂2w 2 = f1 (x)g1 (y) ∂2w ∂2w (x − t) 1 f2 (t) dt − f1 (t) C1 y b (y − ξ) g2 (ξ) dξ + C2 x + C3 y + C4 . w(x. w1 = w(x. where C1 . Other Equations ∂x∂y The substitution 1. where C1 . is also a solution of the equation. 7. y) + C1 x + C2 y + C3 . g1 (ξ) where C1 . 3◦ . Degenerate solutions for f2 g2 = 0: w(x. y) = ϕ(x)ψ(y) + C1 x + C2 y + C3 . C4 g1 (y)ψψyy − (ψy )2 = 0. . Then the function 10. y) = ϕ(z) + C1 x2 + C2 xy + C3 y 2 + C4 x + C5 y. . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 469 + f2 (x)g2 (y). 11. where C1 and C2 are arbitrary constants and ϕ = ϕ(z) is an arbitrary function. y) = ϕ(y) + C1 x + C2 . 2◦ .2. Generalized separable solution for f2 g2 = 0: w(x. ∂x∂y ∂x2 ∂y 2 1◦ . y) is a solution of this equation. .2. Suppose w(x.5.2.7. ∂x2 ∂y 2 © 2004 by Chapman & Hall/CRC . 4◦ .

. Rozendorn (1984). Equation (3) is ﬁnally reduced.2. g(x). with the change of variable U (f ) = fx . R. to a ﬁrst-order linear equation. equation (1) is ﬁrst reduced. w is the stream function for the wind velocity. C4 are arbitrary constants. 2 2 a 1 C1 + a 2 C2 b2 C1 b2 C2 x + C3 e−λx − x + C4 . 2f gxx + a1 gxx + b1 gx − 4fx gx + 2b2 f = 0. . y) is a solution of the equation in question. + f (x) dx.2.2: ∂ 2U ∂x∂y 4. to ﬁrst-order linear equation. Traveling-wave solution: w = C3 exp − b1 C1 + b2 C2 (C1 x + C2 y) + C4 . ∂x∂y ∂x ∂y This equation is used in meteorology for describing horizontal air ﬂows. C4 are arbitrary constants. is also a solution of the equation. 4◦ . Remark. b2 ). with the substitution V (x) = hx . Then the function w1 = w(x + C1 . where C1 . . g2 = f (x). where the functions f (x). . The solutions of Items 3◦ and 4◦ can be used to obtain two other solutions by means of the following renaming: (x. . . y + C2 ) + C3 (b2 x − b1 y) + C4 . ∂w ∂w =Φ ∂y ∂x where Φ(u) is an arbitrary function. 2◦ . To this end. (1) (2) (3) 2f hxx + a1 hxx + b1 hx + 2a2 f + b2 g − (gx )2 = 0. ◦ 1 . Generalized separable solution linear in y: w = y(C1 e−λx + C2 ) + 2 C1 −2λx e + 2a1 λ= where C1 . a2 . Generalized separable solution quadratic in y: w = f (x)y 2 + g(x)y + h(x). © 2004 by Chapman & Hall/CRC . b1 ) (y. a1 . and x and y are coordinates on the earth surface. y) − 1 a2 x2 − 2 a1 y 2 2 leads to an equation of the form 7. Suppose w(x. ∂2w 2 2 = ∂2U ∂2U + b − a 1 a2 . and h(x) are determined by the system of ordinary differential equations 2f fxx + a1 fxx + b1 fx − 4(fx )2 = 0. 3. b1 b1 b1 a1 3◦ . This system can be fully integrated.470 2. . . ∂2w EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS 2 ∂x∂y First integral: = ∂2w ∂2w ∂x2 ∂y 2 + f (x) ∂2w ∂x∂y . Equation (2) is linear in g and a fundamental system of solutions of the corresponding homogeneous equation has the form g 1 = 1. = ∂x2 ∂y 2 ∂x2 f ¢e ∂2w ∂2w + a1 ∂2w + a2 Reference: E. ∂x2 ∂y 2 ∂2w ∂y 2 + b1 ∂w + b2 ∂w . ∂x∂y ∂x2 ∂y 2 ∂x2 ∂y 2 The substitution 1 w = U (x. ∂2w ∂2w ∂2w 2 ∂2w ∂2w = + a1 + a2 + b.

C7 are arbitrary constants.3) U (x. ∂x2 ∂x∂y ∂x2 ∂y 2 ∂y 2 This equation occurs in plane problems of plasticity. y) is a solution of the equation in question.2. ∂x∂y ∂x2 ∂y 2 Generalized separable solution quadratic in y: = f (x) w(x. ∂2w 2 ∂2w ∂2w ∂2w ∂2w – = .7. ∂w ∂x and apply the Legendre transformation (for details. C3 y + C4 ) + C5 x + C6 y + C7 . . . 2f (x)ϕχxx − (ψx )2 + g(x)χ + h0 (x) = 0. w is the generating function. ∂X 2 ∂X∂Y ∂Y 2 This equation is hyperbolic. . y) = ϕ(x)y 2 + ψ(x)y + χ(x). where the functions ϕ = ϕ(x). Z F = √ 1 + X2 © 2004 by Chapman & Hall/CRC . 7. ∂x∂y ∂x There are generalized separable solutions linear and quadratic in y: w(x. + h1 (x) ∂y ∂x ∂y There is a generalized separable solution of the form + g1 (x) w(x. y) = ψ2 (x)y 2 + ψ1 (x)y + ψ0 (x). w(x. ψ = ψ(x). y) = ϕ1 (x)y + ϕ0 (x). .2. The transformation (1 + X 2 )2 t = arctan X. 6. ∂2w ∂x∂y 2 ∂2w ∂2w ∂2w = f1 (x) 2 ∂2w ∂2w ∂x2 ∂y 2 + f2 (x)w + f3 (x)y 2 + f4 (x)y + f5 (x) 2 ∂2w ∂y 2 2 ∂ w ∂2w ∂ w + g2 (x)y + g3 (x) + g4 (x)w + g5 (x)y 2 + g6 (x)y + g7 (x) ∂x2 ∂x∂y ∂y 2 2 ∂w ∂w ∂w + h2 (x) + h3 (x)y + h4 (x) + 1 (x)w + 2 (x)y 2 + 3 (x)y + 4 (x). ∂x Y = Z =x ∂U ∂U +y −U ∂x ∂y 2◦ . 2f (x)ϕψxx − 4ϕx ψx + g(x)ψ + h1 (x) = 0. 2 5. and χ = χ(x) are determined by the system of ordinary differential equations 2f (x)ϕϕxx − 4(ϕx )2 + g(x)ϕ + h2 (x) = 0. Introduce the new variable h to obtain a second-order linear equation: ∂2Z ∂2Z ∂2Z + 2XY (1 + X 2 ) + Y 2 (X 2 − 1) = 0. ∂y X= ∂U . see Subsection S. Suppose w(x. ∂2w 2 = ∂2w ∂2w ∂x2 ∂y 2 + f1 (x) ∂2w ∂x2 + f2 (x) ∂2w ∂y 2 + g1 (x) ∂w + g(x)w + h2 (x)y 2 + h1 (x)y + h0 (x). y) = ∂U . y) = ϕ(x)y 2 + ψ(x)y + χ(x). −2 w1 = C1 w(C1 x + C2 . g g g g 1◦ . 8. Then the functions where C1 . EQUATIONS QUADRATIC IN THE HIGHEST DERIVATIVES 471 + g2 (x) ∂w ∂y . are also solutions of the equation. ξ= 1 2 (1) ln(1 + X 2 ) − ln Y .

