This action might not be possible to undo. Are you sure you want to continue?

LECTURE NOTES IN PARTIAL DIFFERENTIAL EQUATIONS Fourth Edition, February 2011 by ˇ Tadeusz STYS

University of Botswana

ii

Contents

1 Solution of Partial Diﬀerential Equations 1.1 The General Solution of PDE . . . . . . . . . . . . . . . . . . 1.2 First Order PDE with Constant Coeﬃcients . . . . . . . . . . 1.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Classiﬁcation of Partial Diﬀerential Equations of the Second Order 2.1 Hyperbolic, Elliptic and Parabolic Equations . . . . . . . . . . 2.2 The Standard Form of Hyperbolic, Elliptic and Parabolic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Transformation of a Hyperbolic Equation into a Standard Form 2.4 Transformation of an Elliptic Equation into the Standard Form 2.5 Transformation of a Parabolic Equation into the Standard Form 2.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Hyperbolic Equations 3.1 The Initial Value Problem for Wave Equation . . . . . . . . . 3.1.1 D’Alembert’s Solution . . . . . . . . . . . . . . . . . . 3.1.2 The Initial Boundary Value Problem for Wave Equation 3.2 Solution to the Finite Vibrating String by Separation of Variables 3.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Parabolic Equations 4.1 Initial Boundary Value Problem . . . . . . . . . . . . . . . . . 4.2 Solution by Separation of Variables . . . . . . . . . . . . . . . 4.3 Transformation of Non-homogeneous Boundary Conditions to Homogeneous . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 More Developed Heat Equation . . . . . . . . . . . . . . . . . 4.5 Non-homogeneous Heat Equation . . . . . . . . . . . . . . . . 4.6 Fundamental Solution for the Heat Equation . . . . . . . . . . 4.7 Fundamental Formulae . . . . . . . . . . . . . . . . . . . . . . 4.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iii

1 1 3 7 9 9 11 12 18 21 25 27 27 27 31 31 40 43 43 48 50 53 54 57 58 60

. . .5 The Maximum Principle for Helmholz-Poisson Equation . . 83 5. . . 80 5.3 The Maximum Principle for Laplace Equation . . . . .iv 5 Elliptic Equations 63 5. .4 The Maximum Principle for Poisson Equation . . . .3 Green’s Function . . . . . .7. . . . . . . . . . . 73 5. . .1 Fundamental Solution of Laplace Equation .2 Boundary Value Problems for Laplace Equation . . 76 5.1 Laplace Equation . . . .2 Theorem on representation of harmonic functions . . . . . . 70 5. . . . . . . . . 65 5. . . . . . 89 . . . . . 66 5. . . . . .7 Boundary Value Problem for Laplace’s Equation in a Disk . . . .9 Exercises . . . . . . . . . . . . 88 5. . . . . 63 5. . . . . . . . .7. . . . .7. . .8 Helmholz Equation and Eigenvalue Problem. . . . . . . 78 5. . . . . . . . . . . . . . .6 Boundary Value Problem for Laplace’s Equation in a Rectangle 74 5. . . . .

It is assumed that the students have basic knowledge in Real Analysis. The notes have been used for teaching the course MAT426 (PDE). University of Botswana.v PREFACE These lecture notes are designed for undergraduate students as a complementary reading text to an introductory course on Partial Diﬀerential Equations. Partial Diﬀerential Equations at the Faculty of Science. University of Botswana Department of Mathematics. February 2011 ˇ Tadeusz STYS .

let us integrate both sides of the equation (1. z < ∞ Solution. z < ∞ 1 . y.1 The General Solution of PDE The general solution of a partial diﬀerential equation (PDE) is considered as a collection of all possible solutions of a given equation.2) in following form: u(x.2) with respect the variable y.1) with respect to the variable x. Similarly. y) = x2 + y x + f(y) (1. y < ∞ (1. and let us integrate both sides of equation (1. Example 1. we obtain u(x.5) (1. z) = x + 2 y + z. Then. y.3) for arbitrary diﬀerentiable function f(y). z). where f(y) is any diﬀerentiable function. −∞ < x. we obtain all solutions of equation (1. z).Chapter 1 Solution of Partial Diﬀerential Equations 1. −∞ < x. −∞ < x.1 Consider the following ﬁrst order linear PDE equations ux (x. Example 1. Let the variable y in (1. y) = 2 x + y. y. when the variables x and z are ﬁxed.2 Consider the following second and third order linear PDE equations uxy (x. y.1) are of the form (1. Then.1) (1. y) = 2 x + y. −∞ < x.3). Thus.1) be ﬁxed.4) uxyz (x. z) = xy + y 2 + z y + g(x.6) (1. y. y < ∞ (1.2) uy (x. for arbitrary diﬀerentiable function g(x. all solutions of equation (1. z) = x + 2 y + z.

z) 2 2 (1. when the variables y and z are ﬁxed. y) is an antiderivative to f(x.7) with respect to the variable x. we obtain 1 uxy (x. Then. Example 1. y.5) with respect to the variable y.13) . y) = x2y + y 2 x + F (x) + g(y) 2 (1.9) with respect to the variable y. where F (x) is an antiderivative to the arbitrary function f(x). we obtain 1 ux(x. y. all solutions of equation (1. y) with respect to the variable x.8). y) = 2 x y + y 2 + f(x) 2 (1. We note that uxx = 0 and uyy = 0 . we obtain 1 u(x.5) are of the form (1.7) for arbitrary diﬀerentiable function f(x). y) is an antiderivative to the antiderivative F (x. we obtain 1 1 u(x. z) 2 (1. when the variable x is ﬁxed. Finally. z) with respect to the variable x. −∞ < x. z) is an arbitrary diﬀerentiable function of the variables x and y.10) with respect to the variable x. z) = x y z + y 2 z + z 2 y + F (x. when the variable y is ﬁxed. y) with respect to the variable y. we obtain 1 ux(x. let us integrate both sides of equation (1. when the variables x and y are ﬁxed. Then. we integrate equation (1. y. y) + g(x. when the variables x and z are ﬁxed. z) is an antiderivative to g(x.12) Solution. y) = f(y)x + g(y) (1. Thus. y) + G(x.11) where F F (x. Let us integrate both sides of equation (1. we integrate equation (1. Then. y < ∞ (1.3 Find all solutions of the following non-linear PDE equation: (uxx)2 + (uyy )2 = 0. we integrate equation (1.2 Solution. and g(x.10) where F (x. Then.9) Next. Next.8) where F (x) is an antiderivative to the function f(x) and g(y) is an arbitrary diﬀerentiable function of the variable y. z) = x z + 2y z + z 2 + f(x. Then. z) = x2 y z + x y 2 z + x yz 2 + F F (x. Now. and G(x. y) 2 (1. and g(y) is any diﬀerentiable function of the variable y. All solutions of the equation uxx = 0 are in the form u(x.6) with respect to the variable z.