. τ ) is determined by the linear heat equation ∂U ∂ 2 U − = 0. ∂x ∂y References: Yu. Then the function w1 = C1 w(x + C2 . see the books by Tikhonov and Samarskii (1990) and Polyanin (2002). A. C2 . Remark. ∂y w=x ∂w ∂w +y − w. ∂τ ∂ξ 2 4◦ . The original equation is invariant under the Legendre transformation x= p ¢i ∂w . τ= f (t) dt. C2 .3. “Two-dimensional” solutions: q ∂w ∂w – f (t) ∂w ∂ 2 w – g(t) ∂w 2 w = U (z. see Chernousko (1971) and Chernousko and Kolmanovskii (1978). z = y + ϕ(x. Radayev (1988).1. 1. S. The solutions of Items 2◦ and 3◦ are special cases of a more general solution with the form w = U (z. where C1 . 1◦ . τ= f (t) dt + C3 . 2◦ . Radayev. N. I. and the function U = U (z. y + C3 . and L. Yu. Astaﬁev. ∂τ ∂z 2 Reference: A. t) satisﬁes the ﬁrst-order nonlinear partial differential equation ∂ϕ ∂ϕ = g(t). Volosov (2002). “Two-dimensional” solution: w = u(ξ. and C3 are arbitrary constants. where the function ϕ = ϕ(x. z=y 2 x g(t) dt + C1 x 1/2 + C2 . ∂t2 ∂ξ 2 (2) For solutions of equation (2). V. y. Bratus’ and K. ◦ © 2004 by Chapman & Hall/CRC p ¢i 3 . . ∂t ∂x (1) . τ ). and C3 are arbitrary constants. ∂x y= ∂w . η) is determined by the linear heat equation ∂u ∂ 2 u − = 0. C4 are arbitrary constants. ξ = y + C1 x + 1 C1 g(t) dt + C2 . . and the function u = u(ξ. where C1 .472 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS brings (1) to a constant-coefﬁcient linear equation: ∂2F ∂2F = − F. The variable t = T − τ plays the role of “backward” time. τ ). τ= f (t) dt + C3 . t). 7. N. Bellman Type Equations and Related Equations 7. V. Equations with Quadratic Nonlinearities = 0. t) + C4 . τ ). ∂t ∂x ∂x ∂y 2 ∂y This equation occurs in problems of optimal correction of random disturbances and is a consequence of the Bellman equation. where C1 . t) is a solution of the equation in question.3. is also a solution of the equation. . Suppose w(x. Stepanova (2001).

◦ 5 . BELLMAN TYPE EQUATIONS AND RELATED EQUATIONS 473 and the function U = U (z. where λ is an arbitrary constant. F. “Two-dimensional” solution: w = eλx θ(y. F. © 2004 by Chapman & Hall/CRC . Remark. 2002) C2 = ψ(C1 ). x− 2 C1 where ψ = ψ(C1 ) is an arbitrary function and the prime denotes a derivative with respect to the argument. The general integral of equation (3) can be represented in parametric form with the complete integral (4) and the two relations C2 = ϕ(C1 ). t). Zaitsev (2002). and the function θ = θ(y. x− λ2 2 C1 g(t) dt + ϕ (C1 ) = 0. D. The general integral of equation (1) can be represented in parametric form with the complete integral (2) and the two relations (see Kamke. λθ ∂t ∂y ∂y 7◦ . (3) ∂t ∂x ∂x A complete integral of equation (3) is given by ζ = C 1 x + λ2 f (t) + 1 g(t) dt + C2 . − ∂τ ∂z 2 A complete integral of equation (1) is given by ϕ = C1 x + 1 C1 g(t) dt + C2 . where ϕ = ϕ(C1 ) is an arbitrary function. and Moussiaux. where λ is an arbitrary constant. s ¢r Reference: A. and Polyanin. To the solution of Item 2◦ there corresponds ψ(C1 ) = const. ◦ 6 . s ¢r Reference: A. C1 and C2 play the role of parameters. t) is determined by the “two-dimensional” equation 2 ∂2θ ∂θ ∂θ − λf (t)θ 2 − g(t) = 0. 1 g(t) dt + ψ (C1 ) = 0. D. Polyanin and V. Zaitsev (2002). C1 and C2 play the role of parameters. Zaitsev. Polyanin and V.7. t) .3. Cauchy problems and self-similar solutions of the equation for power-law f (t) and g(t) are discussed in Chernousko (1971) and Chernousko and Kolmanovskii (1978). t) is determined by the ﬁrst-order partial differential equation ∂ζ ∂ζ ∂ζ − λ2 f (t) − λ2 g(t) = 0. (2) where C1 and C2 are arbitrary constants. 1965. and the function ζ = ζ(x. “Two-dimensional” solutions: w = exp λy + ζ(x. τ ) is determined by the linear heat equation ∂U ∂ 2 U = 0. C1 (4) t where C1 and C2 are arbitrary constants.

Zaitsev. t) ∂w ∂ 2 w – g(x. The variable t = T − τ plays the role of “backward” time. ∂w ∂w ∂t ∂x EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS – f (t) ∂w ∂ 2 w ∂x ∂y 2 – g(t)h(x) ∂w ∂y 2 = 0. z = y + ϕ(x. 2 where the function U (y. 2◦ . the function ϕ = ϕ(x. where λ is an arbitrary constant and the function ζ = ζ(x. ∂t ∂x ∂x Reference: A. t) is determined by the differential equation (λ is an arbitrary constant) λU 4. “Two-dimensional” solution: w(x. “Two-dimensional” solution: = 0. n + 1 is the dimensionality of the equations of motion of the controllable system (n is a nonnegative integer). t) can be found in Polyanin. ∂w ∂ 2 w n ∂w ∂w 2 ∂w ∂w – f (t) + – g(t)h(x) = 0. ∂w ∂w – f (t) ∂U − λf (t)U ∂t – g(x. t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ϕ ∂ϕ = g(x. t) ∂w ∂y 2 = 0. see Tikhonov and Samarskii (1990) and Polyanin (2002). “Two-dimensional” solutions: w = exp λy + ζ(x. ∂t ∂x ∂x . where λ is an arbitrary constant and the function ζ = ζ(x. ∂w ∂w – f (x. t) = exp λ h(x) dx U (y. h(x) dx leads to an equation of the form 7. t) ∂ 2 U n ∂U + ∂y 2 y ∂y ∂w ∂y 2 − g(t) ∂U ∂y = 0.1. t) .474 2. t) = 0. τ= f (t) dt. see Chernousko and Kolmanovskii (1978). and Moussiaux (2002). τ ) is determined by the linear heat equation ∂U ∂ 2 U − = 0.3. Zaitsev (2002). t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ζ ∂ζ ∂ζ − λ2 f (t) − λ2 g(x.1 for w = w(z. t). F. y. D. w = U (z. (1) ∂t ∂x and the function U = U (z. The substitution z = 3. ∂t ∂x ∂x ∂y 2 y ∂y ∂y This equation occurs in problems of optimal correction of random disturbances and is a consequence of the Bellman equation. t) − λ2 g(x. t). ∂w ∂ 2 w ∂t ∂x ∂x ∂y 2 ◦ 1 . w = exp λy + ζ(x. t). t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ζ ∂ζ ∂ζ − λ2 f (x. For solutions of equation (2). Here. τ ). (2) ∂τ ∂z 2 Complete integrals and the general solutions (integrals) of equation (1) for various g(x. u u © 2004 by Chapman & Hall/CRC w ¢v ∂t ∂x ∂x ∂y 2 “Two-dimensional” solutions: 5. t) = 0. Polyanin and V. y. t). t) .

Equations with Power-Law Nonlinearities 1. where C1 . τ ) is determined by the linear heat equation ∂U ∂ 2 U − = 0. . t) + C4 . C4 are arbitrary constants and the function U = U (z. t) is a solution of the equation in question. Then the function w1 = C1 w(x + C2 . z = y + ϕ(x. . The solutions of Items 2◦ and 3◦ are special cases of the more general solution w = U (z. ∂w ∂t ∂w ∂x k – f (t) ∂w ∂x k ∂2w ∂y 2 – g(t) ∂w ∂y k+1 = 0. 2◦ . τ ). τ= f (t) dt + C3 . τ ). Suppose w(x. where C1 . τ ) is determined by the linear heat equation ∂U ∂ 2 U − = 0. C2 . τ= f (t) dt. . (2) © 2004 by Chapman & Hall/CRC . BELLMAN TYPE EQUATIONS AND RELATED EQUATIONS 475 7. A. τ ). This equation occurs in problems of optimal correction of random disturbances and is a consequence of the Bellman equation. y + C3 . 3◦ . ∂τ ∂z 2 y ¢x Reference: A. ∂τ ∂z 2 A complete integral of equation (1) is given by ϕ = C1 x + 1 k C1 g(t) dt + C2 . . ξ = y + C1 x + 1 k C1 g(t) dt + C2 . y. “Two-dimensional” solution: w = U (z. see Chernousko (1971) and Chernousko and Kolmanovskii (1978). z = y + (x + τ= k C2 ) k+1 f (t) dt + C1 . Volosov (2002). ∂τ ∂ξ 2 4◦ . “Two-dimensional” solution: w = u(ξ. is also a solution of the equation. . . where the function ϕ = ϕ(x. where C1 . η) is determined by the linear heat equation ∂u ∂ 2 u − = 0. (1) and the function U = U (z. C4 are arbitrary constants. t) satisﬁes the ﬁrst-order nonlinear partial differential equation ∂ϕ ∂t ∂ϕ ∂x k = g(t).7. .3. Bratus’ and K.2. 1◦ . (k + 1)k+1 kk g(t) dt + C3 1 k+1 + C4 .3. t). The variable t = T − τ plays the role of “backward” time. S. . and C3 are arbitrary constants and the function u = u(ξ.

and Moussiaux. ¢ Reference: A. τ= f (t) dt. t) . t) is determined by the “two-dimensional” equation k+1 ∂2θ g(t) ∂θ ∂θ − f (t) 2 − = 0. and Moussiaux. © 2004 by Chapman & Hall/CRC . k g(t) dt + ψ (C1 ) = 0. 2. (3) ∂t ∂x ∂x A complete integral of this equation has the form (see Polyanin. ∂t ∂y (λθ)k ∂y 7◦ . where λ is an arbitrary constant and the function ζ = ζ(x.1 for w = w(z. Zaitsev (2002). x−k λk+1 k+1 C1 g(t) dt + ϕ (C1 ) = 0. Zaitsev (2002). Polyanin and V. The general integral of equation (3) can be expressed in parametric form with the complete integral (4) and the two relations C2 = ϕ(C1 ). The substitution z = 3. ◦ 5 . “Two-dimensional” solution: w = eλx θ(y. D. k C1 (4) where C1 and C2 are arbitrary constants. “Two-dimensional” solution: w = exp λy + ζ(x. = 0. [h(x)]1/k dx leads to an equation of the form 7. t) ∂t ∂x ∂x ∂y 2 ∂y ◦ 1 . ∂w ∂t ∂w ∂x k – f (t) ∂w ∂x k ∂2w ∂y 2 – g(t)h(x) ∂w ∂y k+1 = 0. where ϕ = ϕ(C1 ) is an arbitrary function. x − k+1 C1 where ψ = ψ(C1 ) is an arbitrary function and the prime denotes a derivative. t). The general integral of equation (1) can be expressed in parametric form with the complete integral (2) and the two relations (see Kamke. “Two-dimensional” solution: w = U (z.3. t). 2002) C2 = ψ(C1 ). t) is determined by the ﬁrst-order partial differential equation k k ∂ζ ∂ζ ∂ζ − λ2 f (t) − λk+1 g(t) = 0. 2002) ζ = C1 x + λ2 f (t) + λk+1 g(t) dt + C2 . y. F. where λ is an arbitrary constant and the function θ = θ(y. Zaitsev. Zaitsev. Polyanin and V. F. τ ). ¢ Reference: A. and Polyanin. 1965. C1 and C2 play the role of parameters. D.476 EQUATIONS INVOLVING MIXED DERIVATIVES AND SOME OTHER EQUATIONS where C1 and C2 are arbitrary constants. z = y + ϕ(x. ◦ 6 . C1 and C2 play the role of parameters. k+1 ∂w k ∂ 2 w ∂w ∂w ∂w k – f (t) – g(x.2. t). Cauchy problems and self-similar solutions of the equation for power-law f (t) and g(t) are discussed in Chernousko (1971) and Chernousko and Kolmanovskii (1978).

and Moussiaux (2002). t) ∂w ∂y k+1 = 0. 2◦ .7. where λ is an arbitrary constant. w = exp λy + ζ(x. D. t) can be found in Polyanin. see Tikhonov and Samarskii (1990) and Polyanin (2002). For solutions of equation (2). t) . t) ∂ζ ∂x k − λk+1 g(x. (1) and the function U = U (z. the function ϕ = ϕ(x. “Two-dimensional” solution: w = exp λy + ζ(x. t). and the function ζ = ζ(x. Zaitsev. ∂τ ∂z 2 (2) Complete integrals and the general solutions (integrals) of equation (1) for various g(x. ∂t ∂x ∂x “Two-dimensional” solution: 4. Reference: A. τ ) is determined by the linear heat equation ∂U ∂ 2 U − = 0. t) . t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ζ ∂t ¢ ∂ζ ∂x k − λ2 f (t) ∂ζ ∂x k − λk+1 g(x. t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ϕ ∂t ∂ϕ ∂x k = g(x. t) = 0. t) = 0. F. and the function ζ = ζ(x. t) ∂w k ∂2w ∂y 2 – g(x. t) is determined by the ﬁrst-order nonlinear partial differential equation ∂ζ ∂t ∂ζ ∂x k − λ2 f (x. BELLMAN TYPE EQUATIONS AND RELATED EQUATIONS 477 Here. Polyanin and V.3. © 2004 by Chapman & Hall/CRC . where λ is an arbitrary constant. Zaitsev (2002). ∂w ∂w k – f (x.

2◦ . Equations Involving the First Derivative in t Preliminary remarks. .1. ξ = kx + λt. Solution: ◦ © 2004 by Chapman & Hall/CRC . . ∂t ∂x2 1◦ . In the general case. ξ = kx + λt. where A. . is also a solution of the equation.Chapter 8 Second›Order Equations of General Form 8. C5 are arbitrary constants. Equations of the Form ∂w = F w. ∂ w ∂t ∂x ∂x2 ∂w =F ∂2w where C1 . This subsection presents special cases where equation (1) admits exact solutions other than traveling wave (2).1. k. . 2◦ . 1. t) is a solution of this equation. Consider the equation ∂w ∂ 2 w ∂w = F w. Then the function −2 2 w1 = C1 w(C1 x + C2 . k 2 wξξ − λwξ = 0. . is also a solution of the equation. t) = (Ax + B)t + C + ϕ(x). Suppose w(x. where the function ϕ(x) is determined by the ordinary differential equation F ϕxx = Ax + B. . t) = At + B + ψ(ξ). . and C are arbitrary constants. (2) where k and λ are arbitrary constants and the function w(ξ) is determined by the ordinary differential equation F w. Suppose w(x. 3◦ . Generalized separable solution: w(x. t) = F (A)t + 2 Ax2 + Bx + C. ∂w .1. Then the function w(x + C 1 . and λ are arbitrary constants and the function ψ(ξ) is determined by the autonomous ordinary differential equation F k 2 ψξξ = λψξ + A. kwξ . Additive separable solution: 1 w(x. (1) ∂t ∂x ∂x2 1◦ . B. w(x. t + C2 ). B. 2 8. where C1 and C2 are arbitrary constants. equation (1) admits traveling-wave solution w = w(ξ). where A. 4 . t) is a solution of equation (1). C1 t + C3 ) + C4 x + C5 .

Solution: SECOND-ORDER EQUATIONS OF GENERAL FORM 1 w(x. the βi . Special case 1. A2 . Additive separable solution: k + C2 x + aC1 t + C3 . ∂x2 f (z) = Fz (z). C1 . F Θζζ + 1 ζΘζ − Θ = 0. H. β1 β4 − β2 β3 ≠ 0) and the subscripts x and x denote the corresponding partial derivatives. to this end. Solution: w(x. K. k. and the function U (x) is determined by the ordinary differential equation 2aλ exp(λUxx ) + B1 (x2 + A1 x + A2 ) = 0. where the function u(x) is determined by the autonomous ordinary differential equation (u xx )k − u = 0. Special case 2. 2 7◦ . and N. where C1 . B2 . x ζ= √ . 2◦ . Ibragimov (1994). k ≠ 1. and C2 are arbitrary constants. ¯¯ where α. and λ are arbitrary constants and the function U (ξ) is determined by the autonomous ordinary differential equation F k 2 Uξξ + A = λUξ . 6◦ . R. © 2004 by Chapman & Hall/CRC . ξ = kx + λt. F (wxx ) = α α ¡ ¢ References: I. t) = a(1 − k)t + C1 1 1−k u(x) + C2 . whose general solution can be written out in implicit form: £ 1+k 2k u k + C3 1+k −1/2 du = x + C4 . Self-similar solution: x = β 1 x + β2 w + γ 2 . N. k > 0. Akhatov. t) = ∂w brings the original equation to an equation of the form 1. Ibragimov (1989). takes the equation in question to an equation of the ¯ same form. Sh. t where the function Θ(ζ) is determined by the ordinary differential equation w(x. where A. t) = t Θ(ζ). and C3 are arbitrary constants.480 5◦ . The substitution u(x. Generalized separable solution: 1 2 (x + A1 x + A2 ) ln(B1 t + B2 ) + C1 x + C2 . ¯ 2 w x = β3 x + β 4 w x + γ 3 . 1◦ . Gazizov. The transformation ¯ t = αt + γ1 . 2λ where A1 .6. Equation: ∂2w ∂w =a ∂t ∂x2 w(x. t) = U (x) − which is easy to integrate. t) = 1 C x2 2 1 k .3: ∂x ∂u =f ∂t 8◦ . the equation should ﬁrst be solved for U xx .18. B. C2 . Equation: ∂2w ∂w = a exp λ ∂t ∂x2 . ¯ 2 1 w = β1 β4 w + 1 β3 x2 + γ3 x + γ4 t + γ5 + β2 β3 (xwx − w) + γ3 wx + 2 β4 wx . and the γi are arbitrary constants (α ≠ 0. The right-hand side of the equation becomes γ4 β1 β4 − β 2 β3 ¯ ¯¯¯ F (wxx ) + . ∂u ∂x ∂2u . t) = 2 Ax2 + Bx + C + U (ξ). w(x. H. B1 .