16). y) = y 2. Let us note that the solution u(x. y) = f(2x + y) + 2 x f (2x + y). By diﬀerentiation. f(2 ∗ 1 + y) = y 2. y) = 1. b. y < ∞.3 and all solutions of the equation uyy = 0 are in the form u(x. q(x) and r(x). we shall transform the equation a ux + b uy + c u = f(x. b. c and d. we ﬁnd function f. Example 1. Let us consider the following equation a ux + b uy + c u = f(x. we compute x ux−2 x uy = x f(2x+y)+2 x2 f (2x+y)−2 x2 f (2x+y) = x f(2x+y) = u(x. Show that u(x. Then. −∞ < x.13) and (1. we ﬁnd ux (x.15) when f is any diﬀerentiable function.12) have common part which is included in both (1. uy (x. y). (1. Now. (1. and f(x.16). We can choose f(2x + y) = (2x + y − 2)2 . y) = x f (2x + y). g(y). f(t) = (t − 2)2 . u(1.17) where a. −∞ < y < ∞. and c are constant coeﬃcients.16) is the solution of equation (1. Then. for arbitrary constants a. the solutions which have both forms are in the following form: u(x. all solutions of equation (1. y) = a x y + b x + c y + d. applying the condition u(1.15) Solution. y) = x(2x + y − 2)2 satisﬁes the condition u(1. which satisﬁes the condition u(1. Hence. y). y). Thus. y) is a given continuous function. y) = q(x)y + r(x). y) = xf(2 x + y) (1. . Then. Let us consider the case when b = 0. 1. (1.14).2 First Order PDE with Constant Coeﬃcients a2 + b2 > 0. y) = y 2 to the solution (1. y) = y 2. So that.4 Consider the following ﬁrst order PDE equation x ux − 2 x uy = u. Find the solution within the family of solutions (1. Let t = 2 + y and y = t − 2. Let us note that f(2x + y) is the function of one variable t = 2x + y.14) for arbitrary diﬀerentiable functions f(y). −∞ < t < ∞.

19). Thus. z)e b . z) = e b [ b − cz g(w. for g(w. z) b In order to ﬁnd the general solution of equation (1.19) cz 1 v(w.20) ∂z b Integrating both sides of equation (1. we divide the above cz equation by b and multiplying by the factor e b . (1. (1. y). b b cz cz ∂ 1 [e b v(w. we obtain the following general solution of equation (1. we ﬁnd w+az . z).21) or where C(w) is an arbitrary diﬀerentiable function of the variable z. y = z.18) w + az . z = y Hence.4 given in x. equation (1. in the new variables. b In terms of the new coordinates. z)] = g(w. z w = b x − a y. and multiplying by cz − the factor e b . y coordinates to the equation b vz + c v = f( in the new coordinates w. z) = g(w. z)e b . we compute x= a ux + b uy = a(vw wx + vz zx ) + b(vw wy + vz zy ) = (a b − b a)vw + b vz = b vz . we shall solve the equation b vz + c v = g(w. In the case when b = 0. takes the form b vz + c v = f( Now. z) b w+az . z) b (1. z) = f( (1. we have already the equation in the form (1. to obtain cz cz cz c 1 e b vz (w.19). z)e b dz + C(w)]. so that aux + c u = f(x. . z) + e b v(w.20) with respect to z.19) w+a z .17).

z) = 1. we have −e−z vz (w. z) + v(w. z) = 1. y) = v(b x−a y. Below. ∂z . z). z). we can solve equation (1. (1a) Find all solutions of the equation ∂u ∂u − +u=1 ∂x ∂y (1b) Find the solution of equation (1. y) = 2 for y=x (1.5 . by the mapping So that we have w = − x − y. we compute the expression ux − uy = (vw wx + wz zx) − (vw wy + wz zy ) = −vz . z) = f(−w − z. z). z=y y = z. First. y) = u( w+a z . therefore.19) by formula (1. Since g(w. x = −w − z.21) to get the solution v(w. z) = e−z or − ∂ −z [e v] = e−z . we obtain the equation −vz (w. z) = u(x. Example 1. and then to obtain the solution u(x. for which.5 The new function v(w. We consider the new unknown v(w. Multiplying the above equation by e−z . y). c = 1 and the function f(x. z) ≡ u(x. z). we transfer the equation to the form vz + c v = g(w.22) Solution 1a. b Now. y) = u(−w − z. we shall present some examples following the above solution of the ﬁrst order linear equation with constant coeﬃcients. y) = 1. Note that the coeﬃcients a = 1. z) + e−z v(w. b = −1.22) which satisﬁes the condition u(x.

25) . we compute 3 ux − 2 uy = 3(vw wx + vz zx) − 2(vw wy + vz zy ) = −2vz . 2 z − Dividing by −2 and multiplying by the factor e 2 . Solution. (1. We consider the new variables w = 2 x + 3 y.6 By integration with z −e−z v = −e−z + C(w) or Hence. y) = 1 − C(−x − y)ey . So that C(−2x) = −e−x t Let t = −2x. z) = u(−w − z. ∂z 4 (1.23) w − 3z . in the new variables becomes 1 −2vz + v = (w − 3 z). we ﬁnd u(x. z). we introduce the unknown x= v(w. z) = u(x. Then. Solution 1b. x) = 2. y = z. 2 (1. we obtain z z ∂ − 1 − [e 2 v(w. y) = 1 + e 2 e = 1 + e 2 satisﬁes the condition u(x. v(w.6 Find the general solution of the equation 3 ux − 2 uy + u = x. z) = 1 − C(w) ez and u(x. we ﬁnd the solution v(w. For y = x. z)] = − e 2 (w − 3 z).23). x) = 1 − C(−2x)ex = 2.24) z = y. y) = u( Now. Equation (1. we have C(t) = −e 2 and the solution y−x −x − y y u(x. 2 Then. Example 1. Hence w−3z . z) = 1 − C(w)ez for arbitrary diﬀerentiable function C(w).

(b) uxy = x y.7 Integrating both sides of (1. 2 and coming back to the original variables. we obtain a particular solution. 1. Indeed. we ﬁnd z 1 e 2 v(w. 2 2 − where C(w) is an arbitrary diﬀerentiable function of the variable w. we have y y 3ux − 2uy + u = 3 + e 2 + x − 3 + e 2 = x. we have the particular solution y u(x. y) = [2x + 3y − 3y − 6] + e 2 C(2x + 3y) = x − 3 + e 2 C(2x + 3y).26) 2 4 z − w 3z = e 2[ − − 3] + C(w).25) with respect z. For example. . Find the general solution of the equations (a) ux = 3x + 2y. the particular solution is y u(x. Hence. we obtain the general solution of equation (1. (c) uxyz = x + y + z. for C(2x + 3y) = 1. z) = [w − 3 z − 6] + e 2 C(w).23) in the following form y y 1 u(x. z) = − w 4 − z z − 3 e 2 dz + ze 2 dz + C(w) 4 z z z − 3 − 1 − = we 2 + [ze 2 (−2) − e 2 (−2) dz] + C(w) (1. y) = x − 3 + e 2 . Also. y) = x − 3 + e 2 (2x + 3y).3 Exercises Question 1. we ﬁnd the solution z 1 v(w. for C(2x + 3y) = 2x + 3y. 2 Let us observe that choosing the function C(2x + 3y). when w is ﬁxed.

y) = 2 for −∞<y <∞ for −∞<x<∞ . 1.8 Question 2. Find the solution u(x. Find the general solution of the equation ux + 4uy + 2u = 5 2. (b) ux + 2uy + 3u = x + y. (c) ux − uy + u = 0. 1. y) of the equation which satisﬁes the condition u(1. 1) = x Question 4. Find the solution of the equation which satisﬁes the condition u(x. Find all solutions of the equations (a) ux − 2uy + u = x + y. Question 3. Find all solutions of the equation ux + uy − 2u = y 2.