ϕ1 = ϕ2 = The prime denotes a derivative with respect to t. Generalized separable solution: where the functions ϕk = ϕk (t) are determined by the autonomous system of ordinary differential equations ϕ3 = 18aϕ2 . + aw. 2 8 45 aϕ2 ϕ3 . C. . k. ∂t ∂x2 w(x. and C3 are arbitrary constants. 9 aϕ2 . cos2 (Ax + B) Generalized separable solution: where A. Special case. t) = ψ1 (t) + ψ2 (t)x + ψ3 (t)x2 + ψ4 (t)x3 . and the function ϕ(ξ) is determined by the autonomous ordinary differential equation F kϕξ . ξ = kx + λt.1. ∂w =F ∂w . ∂2w where A. t) = ϕ1 (t) + ϕ2 (t)x3/2 + ϕ3 (t)x3 . is also a solution of the equation. Generalized separable solution: w(x. t + C2 ) + C3 eat . Generalized separable solution cubic in x: w(x. 3 ψ3 = 18aψ3 ψ4 . 2◦ . ∂w =F ∂w . t) = At + B + ϕ(ξ). 4 1◦ . t) = θ(x) + C3 + C4 . 2 ψ4 = 18aψ4 . . 3. and λ are arbitrary constants. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t Special case 3. Equation: 481 ∂2w ∂w = a ln . where the functions ψk = ψk (t) are determined by the autonomous system of ordinary differential equations ψ1 = 2aψ2 ψ3 . and D are arbitrary constants. k 2 ϕξξ − λϕξ − A = 0. 2. 3◦ . B. Then the function w1 = w(x + C1 . whose solution can be written out in implicit form. C1 t + C 2 where C1 . ∂t ∂x ∂x2 w(x. C2 . . B. © 2004 by Chapman & Hall/CRC . C4 are arbitrary constants and the function θ = θ(x) is determined by the autonomous ordinary differential equation aθx θxx + C1 θ + C1 C3 = 0. ∂2w where C1 . 2 ψ2 = 2a(2ψ3 + 3ψ2 ψ4 ). this equation has a more complicated exact solution of the form w(x. . ∂t ∂x ∂x2 Apart from a traveling-wave solution. . ∂t ∂x ∂x2 1◦ . t) = (at + C) ln 2A2 (at + C) − 1 + D.8. t) is a solution of this equation. Equation: ∂w ∂ 2 w ∂w =a . Suppose w(x.

482 2◦ . the above solution converts to a traveling-wave solution. ∂w = aw ∂w +F ∂w . where C1 . t) is a solution of this equation. Degenerate solution: w(x. Uζζ + aU Uζ = C2 Uζ + 2C1 . © 2004 by Chapman & Hall/CRC . and C3 are arbitrary constants. 0). ψt = aϕψ + bψ + F (ϕ. ϕt = aϕ2 + bϕ. Then the function 4. is also a solution of the equation. 3◦ . t + C3 ) + C1 bebt . ζ = x + aC1 t2 + C2 t. C2 . t) = − 3◦ . t) = (C1 x + C2 )eat + eat 3◦ . is also a solution of the equation. where C1 and C2 are arbitrary constants and the function U (ζ) is determined by the autonomous ordinary differential equation F Uζ . ∂2w where the functions ϕ(t) and ψ(t) are determined by the system of ordinary differential equations where λ is an arbitrary constant and the function w(z) is determined by the autonomous ordinary differential equation F wz . e−at F (C1 eat . ∂w = aw ∂w +F ∂w . Degenerate solution: SECOND-ORDER EQUATIONS OF GENERAL FORM w(x. ∂t ∂x ∂x ∂x2 1◦ . Suppose w(x. z = x + λt. 0 dτ . Solution: x + C1 1 + aτ τ τF − 1 . t) is a solution of this equation. . and C3 are arbitrary constants. t + C3 ) + C1 . C2 . t) = ϕ(t)x + ψ(t). z = x + λt. Suppose w(x. 2◦ . Traveling-wave solution: w = w(z). 0) dt. Traveling-wave solution: w = w(z). Then the function w1 = w(x + aC1 ebt + C2 . which is easy to integrate (the ﬁrst equation is a Bernoulli equation and the second one is linear in ψ). wzz − λwz + aw = 0. ∂t ∂x ∂x ∂x2 1◦ . In the special case C1 = 0. aτ τ = t + C2 . 2◦ . t) = U (ζ) + 2C1 t. ∂2w w(x. Degenerate solution: w(x. where λ is an arbitrary constant and the function w(z) is determined by the autonomous ordinary differential equation F wz . 5. where C1 . w1 = w(x + aC1 t + C2 . + bw. wzz + awwz − λwz + bw = 0.

g(w) = w1−3k g(w−1 ). wx = wF w−1 . Ibragimov (1994). 2◦ . ∂x t w(y. ¯ ¥ ¢¤ F w. w−3 wx . t) converts a (nonzero) solution w(x. C1 t + C3 ). 2◦ . where C1 . where k and λ are arbitrary constants and the function ϕ(z) is determined by the autonomous ordinary differential equation F k 8. wx (x0 . ϕz 2 ϕzz . and E.k ϕ ϕ = λϕz + ϕ. t) = tϕ(z). R. −1 w1 = C1 w(x + C2 . α2 ) − λ = 0. is also a solution of the equation. t + C3 ). wx . ∂t w ∂x w ∂x2 1◦ . ¯ t = t − t0 . © 2004 by Chapman & Hall/CRC .1. M. z = kx + λ ln |t|. = λ. H. t) dy − F w(x0 . Multiplicative separable solution: w(x. Feix. In the special case it follows from (1) that ¯ F w. Strampp (1982). N. t) of a similar ¯ ¯ ¯ equation: ∂ ¯ ∂w ¯ = F w. Then the function w1 = C1 w(x + C2 . t0 w(x. R. t) = ¯ ¯ ¯ 1 w(x. wx = g(w)(wx )k . t) of the original equation to a solution w(x. F ϕ ϕ This equation has particular solutions of the form ϕ(x) = eαx . Suppose w(x. Solution: w(x. ∂w =F 1 ∂w . 1 ∂2w where C1 . The transformation 6. x=− ¯ x x0 wx = ∂w . is also a solution of the equation. τ ).8. t) is a solution of this equation. ∂t ∂x 1◦ . 1 ∂w 1 ∂ 2 w ∂w = wF . t) is a solution of this equation. Munier. J. and C3 are arbitrary constants. Fijalkow (1984). Suppose w(x. t) = Ceλt ϕ(x). EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t 483 ∂ ∂w = F w. C2 . Then the function 7. where C and λ are arbitrary constants and the function ϕ(x) is determined by the autonomous ordinary differential equation ϕx ϕxx . C2 . . ∂t w ∂x w ∂x2 1◦ . Burgan. and C3 are arbitrary constants. ¯ ¯¯ ¯ ∂t ∂x ¯ where ¯ (1) F w. ¯ References: W. where α is a root of the algebraic (or transcendental) equation F (α. wx = g(w)(wx )k . . τ ) dτ . A. wx . 2◦ .

. C2 . where C. and the function ψ(ξ) is determined by the autonomous ordinary differential equation ψξξ ψξ = βψξ + λψ. respectively. t) is a solution of the equation in question. t) is a solution of this equation. where C. ϕβ−1 F ϕ ϕ 3◦ . = A. and C3 are arbitrary constants. ξ = kx + λ ln(t + C). is also a solution of the equation. Suppose w(x.1. . k. ∂w ∂ 2 w ∂w = eβw F . β w(x. t) = Ceλt ψ(ξ). k 2 Θξξ = λΘξ − . Multiplicative separable solution: w(x. ∂t w ∂x w ∂x2 For the cases β = 0 and β = 1. and the function Θ(ξ) is determined by the autonomous ordinary differential equation 1 eβΘ F kΘξ . Suppose w(x. and C3 are arbitrary constants. 9. ∂w = wβ F 1 ∂w . C2 t + C3 ) + ln C2 . w(x.1. k 2 zz = λΘz + Θ Θ 1−β 10. Then the function 1 w1 = w(x + C1 . β where A and B are arbitrary constants. see equations 8. β where C. β where C1 . and β are arbitrary constants.1. k. t) = − 1 2◦ . Then the function where C1 . C1 t + C3 ). where A and B are arbitrary constants. and λ are arbitrary constants. ψF k . C2 . t) = − 3◦ . t) = (t + C) 1−β Θ(z). ξ = kx + βt. ∂t ∂x ∂x2 1◦ . Θβ F k z . k.1. 1 ∂2w 1◦ . and the function ϕ(x) is determined by the autonomous ordinary differential equation eβϕ F ϕx .7 and 8. 1 ln(Aβt + B) + ϕ(x).8. Solution: SECOND-ORDER EQUATIONS OF GENERAL FORM w(x. Solution: w(z. and the function ϕ(x) is determined by the autonomous ordinary differential equation ϕx ϕxx . k 2 ψ ψ This equation has particular solutions of the form ψ(ξ) = eµξ . 2◦ . λ.484 3◦ . Additive separable solution: © 2004 by Chapman & Hall/CRC . is also a solution of the equation. β−1 w1 = C1 w(x + C2 . ϕxx + A = 0. t) = (1 − β)At + B 1 1−β ϕ(x). z = kx + λ ln(t + C). and the function Θ(z) is determined by the autonomous ordinary differential equation Θ 1 Θ Θ. Solution: 1 ln(t + C) + Θ(ξ). and λ are arbitrary constants.