All the equations of the general form (2. y) in the domain Ω. b(x. y). y) + g(x. y)u = f(x. Classiﬁcation. y). y) and the right side f(x.1) L0 u = a(x. y) 2 + ∂x2 ∂x∂y ∂y ∂u ∂u +d(x. y) (2. y). Elliptic and Parabolic Equations ∂ 2u ∂ 2u ∂ 2u + 2b(x. g(x. The type of an equation is determined by the operator L0 of the sedcond order. e(x. y). y) + c(x. y) of order one does not eﬀect the type of the equation.Chapter 2 Classiﬁcation of Partial Diﬀerential Equations of the Second Order 2. y) are given functions of the variables (x.1) are divided in three the following classes pending on the sign of the discriminant b2 − a c. d(x. 9 .1) where u(x. y) 2 ∂x2 ∂x∂y ∂y ∂u ∂u + e(x. c(x. y)u. y) + c(x. ∂x ∂y We note that the diﬀerential operator L1 u = d(x. y) + e(x. For the classiﬁcation purpose.1 Hyperbolic. y). y) ∂ 2u ∂ 2u ∂ 2u + 2b(x. ∂x py We shall consider the following form of partial diﬀerential equations: Lu ≡ a(x. y) is an unknown function and the coeﬃcients a(x. y) + g(x. we consider the following diﬀerential operator of the second order associated with the main part of equation (2. y).

(2. 1. Also. since the discriminant b2 − a c = 02 − 1 ∗ 0 = 0. ∂x2 ∂y 2 b2 − a c < 0.1) is called hyperbolic equation if the discriminant b2 − a c > 0. 2. for is the elliptic equation. (2.1) is called elliptic equation if the discriminant all (x.10 1. The wave equation ∂ 2u ∂ 2 u − 2 = 0. y) ∈ Ω. 2.1 . . 3. • The heat equation ∂u ∂ 2u − 2 = 0. since the discriminant b2 − a c = 02 − 1 ∗ 1 = −1 < 0. (2. for b2 − a c = 0. the operator L is called elliptic operator if the discriminant b2 − a c < 0. y) ∈ Ω. the operator L is called parabolic operator if the discriminant b2 − a c = 0. the operator L is called hyperbolic operator if the discriminant b2 − a c > 0.1) is called parabolic equation if the discriminant all (x. Example 2. ∂t ∂x is the parabolic equation. since the discriminant b2 − a c = 02 − 1(−1) = 1 > 0. 3. y) ∈ Ω. • Laplace’s equation ∂ 2u ∂ 2u + = 0. for all (x. ∂t2 ∂y is the hyperbolic equation.

x. u. x.11 2. ∂t∂x ∂t ∂x 2. ). ∂t2 ∂x ∂t ∂x 3. ). ∂t ∂x ∂x Here. . x. u.2 The Standard Form of Hyperbolic. x. η = η(x. For the composed function u(ξ. we compute the following derivatives: ∂u ∂u ∂ξ ∂u ∂η = + . In order to transform equation (2. the standard form of a parabolic equation ∂u ∂ 2u ∂u = k 2 2 + f ∗ (t. u. y).1) into its standard form. u. ). . y)). . the second standard form of a hyperbolic equation ∂ 2u ∂ 2u ∂u ∂u − 2 = f ∗ (t. η(x. we consider the new variables ξ = ξ(x. elliptic and parabolic equations are considered: 1. η) = u(ξ(x. 2 ∂x ∂y ∂x ∂y 4. y). Elliptic and Parabolic Equations The following standard or canonical forms of hyperbolic. f ∗ is a function independent of the second derivatives. The ﬁrst standard form of a hyperbolic equation ∂ 2u ∂u ∂u = f ∗ (t. ). ∂x ∂ξ ∂x ∂η ∂x and ∂ 2u ∂ 2u ∂ξ 2 ∂ 2u ∂ξ ∂η ∂ 2u ∂η 2 ∂u ∂ 2ξ ∂u ∂ 2η = 2( ) +2 + ( ) + + ∂x2 ∂ξ ∂x ∂ξ∂η ∂x ∂x ∂η 2 ∂x ∂ξ ∂x2 ∂η ∂x2 ∂ 2u ∂ 2u ∂ξ ∂ξ ∂ 2u ∂ξ ∂η ∂ 2u ∂ξ ∂η ∂ 2u ∂η ∂η = 2 + + + + ∂x∂y ∂ξ ∂x ∂y ∂ξ∂η ∂x ∂y ∂ξ∂η ∂y ∂x ∂η 2 ∂x ∂y ∂u ∂ 2ξ ∂u ∂ 2η + + ∂ξ ∂x∂y ∂η ∂x∂y ∂ 2u ∂ξ 2 ∂ 2u ∂ξ ∂η ∂ 2u ∂η 2 ∂u ∂ 2ξ ∂u ∂ 2η ∂ 2u = 2( ) +2 + ( ) + + ∂y 2 ∂ξ ∂y ∂ξ∂η ∂y ∂y ∂η 2 ∂y ∂ξ ∂y 2 ∂η ∂y 2 ∂u ∂u ∂ξ ∂u ∂η = + ∂y ∂ξ ∂y ∂η ∂y . y). the standard form of an elliptic equation ∂ 2 u ∂ 2u ∂u ∂u + 2 = f ∗ (t.

η. ∂η ∂η ∂x ∂y ∂y ∂x ∂x ∂y is Jacobian of the mapping ξ = ξ(x. (2. ∂ξ∂η ∂ 2u ∂ 2 u − 2 = f ∗ (t.1). y). substituting the above relationship to equation (2. 2. provided that Jacobian J (ξ.1) can be transformed by a transformation ξ = ξ(x.4) to a standard form.2) ∂ξ 2 ∂ξ ∂ξ ∂ξ ) + 2b + c( )2 ∂x ∂x ∂y ∂y ∂ξ ∂η ∂ξ ∂η ∂ξ ∂η ∂ξ ∂η + b( + )+c B=a ∂x ∂x ∂x ∂y ∂y ∂x ∂y ∂y ∂η 2 ∂η ∂η ∂η 2 C = a( ) + 2b + c( ) ∂x ∂x ∂y ∂y 2 2 2 ∂ ξ ∂ ξ ∂ ξ ∂ξ ∂ξ D = a 2 + 2b +c 2 +d +e ∂x ∂x∂y ∂y ∂x ∂y 2 2 2 ∂ η ∂ η ∂ η ∂η ∂η E = a 2 + 2b +c 2 +d +e ∂x ∂x∂y ∂y ∂x ∂y G = g.12 Now. ∂x ∂y ∂y ∂x where ∂ξ ∂ξ ∂ξ ∂η ∂ξ ∂η ∂x ∂y J (ξ. η) = = − = 0. η = η(x. η) = 0. u). every equation of the general form (2. y).3 Transformation of a Hyperbolic Equation into a Standard Form ∂ 2u = f ∗ (t. It may be veriﬁed that ∂ξ ∂η ∂ξ ∂η 2 B 2 − A C = (b2 − a c)( − ) . y).3) Thus. 2 ∂t ∂x The standard forms of a hyperbolic equation in the two variables ξ. we obtain the following equation in terms of the variables ξ and η: A where ∂ 2u ∂ 2u ∂u ∂u ∂ 2u + 2B +C 2 +D +E + Gu = F 2 ∂ξ ∂ξ∂η ∂η ∂ξ ∂η A = a( (2. u). ξ. (2. η are: or . η = η(x. ξ. We note that. the type of the equation remains the same in the new coordinates ξ and η. η. F = f. y).

5) given in the implicit form A = a( ξ(x. y) . y)2 − a(x. ∂ξ 2 ∂η To obtain a standard form put in (2. ηy . along of the characteristic curves the diﬀerentials d ξ = 0 and d η = 0. that is the quadratic equation dx aλ2 − 2b λ + c = 0.7) dy ηx =− . y) (2. ∂x ∂y (2. we dx obtain the following ordinary diﬀerential equation dy dy a( )2 − 2 b + c = 0.8) for λ = . solving equation (2. we ﬁnd the two ordinary diﬀerential equadx tions dy ξx =− .5).5) ∂η 2 ∂η ∂η ∂η 2 C = a( ) + 2b + c( ) = 0. y) = constant Then. dx ξy (2. ∂x ∂y dy Solving the above equations for .5) as characteristics of equation (2. dx ηy dy Substituting relation (2. ∂x ∂x ∂y ∂y (2.13 The wave equations as the representatives ∂ 2u =0 ∂ξ∂η or ∂ 2u ∂ 2 u − 2 = 0. y)c(x. we obtain the following two ordinary diﬀerential equations for characteristics curves: b(x.1). y) − b(x. y)2 − a(x. ξy . ∂x ∂x ∂y ∂y We shall call the curves determined by equations (2.8) dx dx dy Again. y)c(x.6) ∂η ∂η dx + dy = 0. y) = constant and η(x. ηx. to equation (2. y) + b(x. These characteristic curves are solutions of equations (2.2) the conditions: ∂ξ ∂ξ ∂ξ ∂ξ 2 ) + 2b + c( )2 = 0. y) dy = λ1 = dx a(x. y) dy = λ2 = dx a(x. (2.9) b(x.7) between and ξx . that is ∂ξ ∂ξ dx + dy = 0.

The equation for characteristics is (see formula ( 2.10) (2. aλ2 − 2bλ + c = 0. d = 0. we ﬁnd the two equations dx √ dx b + b2 − ac = λ2 = = −3 dt a 1 4 . η. ξy ηx = −λ2 ηy (2. e = 0. we apply the mapping ξ = x − 2t. η = x + 3t . Solution To (a).25 > 0 is positive.13) Hence. we ﬁnd the solutions x = 2t + constant.8)): ( dx 2 dx ) + −6=0 dt dt dy .12) +6 = Solving the above equation for λ = dx b− = λ1 = dt √ b2 − ac = 2.1)) The equation is hyperbolic since the discriminant ∆ = b2 − ac = 6. a (2. c = −6.2 . f = 2 2 (see formula ( 2. The coeﬃcients of the equation are: 1 a = 1. b = − . (a) Find the characteristics for the equation: utt − utx − 6uxx − 5u = 2.12) to its standard form in terms of the variables ξ. (b) Transform equation (2. x = −3t + constant In order to transform equation (2. (2.11) Example 2.14 where λ1 and λ2 are roots of the quadratic equation.12) into its standard form. g = −5. and ξx = −λ1 .

we obtain the second standard from ∂ 2u ∂ 2u 1 − = (5u + 2) 2 2 ∂α ∂β 13 Example 2.15) (b) Transform equation (2. ∂t ∂t ∂x ∂x D = 0.15 Evaluate the coeﬃcients A. B.3 . ∂u ∂α ∂u ∂β ∂u ∂u ∂u = + = + ∂t ∂α ∂t ∂β ∂t ∂α ∂β Hence. (2. we evaluate. D. we ﬁnd ∂ 2u ∂ 2 u ∂α ∂ 2u ∂β ∂ 2u ∂α ∂ 2u ∂β ∂ 2u ∂ 2u = + + + =− 2 + ∂t∂x ∂α2 ∂x ∂β∂α ∂x ∂β∂α ∂x ∂β 2 ∂x ∂α ∂β 2 By the ﬁrst standard form (2.14). the ﬁrst standard form is −13 ∂ 2u − 5u = 2 ∂t∂x or ∂ 2u 1 = − (5u + 2) ∂t∂x 13 (2. E.15) into the canonical form Solution To (a). E = 0. The two characteristic equations are √ √ dy b − b2 − ac x dy b = b2 − ac x = =− . = = dx a y dx a y (2. β =t+x Indeed. G = −5. F = 2 A = a( Thus. F by the formulae (see (2.14) Note that one can obtain from the ﬁrst standard form the second standard form by the mapping α = t − x. G. x > 0.16) . ∂t ∂t ∂x ∂x ∂ξ ∂η ∂ξ ∂η ∂ξ ∂η ∂ξ ∂η 1 13 B=a + b[ + ]+c = −6 − [−2 + 3] − 6 = − . C.2)) ∂ξ 2 ∂ξ ∂ξ ∂ξ ) + 2b + c( )2 = (−2)2 + 2 − 6 = 0. (a) Find equations of the characteristic for the following hyperbolic equation: y 2uxx − x2uyy = 0. y > 0. ∂t ∂t ∂t ∂x ∂x ∂t ∂x ∂x 2 2 ∂η ∂η ∂η 2 ∂η 2 C = a( ) + 2b + c( ) = (3)2 − 3 − 6 = 0.