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

485

**∂ 2 w ∂w ∂w . =F ∂t ∂x2 ∂x This is a special case of equation 8.1.1.2. 11. 1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function
**

−1 w1 = C1 w(x + C2 , C1 t + C3 ) + C4 ,

where C1 , . . . , C4 are arbitrary constants, is also a solution of the equation. 2◦ . Solution: w(x, t) = At + B + ϕ(ξ), ξ = kx + λt,

where A, B, k, and λ are arbitrary constants, and the function ϕ(ξ) is determined by the autonomous ordinary differential equation F kϕξξ /ϕξ = λϕξ + A. If A = 0, the equation has a traveling-wave solution. 3◦ . Solution: w(x, t) = tΘ(z) + C, z = kx + λ ln |t|

where C, k, β, and λ are arbitrary constants, and the function Θ(z) is determined by the autonomous ordinary differential equation F kΘzz /Θz = λΘz + Θ. . ∂t ∂x ∂x This is a special case of equation 8.1.1.2. 12. F ∂x2 1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function w1 = C1 w(x + C2 , t + C3 ) + C4 , where C1 , . . . , C4 are arbitrary constants, is also a solution of the equation. 2◦ . Solution: w(x, t) = At + B + ϕ(z), z = kx + λt, ∂w = ∂w ∂2w ∂w

where A, B, k, and λ are arbitrary constants, and the function ϕ(z) is determined by the autonomous ordinary differential equation kϕz F kϕzz /ϕz = λϕz + A. 3◦ . Solution: w(x, t) = Aeβt Θ(ξ) + B, ξ = kx + λt,

where A, B, k, β, and λ are arbitrary constants, and the function Θ(ξ) is determined by the autonomous ordinary differential equation kΘξ F kΘξξ /Θξ = λΘξ + βΘ. ∂w β ∂ 2 w ∂w ∂w = F . ∂t ∂x ∂x2 ∂x This is a special case of equation 8.1.1.2. For the cases β = 0 and β = 1, see equations 8.1.1.11 and 8.1.1.12, respectively. 13. 1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function

β−1 w1 = C1 w(x + C2 , C1 t + C3 ) + C4 ,

**where C1 , . . . , C4 are arbitrary constants, is also a solution of the equation.
**

© 2004 by Chapman & Hall/CRC

486

SECOND-ORDER EQUATIONS OF GENERAL FORM

**2◦ . Generalized separable solution: w(x, t) = A(1 − β)t + B
**

1 1−β

ϕ(x) + C,

where A, B, and C are arbitrary constants, and the function ϕ(x) is determined by the autonomous ordinary differential equation ϕx 3◦ . Solution:

1 β

F ϕxx /ϕx = Aϕ. z = kx + λ ln(t + A),

w(x, t) = (t + A) 1−β Θ(z) + B,

where A, B, k, and λ are arbitrary constants, and the function Θ(z) is determined by the autonomous ordinary differential equation k β Θz ∂w ∂t ∂w ∂x

2 β

F kΘzz /Θz = λΘz + 1 ∂2w

1 Θ. 1−β

14.

= wF

+ aw2 ,

w ∂x2

.

**This is a special case of equation 8.1.2.11. 15. ∂w ∂t = wF 1 ∂2w w ∂x2 ,w ∂2w ∂x2 – ∂w ∂x
**

2

.

**This is a special case of equation 8.1.2.12. 16. ∂w ∂t = wF ∂2w ∂x2 , 2w ∂2w ∂x2 – ∂w ∂x
**

2

+G

∂2w ∂x2

, 2w

∂2w ∂x2

–

∂w ∂x

2

.

**This is a special case of equation 8.1.2.13.
**

2 8.1.2. Equations of the Form ∂w = F t, w, ∂w , ∂ w ∂t ∂x ∂x2

1.

∂w ∂t

= F t,

∂w ∂ 2 w , ∂x ∂x2

+ aw.

Suppose w(x, t) is a solution of this equation. Then the function w1 = w(x, t) + Ceat , where C are arbitrary constants, is also a solution of the equation. 2. ∂w ∂t + f (t)w ∂w ∂x = F t, ∂w ∂ 2 w , ∂x ∂x2 + g(t)w.

Suppose w(x, t) is a solution of this equation. Then the function w1 = w x + ψ(t), t + ϕ(t), ϕ(t) = C exp g(t) dt , ψ(t) = − f (t)ϕ(t) dt,

**where C is an arbitrary constant, is also a solution of the equation.
**

© 2004 by Chapman & Hall/CRC

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

487

1 ∂2w ∂w . = wF t, ∂t w ∂x2 Multiplicative separable solutions: 3. w(x, t) = A exp λx + F (t, λ2 ) dt , F (t, λ2 ) dt , F (t, −λ2 ) dt ,

w(x, t) = A cosh(λx) + B sinh(λx) exp w(x, t) = A cos(λx) + B sin(λx) exp where A, B, and λ are arbitrary constants. ∂t w ∂x w ∂x2 Multiplicative separable solution: 4. ∂w = wF t, 1 ∂w , 1 ∂2w .

w(x, t) = A exp λx + where A and λ are arbitrary constants. . ∂t ∂x ∂x2 1◦ . Multiplicative separable solution for k ≠ −1: 5. ∂w = wF t, wk ∂w , w2k+1 ∂2w

F (t, λ, λ2 ) dt ,

w(x, t) = C1 (k + 1)x + C2

1 k+1

ϕ(t),

**where the function ϕ = ϕ(t) is determined by the ﬁrst-order ordinary differential equation
**

2 ϕt = ϕF t, C1 ϕk+1, −kC1 ϕ2k+2 .

2◦ . For k = −1, see equation 8.1.2.4. 6. 1 ∂w 1 ∂ 2 w ∂w = f (t)wβ Φ , ∂t w ∂x w ∂x2 The transformation w(x, t) = G(t)u(x, τ ), τ= + g(t)w.

f (t)Gβ−1 (t) dt,

G(t) = exp

g(t) dt ,

leads to a simpler equation of the form 8.1.1.9: 1 ∂u 1 ∂ 2 u ∂u = uβ Φ , , ∂τ u ∂x u ∂x2 which has a traveling-wave solution u = u(Ax + Bτ ) and a solution in the multiplicative form u = ϕ(x)ψ(τ ). 7. ∂w k ∂2w ∂w = f (t) Φ ∂t ∂x ∂x2 Generalized separable solution: ∂w ∂x + g(t)w + h(t).

w(x, t) = ϕ(t)Θ(x) + ψ(t),

© 2004 by Chapman & Hall/CRC

488

SECOND-ORDER EQUATIONS OF GENERAL FORM

where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = Af (t)ϕk + g(t)ϕ, ψt = g(t)ψ + Bf (t)ϕk + h(t), (1) (2)

C is an arbitrary constant, and the function Θ(x) is determined by the second-order ordinary differential equation k Θx Φ Θxx /Θx = AΘ + B. (3) The general solution of system (1), (2) is expressed as ϕ(t) = G(t) C − kA ψ(t) = DG(t) + G(t) f (t)G

k−1

(t) dt

1 1−k

,

G(t) = exp dt , G(t)

g(t) dt ,

Bf (t)ϕk (t) + h(t)

where A, B, C, and D are arbitrary constants. For k = 1 and Φ(x, y) = Φ(y), a solution to equation (3) is given by Θ(x) = αeλx − B/A, where α is an arbitrary constant and λ is determined from the algebraic (or transcendental) equation λΦ(λ) = A. 8. ∂w ∂ 2 w ∂w = f (t)eβw Φ , ∂t ∂x ∂x2 The transformation w(x, t) = u(x, τ ) + G(t), + g(t).