To (b).16 Note that this is equivalent to setting A = C = 0. Solving the above ordinary diﬀerential equations by the method of separating variables. We ﬁnd the standard form of the hyperbolic equation in the new variables ξ = y 2 − x2 . y 2 + x2 = constant. D.17). β = ξ − η. (2. we compute (see (2. η in terms of the variables α. B. we ﬁnd the equations for the characteristics in the implicit form y 2 − x2 = constant. we obtain the following equation −(x2 + y 2)uξ + (y 2 − x2)uη uξη = . b = 0. we ﬁnd the ﬁrst standard form uξη = ηuξ − ξuη 2(η 2 − ξ 2 ) We can ﬁnd the second canonical form of equation (2. e = 0. . d = 0. C.2)) A = a(ξx )2 + 2 b ξx ξy + c (ξy )2 = y 2(−2x)2 − x2(2y)2 = 0 B = aξx ηx + b(ξx ηy + ξy ηx ) + cξy ηy = y 2 (−2x)(2x) − x2 (2y)(2y) = −8x2 y 2 C = a(ηx )2 + 2 b ηx ηy + c (ηy )2 = y 2(−2x)2 − x2(2y)2 = 0 D = aξxx + 2bξxy + cξyy + dξx + eξy = −2((x2 + y 2 ) E = aηxx + 2bηxy + cηyy + dηx + eηy = 2(y 2 − x2 ) F = 0. E. F. (2. β to obtain the second canonical form. G = 0.17).17) Having the coeﬃcients a = y 2. G in the equation Auξξ + 2Buξη + Cuηη + Duξ + Euη + Gu = F. f = 0. we compute the coeﬃcients A. Substituting the above coeﬃcients to the equation (2. we can rewrite equation (2. Now. g = 0 in the equations auξξ + auξη + auηη + auξ + euη + fu = g. c = −x2. given in variables ξ. introducing the new variables α = ξ + η. by the equations of the characteristics.18) 8x2 y 2 Hence.18). Then. η = y 2 + x2 .

Consider th equation y ∂ 2u ∂ 2u ∂ 2u + (x + y) + x 2 = 0. we consider λ1 = dy x = . x and λ2 = 1.17 Then. r2 = = = 2y 2y 1. (x + y) − (x + y)2 − 4xy dy 2 dy ) − (x + y) + x = 0. ∂x2 ∂x∂y ∂y (2. we compute uξ = uααξ + uβ βξ = uα + uβ uη = uααη + uβ βη = uα − uβ uξη = uαααη + uαβ βη + uβααη + uββ βη = uαα − uββ Hence. 2.4 . by substituting. 2αβ 1. except the line y = x. 1 Solution. We have a = y. Then. The equation for the characteristics is: y( Then. y plane. b = (x + y) c = x. uαα − uββ = − Example 2. if x ≥ y if x > y. where (2. we ﬁnd the roots x . equation (2. we obtain the second canonical from equation (2.20) if x ≤ y if x > y. x 2 − 4xy (x + y) − (x + y) (x + y) + |x − y| y . y Hence.15). dx dx (2. dx y dy =1 dx (2. Find the range of x and y for which the equation is hyperbolic.19) βuα + αuβ . 4 is positive for all real x = y . we obtain the following two equations for characteristics So that. The dis2 criminant 1 b2 − a c = (x − y)2 > 0. Transform the equation to a canonical form.19) is hyperbolic on the whole x.21) . d = e = g = f = 0.19) becomes the parabolic equation. (x + y) − |x − y| r1 = = = y 2y 2y 1.

y)u = f(x.8) for characteristic will be also complex. (2. y) ∂ 2u ∂ 2u ∂ 2u + 2b(x. y). y) 2 + ∂x2 ∂x∂y ∂y ∂u ∂u +d(x.3) is determined by the conjugate roots of quadratic equation (2. that is. ∂x py (2. y) + c(x. Therefore. In the new variables ξ and η. y) = y − x = constant. dx dx . Then the roots λ1 and λ2 of the quadratic equation (2.22) To transform the hyperbolic equation to the standard form.10). y) and η(x. F = 0.4 Transformation of an Elliptic Equation into the Standard Form Let us recall that the equation a(x. the hyperbolic equation (2. y) of equation (2. we compute A = a(ξx)2 + 2bξx ξy + c (ξy )2 = y(−2x)2 + (x + y)(−2x)(2y) + (2y)2 = 0 B = aξxηx + b(ξxηy + ıy ηx ) + cξy ηy = = y(−2x)(−1) + 1 (x + y)(−2x − 2y) + 2xy = −(x − y)2 = −ψ 2.10) are complex and hence the solution ξ(x. y) and η(x. Now.8) for the chracteristics This characteristic equation is a( dy 2 dy ) − 2 b + c = 0. (2. we ﬁnd ϕ(x. mapping (2. Then .23) ψ(x. 2 C = a(ηx)2 + 2bηx ηy + c(ηy )2 = y − (x + y) + x = 0 D = aξxx + 2bξxy + cξyy + dξx + eξy = −2(y − x) = −2ψ. y) + e(x. y) are complex solutions of the diﬀerential equation (2. η = y − x.18 Solving the above equations. (2. by λ − iµ and λ + iµ. y) = y 2 − x2 = constant. we consider the mapping ξ = y 2 − x2 . the functions ξ(x.24) ∂ξ∂η ∂ξ 2.25) is elliptic when the discriminant b2 − ac < 0.19) takes the standard form ∂ 2u ∂u η + = 0. y) + g(x. E = aηxx + 2bηxy + cηyy + dηx + eηy = 0.

By the real mapping. we apply the mapping ξ(x.28) . y) = α − iβ(x. we transform the equation in x.2) for the evaluation of the coeﬃcients A. We consider the complex mapping ξ(x. y) + iβ(x. y). G. C. ∂ξ 2 ∂η as the representative. D. y) = α(x. 2 ∂ξ ∂η A ∂ξ A ∂η A A with the Laplace’s equation ∂ 2u ∂ 2u + 2 = 0. ∂x2 ∂x∂y ∂y Transform the equation into the standard form. y) = β(x. η using formulae (2. y) − iβ(x. F . y) = α(x. y) = constant is the complex solution of the characteristic equation then the conjugate η(x. y) = α(x. if ξ(x. E. Then we arrive to the equation A( ∂ 2u ∂ 2 u ∂u ∂u + 2) + D +E + Gu = F ∂ξ 2 ∂η ∂ξ ∂η (2. η(x. y) + iµ(x. y) as the real and the imaginary parts of the complex solution.26) Thus. y). y). y) dx (2. we obtain the standard form ∂ 2u ∂ 2 u D ∂u E ∂u G F + 2+ + + u= . y) is also the complex solution of the characteristic equation. dx dy = λ(x. y variables to the equation in the new variables ξ. η(x. Example 2.5 .29) (2.19 Soloving the quadratic equation aλ2 − 2bλ + c = 0 for λ = dy . y) + iβ(x. Dividing by A. y) = α(x. we get two equations dx dy = λ(x. (2. y). y) − iµ(x.27) with A = C and B = 0. In order to obtain a real mappipng which leads the original equation to the standard form. B. Determine type of the equation ∂ 2u ∂ 2u ∂ 2u −2 + 2 2 = 0.

the standard form of the equation in the new variables is: ∂ 2u ∂ 2u + 2 = 0. η(x. it is the elliptic equation. ∂ξ 2 ∂η Example 2. g = 0. The discriminant b2 − ac = −y < 0 for y > 0. f = 0. f = 0. d = 0. So that.6 . The characteristics of this equation are deﬁned by the equation ( Solving the equation for λ = dy 2 dy ) − 2 + 2 = 0. by formulae (2. d = 0. we ﬁnd the following complex mapping ξ(x. E = 0. dx dy = −1 − i. e = 0. b = 0. G = 0. Apply the real mapping ξ = x + y and η = x. dx y > 0. ∂x2 ∂x∂y ∂y ∂ξ ∂η Thus. it is the elliptic equation. Determine type of the equation ∂ 2u ∂ 2u + y 2 = 0. dx Integrating the above equations.2). Solution. So that. c = 2.20 Solution. ∂x2 ∂y y > 0. The discriminant b2 − ac = −1 < 0. Then. (2. dx dx dy . We have data: a = 1. In terms of ξ. g = 0. The characteristics of this equation are deﬁned by the equation ( dy 2 ) + y = 0. y) = y + x − ix. b = −1. c = y.30) Transform the equation into the standard form. e = 0. we have the two equations dx dy = −1 + i.31) (2. y) = y + x + ix. . We have data: a = 1. we ﬁnd A = a(ξx )2 + 2bξx ξy + c(ξy )2 = 1 B = aξx ηx + b(ξx ηy + ξy ηx ) + cξy ηy = 0 C = a(ηx)2 + 2bηx ηy + c(ηy )2 = 1 D = 0. η we get ∂ 2u ∂ 2u ∂ 2u ∂ 2u ∂ 2u −2 + 2 2 = 2 + 2. F = 0.

we ﬁnd only one solution ξ(x. the standard form of the equation in the new variables is: ∂ 2u ∂ 2u 1 ∂u + 2− = 0. (2. = 2 ∂x ∂ξ ∂x2 ∂ξ 2 ∂u 1 ∂u ∂ u 1 ∂ 2u 1 ∂u =√ . y). y) = x − 2i y.32) Thus. Then.21 Solving the equation for λ = dy . y) = constant of the equation a dy b = dx a In this case. where η(x. η we get ∂ 2u ∂ 2u ∂ 2 u ∂ 2 u 1 ∂u +y 2 = 2 + 2 − = 0. y). we have the two equations dx dy √ = −i y. ∂ξ 2 ∂η η ∂η y > 0. ∂x2 ∂y ∂ξ ∂η η ∂η y > 0. y) = constant is an arbitrary family of curves such that Jacobian J (ξ. = − √ 3 2 2 ∂y y ∂η ∂y y ∂η 2 y ∂η In terms of ξ. Then. dy √ = i y. η) = ξx ηy − ξy ηx = 0. Because ξx b = −λ = − ξy a . η = η(x.5 Transformation of a Parabolic Equation into the Standard Form For a parabolic equation there is one repeating root of equation (2. y > 0 √ Let ξ = x and η = 2 y. dx y > 0. dx Integrating the above separate variables equations. 2.10) equal b to . we ﬁnd the following complex mapping √ √ ξ = ξ(x. we consider the mapping ξ = ξ(x. we ﬁnd ∂u ∂u ∂ 2u ∂ 2u = . y) = x + 2i y. η = η(x.

Multiplying this equation by b and using the equality b2 = ac. . Therefore. In order to determine the type of the equation. f = 0. ∂ξ∂η bξx + cξy = 0 Example 2. g = 1. we are free to choose any function for which the Jacobian ξt ηx − ξx ηt = 0. to its standard form. we ﬁnd abξx + b2ξy = 0 abξx + ac2ξy = 0 a(bξx + cξy ) = 0 Hence. we have B = aξxηx + b(ξx ηy + ξy ηx ) + cξy ηy = ηx (aξx + bξy ) + ηy (bξx + cξy ) = 0. b = −1. Let us note that for η. c = 1. we obtain the solution t + x = constant. Since the discriminant equals to zero. t ≥ 0. From the general form of a linear PDE of the second order (2.22 therefore aξx + bξy = 0.7 Transform the following equation: utt − 2utx + uxx − ut − ux + u = 0. the term with ∂ 2u in the canonical form is absent. −∞ ≤ x ≤ ∞. η = t. there is one family of characteristics determined by the ordinary diﬀerential equation dx b = = −1 dt a Hence. d = −1. Now. we compute the discriminant b2 − a c = (−1)2 − 1 = 0.1). the equation is parabolic one. we ﬁnd the coeﬃcients a = 1. we choose the mapping ξ = t + x. e = −1. Then. Solution.

2).23 Now.33) Transform the equation into the standard form. Let η(x. . y) is an arbitrary function such that Jacobian ξx ψy − ξy ψx = 0. 2 2 C = x2ηx + 2xyηx ηy + y 2ηy = x2 . η(x. D = a ξtt + 2b ξtx + c (ξx )2 + d ξt + e ξx = 0 − 2 ∗ 0 + 0 − 1 − 1 = −2. y) = x. or 1 1 1 uξ = uηη + uη − u. Solution. we obtain the standard form of the equation uηη − 2uξ − uη + u = 0. dx x It is easy to ﬁnd the solution y = kx. F = 0. In order to ﬁnd the canonical form of the equation. G = 1. we consider the mapping ξ = y − kx. we compute the coeﬃcients A = a(ξt )2 + 2bξt ξx + c(ξx )2 = 1 − 2 + 1 = 0. y). Here.8 . The discriminant b2 − ac = x2 y 2 − x2y 2 = 0. ∂x2 ∂x∂y ∂y (2. so that it is a parabolic equation for all real x and y. where k is a generic constant. B = aξt ηt + b(ξt ηx + ξx ηt ) + c ξx ηx = 1 − (0 + 1) + 0 = 0 C = a (ηt )2 + 2b ηt ηx + c(ηx )2 = 1 − 2 ∗ 0 + 0 = 1. Hence. we solve the characteristics equation dy y = . E = a ηtt + 2b ηtx + c (ηx )2 + d ηt + e ηx = 0 − 2 ∗ 0 + 0 − 1 − 0 = −1. B = x2ξx ηx + xy(ξxηy + ξy ηx ) + y 2ξy ηx = 0. we ﬁnd the coeﬃcients 2 2 A = x2ξx + 2xyξx ξy + y 2ξy = (kx − y)2 = 0. Then. 2 2 2 Example 2. Determine type of the equation x2 ∂ 2u ∂ 2u ∂ 2u + 2xy + y 2 2 = 0. Then. η = η(x. by the general form (2.

Much of the theoretical work on the properties of solutions to hyperbolic problems assume the equation has been written in the canonical form uξξ − uηη = Φ(ξ. . u. Many computer programs have been written to ﬁnd the numerical solution of the canonical form. we can always come back to the original variables.24 Hence. uη ). The three major classiﬁcations as elliptic. ∂η 2 Applications of the canonical form of elliptic.33) takes the canonical form ∂ 2u = 0. η. in the new variables ξ and η equation (2. Having the numerical solution in the new variables. parabolic and hyperbolic equations 1. in fact classify physical problems into three basic physical types: steady-state problems. diﬀusion and wave propagation. 2. uξ . parabolic and hyperbolic equations. The mathematical solutions of these three types of equations are very diﬀerent. 3.