τ=

f (t) exp βG(t) dt,

G(t) =

g(t) dt,

leads to a simpler equation of the form 8.1.1.10: ∂u ∂ 2 u ∂u = eβu Φ , , ∂τ ∂x ∂x2 which has a traveling-wave solution u = u(Ax+Bτ ) and an additive separable solution u = ϕ(x)+ψ(τ ). 9. ∂w ∂t ∂x With the transformation = f (t)Φ w, ∂w , ∂2w ∂x2 + g(t) ∂w ∂x .

w = U (z, τ ), one arrives at the simpler equation

z =x+

g(t) dt,

τ=

f (t) dt

∂U ∂ 2 U ∂U = Φ U, , ∂τ ∂z ∂z 2 which has a traveling-wave solution U = U (kz + λτ ). 10. + aw2 . ∂t ∂x2 Multiplicative separable solutions: √ √ w(x, t) = C1 sin x a + C2 cos x a = wF t, w w(x, t) = C1 sinh x ∂w ∂2w

,

exp

F (t, 0) dt exp F (t, 0) dt

if a > 0, if a < 0,

|a| + C2 cosh x |a|

**where C1 and C2 are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

489

. ∂t ∂x w ∂x2 1◦ . Multiplicative separable solution for a > 0: √ √ w(x, t) = C1 sin x a + C2 cos x a 11.

∂w

= wF t,

∂w

2

+ aw2 ,

1 ∂2w

ϕ(t),

where C1 and C2 are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ordinary 2 2 differential equation ϕt = ϕF t, a(C1 + C2 )ϕ2 , −a . 2◦ . Multiplicative separable solution for a < 0: w(x, t) = C1 e

√ |a| x

+ C2 e−

√ |a| x

ϕ(t),

where C1 and C2 are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt = ϕF t, 4C1 C2 aϕ2 , −a .

Example. For C1 C2 = 0, a solution is given by w(x, t) = C exp where C is an arbitrary constant.

¦

|a| x +

F (t, 0, −a) dt ,

12.

1 ∂2w ∂2w ∂w = wF t, ,w – ∂t w ∂x2 ∂x2 1◦ . Multiplicative separable solution:

∂w ∂x

2

.

w(x, t) = C exp λx + where C and λ are arbitrary constants. 2◦ . Multiplicative separable solution:

F (t, λ2 , 0) dt ,

w(x, t) = (Aeλx + Be−λx )ϕ(t), where A, B, and λ are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt = ϕF t, λ2 , 4ABλ2 ϕ2 . 3◦ . Multiplicative separable solution: w(x, t) = [A sin(λx) + B cos(λx)]ϕ(t), where A, B, and λ are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt = ϕF t, −λ2 , −λ2 (A2 + B 2 )ϕ2 .

¨ ¢§

Reference: Ph. W. Doyle (1996), the case ∂t F ≡ 0 was treated.

∂t ∂x2 ∂x2 ∂x Generalized separable solution quadratic in x:

13.

∂w

= wF t,

∂2w

, 2w

∂2w

–

∂w

2

+ G t,

∂2w ∂x2

, 2w

∂2w ∂x2

–

∂w ∂x

2

.

w = ϕ1 (t)x2 + ϕ2 (t)x + ϕ3 (t), where the functions ϕ1 , ϕ2 , and ϕ3 are determined by the system of ﬁrst-order ordinary differential equations ϕ1 = ϕ1 F t, 2ϕ1 , 4ϕ1 ϕ3 − ϕ2 , 2 ϕ2 = ϕ2 F t, 2ϕ1 , 4ϕ1 ϕ3 − ϕ2 , 2 ϕ3 = ϕ3 F t, 2ϕ1 , 4ϕ1 ϕ3 − ϕ2 + G t, 2ϕ1 , 4ϕ1 ϕ3 − ϕ2 . 2 2 It follows from the ﬁrst two equations that ϕ2 = Cϕ1 , where C is an arbitrary constant.

© 2004 by Chapman & Hall/CRC

490

SECOND-ORDER EQUATIONS OF GENERAL FORM

2 8.1.3. Equations of the Form ∂w = F x, w, ∂w , ∂ w ∂t ∂x ∂x2

Preliminary remarks. Consider the equation ∂w ∂ 2 w ∂w = F x, w, , . ∂t ∂x ∂x2 Suppose that the auxiliary ordinary differential equation w = F (x, w, wx , wxx ) is reduced, by a linear transformation x = ϕ(z), w = ψ(z)u + χ(z) and the subsequent division of the resulting equation by ψ(z), to the autonomous form u = F(u, uz , uzz ), where F = F/ψ. Then, the original equation (1) can be reduced, by the same transformation x = ϕ(z), to the equation w(x, t) = ψ(z)u(z, t) + χ(z), (1)

∂u ∂ 2 u ∂u , = F u, , ∂t ∂z ∂z 2 which has a traveling-wave solution u = u(kz + λt). The above allows using various known transformations of ordinary differential equations (see Kamke, 1977, and Polyanin and Zaitsev, 2003) for constructing exact solutions to partial differential equations. If the original equation is linear, then such transformations will result in linear constantcoefﬁcient equations. 1. . ∂t ∂x2 Generalized separable solution: w(x, t) = Axt + Bt + C + ϕ(x), where A, B, and C are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation F x, ϕxx = Ax + B. ∂t ∂x ∂x2 Additive separable solution: 2. ∂w = F x, ∂w , ∂2w . w(x, t) = At + B + ϕ(x), ∂w = F x, ∂2w

where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation F x, ϕx , ϕxx = A. . ∂t ∂x ∂x ∂x2 1◦ . Suppose w(x, t) is a solution of this equation. Then the function 3. w1 = w(x + C1 e−at , t + C2 ), where C1 and C2 are arbitrary constants, is also a solution of the equation.

© 2004 by Chapman & Hall/CRC

∂w

= ax

∂w

+ F w,

∂w

,

∂2w

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

491

2◦ . Generalized traveling-wave solution: w = w(z), z = x + Ce−at ,

where C is an arbitrary constant and the function w(z) is determined by the ordinary differential equation F w, wz , wzz + azwz = 0. 4. ∂w ∂t =F ∂w ∂x ,x ∂2w ∂x2 .

**1◦ . Suppose w(x, t) is a solution of this equation. Then the function
**

−1 w1 = C1 w(C1 x, C1 t + C2 ) + C3 ,

where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. 2◦ . Additive separable solution: w(x, t) = At + B + ϕ(x), where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation F ϕx , xϕxx = A. 3◦ . Solution: w(x, t) = tΘ(ξ) + C, ξ = x/t, where C is an arbitrary constant, and the function Θ(ξ) is determined by the ordinary differential equation F Θξ , ξΘξξ + ξΘξ − Θ = 0. 5. ∂w ∂2w ∂w = F w, x , x2 . ∂t ∂x ∂x2

©

The substitution x = ez leads to the equation ∂w ∂ 2 w ∂w ∂w = F w, , − , ∂t ∂z ∂z 2 ∂z which has a traveling-wave solution w = w(kz + λt). 6. ∂w ∂t = xk F w, x ∂w ∂x , x2 ∂2w ∂x2 .

**1◦ . Suppose w(x, t) is a solution of this equation. Then the function
**

−k w1 = w(C1 x, C1 t + C2 ),

where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . Self-similar solution: w(x, t) = w(z), z = xt1/k ,

where the function w(z) is determined by the ordinary differential equation kz k−1 F w, zwz , z 2 wzz − wz = 0.

© 2004 by Chapman & Hall/CRC

492 7.

SECOND-ORDER EQUATIONS OF GENERAL FORM

∂w ∂w ∂2w ∂w + ax = xk F w, x , x2 . 2 ∂t ∂x ∂x ∂x Passing to the new independent variables z = xeat , we obtain an equation of the form 8.1.3.6: ∂w 2 ∂ 2 w ∂w = z k F w, z ,z . ∂τ ∂z ∂z 2 8. ∂w ∂ 2 w ∂w . = eλx F w, , ∂t ∂x ∂x2 1◦ . Suppose w(x, t) is a solution of this equation. Then the function w1 = w(x + C1 , e−λC1 t + C2 ), where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . Solution: w(x, t) = w(z), z = λx + ln t, τ= 1 1 − e−akt , ak

where the function w(z) is determined by the ordinary differential equation ez F w, λwz , λ2 wzz − wz = 0. . ∂t w ∂x w ∂x2 1◦ . Suppose w(x, t) is a solution of this equation. Then the function 9. w1 = C1 w(x, t + C2 ), where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . Multiplicative separable solution: w(x, t) = eλt ϕ(x), where λ is an arbitrary constant and the function ϕ(x) is determined by the ordinary differential equation F x, ϕx /ϕ, ϕxx /ϕ = λ. 10. ∂w = wβ F x, 1 ∂w , 1 ∂2w w ∂x2 . ∂w = wF x, 1 ∂w , 1 ∂2w

∂t w ∂x For β = 1, see equation 8.1.3.9.

**1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function
**

β−1 w1 = C1 w(x, C1 t + C2 ),

**where C1 and C2 are arbitrary constants, is also a solution of the equation. 2◦ . Multiplicative separable solution: w(x, t) = (1 − β)At + B
**

1 1−β

ϕ(x),

where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation ϕβ−1 F x, ϕx /ϕ, ϕxx /ϕ = A.

© 2004 by Chapman & Hall/CRC

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

493

∂w ∂ 2 w ∂w . = eβw F x, , ∂t ∂x ∂x2 1◦ . Suppose w(x, t) is a solution of this equation. Then the function 1 w1 = w(x, C1 t + C2 ) + ln C1 , β where C1 and C2 are arbitrary constants, is also a solution of the equation. 11. 2◦ . Additive separable solution: 1 ln(Aβt + B) + ϕ(x), β where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation eβϕ F x, ϕx , ϕxx + A = 0. w(x, t) = − . ∂t ∂x ∂x 1◦ . Suppose w(x, t) is a solution of this equation. Then the function w1 = C1 w(x, t + C2 ) + C3 , where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. F x, ∂x2 ∂w = ∂w ∂2w ∂w

12.

2◦ . Additive separable solution: w(x, t) = At + B + ϕ(x), where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation ϕx F x, ϕxx /ϕx = A. 3◦ . Generalized separable solution: w(x, t) = Aeµt Θ(x) + B where A, B, and µ are arbitrary constants, and the function Θ(x) is determined by the ordinary differential equation Θx F x, Θxx /Θx = µΘ. 13. ∂w ∂t ∂x ∂x2 For β = 1, see equation 8.1.3.12. = ∂w

β

F x,

∂2w

∂w ∂x

.

1◦ . Suppose w(x, t) is a solution of the equation in question. Then the function β−1 w1 = C1 w(x, C1 t + C2 ) + C3 , where C1 , C2 , and C3 are arbitrary constants, is also a solution of the equation. 2◦ . Additive separable solution: w(x, t) = At + B + ϕ(x), where A and B are arbitrary constants, and the function ϕ(x) is determined by the ordinary differential equation β ϕx F x, ϕxx /ϕx = A. 3◦ . Generalized separable solution: w(x, t) = A(1 − β)t + C1 1−β Θ(x) + B + C2 , where A, B, C1 , and C2 are arbitrary constants, and the function Θ(x) is determined by the ordinary differential equation β Θx F x, Θxx /Θx = AΘ + AB.

1

© 2004 by Chapman & Hall/CRC

494

SECOND-ORDER EQUATIONS OF GENERAL FORM

2 8.1.4. Equations of the Form ∂w = F x, t, w, ∂w , ∂ w ∂t ∂x ∂x2

1.

∂w

∂t ∂x The transformation

=a

∂w

m

∂2w ∂x2

n

+ f (t)x + g(t)

∂w ∂x

+ h(t)w. τ= F m+2n (t)H m+n−1 (t) dt,

w(x, t) = u(z, τ )H(t),

z = xF (t) +

g(t)F (t) dt,

**where the functions F (t) and H(t) are given by F (t) = exp leads to the simpler equation f (t) dt , H(t) = exp
**

m n

h(t) dt ,

∂2u ∂u ∂u . =a ∂τ ∂x ∂z 2 The last equation admits a traveling-wave solution, a self-similar solution, and a multiplicative separable solution. 2. ∂w ∂t ∂x2 The transformation z = xG(t) + = f (w) ∂2w

k

+ xg(t) + h(t) τ=

∂w ∂x

. G(t) = exp

k

h(t)G(t) dt,

G2k (t) dt,

g(t) dt ,

∂2w ∂w = f (w) . ∂τ ∂z 2 The last equation admits a traveling-wave solution and a self-similar solution. 3. ∂w ∂ 2 w ∂w = F x, t, , + aw. ∂t ∂x ∂x2 Suppose w(x, t) is a solution of this equation. Then the function w1 = w(x, t) + Ceat , where C are arbitrary constants, is also a solution of the equation. ∂w ∂ 2 w ∂w = F ax + bt, w, , . ∂t ∂x ∂x2 Solution: w = w(ξ), ξ = ax + bt, where the function w(ξ) is determined by the ordinary differential equation 4. F ξ, w, awξ , a2 wξξ − bwξ = 0. ∂t ∂x Passing to the new independent variables z = xG(t), τ= 5. ∂w = f (t)xk Φ w, x ∂w , x2 ∂2w ∂x2 + xg(t) ∂w ∂x .

leads to the simpler equation

f (t)G−k (t) dt,

G(t) = exp

g(t) dt ,

we obtain a simpler equation of the form 8.1.3.6: ∂w 2 ∂ 2 w ∂w = z k Φ w, z ,z . ∂τ ∂z ∂z 2

© 2004 by Chapman & Hall/CRC

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

495

f (x) ∂ 2 w ∂w . = wF t, ∂t w ∂x2 Multiplicative separable solution: 6. w(x, t) = ϕ(x) exp F (t, λ) dt ,

where the function ϕ = ϕ(x) satisﬁes the linear ordinary differential equation f (x)ϕ xx = λϕ. 7. ∂w ∂t w ∂x2 Generalized separable solution: w(x, t) = eλx E(t) A + = wΦ t, 1 ∂2w + f (t)eλx . f (t) dt + Be−λx E(t), E(t)

E(t) = exp

Φ(t, λ2 ) dt ,

where A, B, and λ are arbitrary constants. 8. ∂w ∂t w ∂x2 Generalized separable solution: = wΦ t, 1 ∂2w + f (t)eλx + g(t)e–λx . f (t) dt + e−λx E(t) B + E(t) Φ(t, λ2 ) dt ,

w(x, t) = eλx E(t) A +

g(t) dt , E(t)

E(t) = exp where A, B, and λ are arbitrary constants.

∂t w ∂x2 w ∂x2 There is a generalized separable solution of the form 1 ∂2w

9.

∂w

= wF1 t,

1 ∂2w

+ eλx F2 t,

1 ∂2w

+ e–λx F3 t,

1 ∂2w w ∂x2

.

w(x, t) = eλx ϕ(t) + e−λx ψ(t). 10. ∂w ∂t w ∂x2 Generalized separable solution: = wΦ t, + f (t) cosh(λx) + g(t) sinh(λx). f (t) dt + sinh(λx)E(t) B + E(t) Φ(t, λ2 ) dt ,

w(x, t) = cosh(λx)E(t) A +

g(t) dt , E(t)

E(t) = exp where A, B, and λ are arbitrary constants. 11. ∂w ∂t w ∂x2 Generalized separable solution: = wΦ t, 1 ∂2w + f (t) cos(λx).

**w(x, t) = cos(λx)E(t) A + E(t) = exp where A, B, and λ are arbitrary constants.
**

© 2004 by Chapman & Hall/CRC

f (t) dt + B sin(λx)E(t), E(t) Φ(t, −λ2 ) dt ,

496 12.

SECOND-ORDER EQUATIONS OF GENERAL FORM

1 ∂2w ∂w + f (t) cos(λx) + g(t) sin(λx). = wΦ t, ∂t w ∂x2 Generalized separable solution: w(x, t) = cos(λx)E(t) A + f (t) dt + sin(λx)E(t) B + E(t) Φ(t, −λ2 ) dt , g(t) dt , E(t)

E(t) = exp where A, B, and λ are arbitrary constants.

∂t w ∂x2 w ∂x2 There is a generalized separable solution of the form

13.

∂w

= wF1 t,

1 ∂2w

+ cos(λx)F2 t,

1 ∂2w

+ sin(λx)F3 t,

1 ∂2w w ∂x2

.

w(x, t) = cos(λx)ϕ(t) + sin(λx)ψ(t). 14. ∂w ∂t w ∂x w ∂x2 Multiplicative separable solution: w(x, t) = eλx E(t) A + = wΦ t, 1 ∂w , 1 ∂2w + f (t)eλx .

f (t) dt , E(t)

E(t) = exp

Φ(t, λ, λ2 ) dt ,

where A, B, and λ are arbitrary constants. 15. ∂w = f (t)wβ Φ x, 1 ∂w w ∂x , 1 ∂2w w ∂x2 + g(t)w.