Question 2.2 Transform the equation utx + ut + ux + u = f(t. y) Question 2.25 2.6 Exercises Question 2. 5. uxx − 8uxy + 2uyy + xux − yuy = 0. in the lower half-plane y < 0. Find characteristics curves of the above three equations. 2. Question 2. 4uxx + 2uxy + uyy − uy = 0. 3.4 Show that all linear partial diﬀerential equations of the second order in two variables x and y of elliptic type with constant coeﬃcients can be transformed into the canonical form uxx + uyy + gu = f(x. 4. 9uxx + 12uxy + 4uyy + ux = 0.3 Find characteristics of Tricomi’s equation y uxx + uyy = 0. x).5 Show that all linear partial diﬀerential equations of the second order in two variables t and x of hyperbolic type with constant coeﬃcients can be transformed into the canonical form utt − uxx + gu = f(t.η − uξξ + uξ + uη + u = g(ξ. Transform the above three equations into their canonical forms. Question 2.1 Classify the following equations: 1. x) . Transform Tricomi’s equation into the canonical form in the upper-half of the plane when y > 0. into the equation uξ. η).

26 .

The solution u(t.1) are dx 2 dx dx ) − k 2 = 0. 27 (3.1) for give functions φ0 . ut (0.Chapter 3 Hyperbolic Equations 3. φ1 and the coeﬃcient k.2) 3. Let us write equation (3.1). to obtain the ﬁrst standard form of equation (3.1 The Initial Value Problem for Wave Equation utt = k 2 uxx. x) = φ1 (x). Find the solution u(t. x) = [φ0(x − kt) + φ0 (x + kt)] + 2 2k x+kt x−kt φ1 (ξ)dξ.1. = k.3) . −∞ < x < ∞. uξη = 0. Note that the characteristic equations for the wave equation (3. This problem describes the motion of an inﬁnite string under initial conditions. and x + kt = constant. x) is given by the formula 1 1 u(t. The problem was solve by French mathematician D’Alembert. Step1. x) of the initial value problem (3. dt dt dt The solutions of the equations for characteristics are: ( x − kt = constant ξ = x − kt. 0 < t < ∞. x) = φ0 (x). (3.1 D’Alembert’s Solution We solve initial value problem (3. u(0.1) in terms of the new variables η = x + kt. = −k.1) in four steps.

we integrate the equation with respect to the variable η to obtain u(ξ. to obtain u(t. we ﬁnd the ﬁrst standard form of equation (3. η) = φ(η). (3. we integrate equation (3. η) = sin η + ξ 2 u(ξ. η) given in terms of the variables ξ and η to the original variables t and x To ﬁnd all solutions in terms of the original variables t and x.3).η = 0. one can easily check that the functions u(ξ. we transform the solution u(ξ. η) = 1 + tan ξ. η) = ϕ(η) + ψ(ξ). η) = η 2 + eξ u(ξ. x) = ϕ(x + kt) + ψ(x − kt).28 Simple application of the chain rule gives ux = uξ + uη ut = k(−uξ + uη ) uxx = uξξ + 2uξη + uηη utt = k 2 (uξξ − 2uξη + uηη ) By substitution to equation (3. Thus. For example. ϕ(η) = φ(η dη. Step 3.4) η = x + kt. In step 2. η are solutions of the equation uξ. to u(ξ.1). . Step 2. to obtain the solution uη (ξ. In the step 3. where ϕ(η) and ψ(ξ) are any diﬀerentiable functions of the variables ξ and η. Secondly. we substitute ξ = x − kt.1). all solutions of equation (3. This completes step 2.3) have the following form: u(ξ. ﬁrstly with respect to the variable ξ. This completes step 1. η) = ϕ(η) + ψ(ξ). η) = ϕ(η) + ψ(ξ).

For example. Thus. Step 4. (3. x) = φ1(x) This means to ﬁnd functions ϕ and ψ such that ϕ(x) + ψ(x) = φ0 (x).5). x) = (x + kt)2 .8) 1 1 ϕ(x + kt) = φ0 (x + kt) + 2 2k 1 1 ψ(x − kt) = φ0(x − kt) − 2 2k K 2 0 x−kt K φ1 (s) ds − 2 0 x+kt φ1 (s) ds + . the functions u(t. u(t.7) φ1(ξ) dξ + K. φ1 (ξ) dξ + (3. x) = sin(x − kt) + (x + kt)2 .6) from 0 to x to obtain ϕ(x) − ψ(x) = for a constant K.4) (all solutions) represents the sum of any two moving waves. x) = sin(x − kt). 1 k x 0 (3. we choose that one which satisﬁes the initial value conditions u(0. Solving the two equations ϕ(x) + ψ(x) = φ0(x) ϕ(x) − ψ(x) = we ﬁnd 1 k x 0 φ1(ξ) dξ + K. x) = φ0 (x). Each wave moves in opposite directions with velocity k. we substitute to formula (3.5) with any diﬀerentiable functions ϕ and ψ. among all solutions of the form u(t. we choose from all solutions given by formula (3. two oppositely moving waves This ends step 3.6) We now integrate the second equation of (3. ut(0.29 The general solution (3.4) the solution u(t. K 2 0 x K φ1(ξ) dξ − . x) which satisﬁes the initial-value conditions. kϕ (x) − kψ (x) = φ1(x). 2 0 x 1 1 ϕ(x) = φ0 (x) + 2 2k 1 1 ψ(x) = φ0 (x) − 2 2k Now. one − right moving wave one − left moving wave u(t. In the step 4. x) = ϕ(x + kt) + ψ(x − kt) (3.

x) = sin x. x) = 0. Let us consider the initial value problem utt − k 2 uxx = 0. x+kt x−kt φ1 (ξ)dξ. with the initial conditions u(0. x) = 0. Let us consider the initial problem with the initial velocity utt − k 2 uxx = 0. by D’Alembert’s formula the solution is 1 u(t. x) is u(t. x) = [sin(x − kt) + sin(x + kt)] 2 2.1 . For φ1 (x) = 0. x) = sin x. the solution of the initial-value problem is 1 1 u(t. t ≥ 0. −∞ < x < ∞. 2k Example 3. x) = [φ0(x − kt) + φ0 (x + kt)] + 2 2k Examples 1. −∞ ≤ x < ∞. t > 0.30 Hence. −∞ < x < ∞. For φ0 (x) = 0. ut(0. . u(0. ut (0. x) = cos 3x.9) ut(0. x) = sin 3x. u(0. (3. Solve the following initial value problem: utt − 25uxx = 0. t > 0. by D’Alembert’s formula the solution u(t. x) = 1 2k x+kt x−kt sin ξ dξ = 1 [cos(x + kt) − cos(x − kt)].