∂t The transformation

w(x, t) = G(t)u(x, τ ),

τ=

f (t)Gβ−1 (t) dt,

G(t) = exp

g(t) dt ,

**leads to a simpler equation of the form 8.1.3.10: 1 ∂u 1 ∂ 2 u ∂u = uβ Φ x, , , ∂τ u ∂x u ∂x2 which has a multiplicative separable solution u = ϕ(x)ψ(τ ). 16. ∂w = f (t) ∂w
**

k

∂t ∂x Generalized separable solution:

Φ x,

∂2w ∂x2

∂w ∂x

+ g(t)w + h(t).

w(x, t) = ϕ(t)Θ(x) + ψ(t), where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations ϕt = Af (t)ϕk + g(t)ϕ, ψt = g(t)ψ + Bf (t)ϕk + h(t), (1) (2)

**C is an arbitrary constant, and the function Θ(x) is determined by the second-order ordinary differential equation k (3) Θx Φ x, Θxx /Θx = AΘ + B.
**

© 2004 by Chapman & Hall/CRC

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

497

**The general solution of system (1), (2) is expressed as ϕ(t) = G(t) C − kA ψ(t) = DG(t) + G(t) f (t)Gk−1 (t) dt
**

1 1−k

,

G(t) = exp dt , G(t)

g(t) dt ,

Bf (t)ϕk (t) + h(t)

where A, B, C, and D are arbitrary constants. For k = 1 and Φ(x, y) = Φ(y), a solution of equation (3) is given by Θ(x) = αeλx − B/A, where α is an arbitrary constant, and λ is found from the algebraic (or transcendental) equation λΦ(λ) = A. 17. ∂w = f1 (t)w + f0 (t) ∂w

k

∂t ∂x Generalized separable solution:

Φ x,

∂2w ∂x2

∂w ∂x

+ g1 (t)w + g0 (t).

w(x, t) = ϕ(t)Θ(x) + ψ(t), where the functions ϕ(t) and ψ(t) are determined by the system of ﬁrst-order ordinary differential equations (C is an arbitrary constant): ϕt = Cf1 (t)ϕk+1 + g1 (t)ϕ, ψt = Cf1 (t)ϕ + g1 (t) ψ + Cf0 (t)ϕ + g0 (t),

k k

(1) (2) (3)

**and the function Θ(x) is determined by the second-order ordinary differential equation Θx
**

k

Φ x, Θxx /Θx = C.

−1/k

The general solution of system (1), (2) is expressed as ϕ(t) = G(t) A − kC ψ(t) = Bϕ(t) + ϕ(t) f1 (t)Gk (t) dt , G(t) = exp dt , ϕ(t) g1 (t) dt ,

Cf0 (t)ϕk (t) + g0 (t)

where A, B, and C are arbitrary constants. Further, we assume that Φ is independent of x explicitly, i.e., Φ(x, y) = Φ(y). For Φ(0) ≠ 0 and Φ(0) ≠ ∞, particular solution to equation (3) has the form Θ(x) = αx + β, where α k Φ(0) = C and β is an arbitrary constant. For k = 0, the general solution of equation (3) is expressed as Θ(x) = αeλx + β, where α and β are arbitrary constants, and λ is determined from the algebraic (transcendental) equation Φ(λ) = C. 18. ∂w ∂ 2 w ∂w = f (t)eβw Φ x, , ∂t ∂x ∂x2 The transformation w(x, t) = u(x, τ ) + G(t), τ= + g(t).

f (t) exp βG(t) dt,

G(t) =

g(t) dt

**leads to a simpler equation of the form 8.1.3.11: ∂u ∂ 2 u ∂u = eβu Φ x, , , ∂τ ∂x ∂x2 which has an additive separable solution u = ϕ(x) + ψ(τ ).
**

© 2004 by Chapman & Hall/CRC

498 19.

SECOND-ORDER EQUATIONS OF GENERAL FORM

∂w ∂ 2 w ∂w ∂2w ∂2w ∂w , , 2w – 2x . = wF t, –x + x2 ∂t ∂x2 ∂x ∂x2 ∂x ∂x2 Multiplicative separable solution: w(x, t) = (C2 x2 + C1 x + C0 )ϕ(t), where C0 , C1 , and C2 are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ordinary differential equation ϕt = ϕF t, 2C2 ϕ, C1 ϕ, 2C0 ϕ .

¢

Reference: Ph. W. Doyle (1996), the case ∂t F ≡ 0 was treated.

∂t ∂x Additive separable solution:

20.

∂w

= F x, t, w,

∂w

G x,

∂w ∂x

,

∂2w ∂x2

+ h(t).

w(x, t) = ϕ(x) +

h(t) dt,

where the function ϕ(x) is determined by the ordinary differential equation G x, ϕx , ϕxx = 0. 21. ∂w = F x, t, w, ∂w G x, 1 ∂w w ∂x , 1 ∂2w w ∂x2 + h(t)w.

∂t ∂x Multiplicative separable solution:

w(x, t) = C exp

h(t) dt ϕ(x),

**where the function ϕ(x) is determined by the ordinary differential equation G x, ϕx /ϕ, ϕxx /ϕ = 0. 22. ∂2w ∂w = g0 (t)F0 ∂t ∂x2 + h(t) + xg1 (t)F1 ∂w
**

2

∂2w ∂x2

+ x2 g2 (t)F2 ∂w ∂x

∂2w ∂x2

∂x There is a generalized separable solution of the form

w(x, t) = x2 ϕ(t) + xψ(t) + χ(t). 23. ∂2w ∂2w ∂w = x2 f2 t, + xf1 t, ∂t ∂x2 ∂x2 Generalized separable solution quadratic in x: w(x, t) = x2 ϕ(t) + x + f0 t, ∂2w . ∂x2

f1 (t, 2ϕ) dt +

f0 (t, 2ϕ) dt + C1 x + C2 ,

where C1 and C2 are arbitrary constants, and the function ϕ = ϕ(t) is determined by the ﬁrst-order ordinary differential equation ϕt = f2 (t, 2ϕ). ∂t ∂x2 ∂x2 There is a generalized separable solution of the form 24. ∂w = x2 f2 t, ∂2w + xf1 t, ∂2w + f0 t, ∂2w ∂x2 + g(t)w.

w(x, t) = x2 ϕ(t) + xψ(t) + χ(t).

© 2004 by Chapman & Hall/CRC

+ p0 (t) + xp1 (t)

+ q(t)w +

0 (t)

+ x 1 (t) + x2 2 (t).

8.1. EQUATIONS INVOLVING THE FIRST DERIVATIVE IN t

499

2 8.1.5. Equations of the Form F x, t, w, ∂w , ∂w , ∂ w ∂t ∂x ∂x2

=0

1. F at + bx, w, Solution:

∂w ∂t

,

∂w ∂x

,

∂2w ∂x2

= 0.

w = w(ξ), ξ = at + bx, where the function w(ξ) is determined by the ordinary differential equation F ξ, w, awξ , bwξ , b2 wξξ = 0. 2. F t, = 0. w ∂t w ∂x2 1◦ . Multiplicative separable solution: w(x, t) = A cosh(λx) + B sinh(λx) ϕ(t), where A, B, and λ are arbitrary constants, and the function ϕ(t) is determined by the ﬁrst-order ordinary differential equation F t, ϕt /ϕ, λ2 = 0. 2◦ . Multiplicative separable solution: w(x, t) = A cos(λx) + B sin(λx) ϕ(t), where A, B, and λ are arbitrary constants, and the function ϕ(t) is determined by the ﬁrst-order ordinary differential equation F t, ϕt /ϕ, −λ2 = 0. w ∂t w ∂x w ∂x2 Multiplicative separable solution: 3. F t, 1 ∂w , 1 ∂w , 1 ∂2w = 0. 1 ∂w , 1 ∂2w

w(x, t) = Aeλx ϕ(t), where A and λ are arbitrary constants, and the function ϕ(t) is determined by the ﬁrst-order ordinary differential equation F t, ϕt /ϕ, λ, λ2 = 0. w ∂t w ∂x w ∂x2 Multiplicative separable solution: 4. F x, 1 ∂w , 1 ∂w , 1 ∂2w = 0.

w(x, t) = Aeλt ϕ(x), where A and λ are arbitrary constants, and the function ϕ(x) is determined by the second-order ordinary differential equation F x, λ, ϕx /ϕ, ϕxx /ϕ = 0. 5. F1 t, ∂w ∂t Additive separable solution: + F2 x, ∂w ∂ 2 w , ∂x ∂x2 = kw.

w(x, t) = ϕ(t) + ψ(x), where the functions ϕ(x) and ψ(x) are determined by the ﬁrst- and second-order ordinary differential equations F1 t, ϕt − kϕ = C, F2 x, ψx , ψxx − kψ = −C, and C is an arbitrary constant.

© 2004 by Chapman & Hall/CRC

500 6. F1 t,

SECOND-ORDER EQUATIONS OF GENERAL FORM

1 ∂w 1 ∂ 2 w 1 ∂w + wk F2 x, = 0. , w ∂t w ∂x w ∂x2 Multiplicative separable solution: w(x, t) = ϕ(t)ψ(x), where the functions ϕ(t) and ψ(x) are determined by the ﬁrst- and second-order ordinary differential equations ϕ−k F1 t, ϕ