0) − γ0 u(t.31 Solution. 3. L) = ψ1(t). (3. Force speciﬁed on the boundaries ux(t. 0) = 0. . u(t.10) u(0. 2 30 3. t ≥ 0. L) = 0. 0 < t < ∞. 0 < t < ∞. u(t. x) = φ0 (x). 2.1.2 The Initial Boundary Value Problem for Wave Equation utt = k 2uxx . 0) = ψ0(t). L) = ψ1(t). ux(t. u(0. 2 2k x−kt 1 1 x+5t = [sin 3(x − 5t) + sin 3(x + 5t)] + cos 3ξ)dξ 2 10 x−5t 1 1 = [sin 3(x − 5t) + sin 3(x + 5t)] + [sin 3(x + 5t) − sin 3(x − 5t)].2 Solution to the Finite Vibrating String by Separation of Variables utt = k 2 uxx. Apply D’Alembert’s formula for k = 5. x) = φ1 (x). L) = ψ1 (t). x) = [φ0(x − kt) + φ0 (x + kt)] + φ1(ξ)dξ. Let us consider the following initial boundary value problem: 0 < x < L. L]. x) = φ1 (x). k 2 > 0. ut (0. 0 < x < L. Controlled end points boundary conditions u(t. When variable x ∈ [0. (3. 3. Elastic attachment ux (t. 0) = ψ0(t). ux (t. the following three kind of boundary conditions are considered: 1. u(t. L − γ1 u(t. x) = φ0 (x). 0) = ψ0(t).11) To solve the initial boundary value problem ut (0. φ0 (x) = cos 3x and φ1 (x) = sin 3x to obtain 1 1 x+kt u(t.

32 we start by seeking standing wave solutions to the wave equation. Case 2. k2 or 1 T (t) X (x) = =λ 2 T (t) k X(x) where λ is a constant. Next. (3. we consider the following three cases: Case 1. λ < 0. Hence. So.13) X(x) = C sin(βx) + D cos(βx). we ﬁnd roots of the quadratic functions P2 (α) = α2 − k 2 λ = 0. X(x) = Cx + D. where A.12) T (t) = At + B. X (x) − λX(x) = 0. λ = 0 If λ = 0 then there are linear solutions to the equations (3. that is.12) In order to solve these ordinary diﬀerential equations. x) = X(x)T (t) Substituting this expression to the wave equation 1 utt = uxx k2 we ﬁnd 1 T (t)X(x) = X (x)T (t).12) take the form T (t) = Aekβt + Be−kβt. solutions of the following form: u(t. the solutions are T (t) = A sin(kβt) + B cos(kβt). X(x) = Ceβx + De−βx . (3. the solution either it is trivial (u(t. So that. for λ = −β 2. λ > 0. we obtain the two ordinary diﬀerential equations T (t) − k 2 λT (t) = 0. √ If √ < 0 then there are two complex roots α1 = −ik −λ and α2 = λ ik −λ. B. C and D are constants to be determined by the initial and boundary conditions. x) = 0) or unbounded because of initial boundary conditions. Case 3. In this case the solution can be trivial (u(t. . If λ = β 2 > 0 then the solutions of equations (3. in this case. x) ≡ 0) or unbounded and feasible because of the initial value conditions. Q(α) = α2 − λ = 0.

m=n 2 we can ﬁnd the coeﬃcients L 2 nπx An = φ1 (x) sin dx. n = 1. ..15) satisﬁes the wave equation and homogenous boundary conditions. u2(t. So that.15) into the initial conditions u(0. x). we apply the homogeneous boundary conditions plugging into u(t. gives the two equations ∞ ut(0. we ﬁnd nπ βn = . x) = φ1 (x). So that... Bn . ∞ Bn sin n=1 nπx = φ0(x). Now. 2. nπk 0 L L 2 Bn = L L 0 nπx φ0 (x) sin dx. Then. sin(βL) = 0.. u(t. x).. the function ∞ (3. L We note that for Tn (t) = An sin(kβn t) + Bn cos(kβn t). L) = 0.17) . t ≥ 0. mπx nπx sin sin dx = L L L 0 .14) u(t. m = n. x) = n=1 sin nπkt nπkt nπx [An sin + Bn cos ] L L L (3. Because the wave equation is linear one.. x) = sin(βn x)[An sin(kβn t) + Bn cos(kβn t))] un (t. 2.33 Let us consider the solution given by formula (3.16) Using the orthogonality condition 0. x) = φ0(x).13). The constant β has to satisfy the equation sin(βL) = 0. D = 0. 0) = T (t)X(0) = D[A sin(kβt) + B cos(kβt)] = 0. we obtain u(t.. L An n=1 nπk nπx sin = φ1 (x) L L (3.. L (3. 0) = u(t. x) is the solution of the wave equation which satisﬁes the homogeneous boundary conditions for arbitrary constants An .. when λ < 0 . therefore every linear combination of u1(t. Xn (x) = sin(βn x) un (t. . . Substituting sum (3. is also a solution of the wave equation which satisﬁes the homogeneous boundary conditions. n = 1. L) = T (t)X(L) = C sin(βL)[A sin(kβt) + B cos(kβt)] = 0.

. the solution u(t. .15) with the constants An and Bn ..34 for n = 1... determined by the formulae (3. 2. x) of the initial boundary problem is given by formula (3..17). 2. . .. n = 1. Finally.

x) = X(x)T (t) ¥ © ¤ ¥ We shall now make the following observations: 1.35 We present these three cases below on the following diagram : § Solution u(t.18) if the initial velocity φ1(x) = 0 with the homogeneous boundary conditions. x) § ¦ § λ < 0. when the initial position of the vibrating string u(0. λ = −β 2 c ¦ ¥ $$$ $$ $$$ $$$ c c ¤ § c ¤ § Possible values of λ ¤ ¥ ¦ λ=0 ¦ T (t) = A sin(kβt) + B cos(kβt) X(x) = C sin(βx) + D cos(βx) d d § ¤ ¥ § ¦ ¥ ¦ λ > 0. adding the terms of each simple vibration. λ = β 2 c (k2β)2t ¤ ¥ −(k2 β)2 t T (t) = Ae + Be X(x) = Ceβx + De−βx ¤ ¤ ¥ d d ¦ d d § d ¦ T (t) = At + B X(x) = Cx + D c c u(t. Let the function ∞ nπx φ0 (x) = Bn sin . x) = φ0 (x) is present. simple sine vibration of a string is given by the term Bn sin nπx nπkt cos . x) = n=1 Bn sin nπx nπkt cos L L (3. consider the initial state of a . we obtain the solution of the initial boundary problem. Let us note that the solution takes the following form: ∞ u(t. Namely. L n=1 Then. L L Thus.

from the initial conditions. the solution of such initial boundary value problem is m u(t. suppose that the initial string position is φ0(x) = sin πx 3πx 5πx + 0.25 sin cos . The n-th term in the solution (3. and bn = 0 for n > m. that is u(t. L L L L L L 2.19) We can obtain formula (3. . x) = sin πkt 3πx 3πkt 5πx 5πkt πx cos + 0.5 sin cos + 0.18). L L L The overall response to this initial condition would then be the sum of the responses to each term. Then. So. L L (3.25 sin .. we compute the coeﬃcients Bn = nπx 2 L φ0(x) sin dx L 0 L L 2 m sπx nπx = bs sin sin dx L s=1 L L 0 = bn . Indeed. This harmonic can be rewritten in the following form Rn sin nπx nπk cos (t − δn ).36 vibrating string with ﬁxed end points m φ0 (x) = n=1 bn sin nπx .20) ..15) sin nπx nπkt nπkt [An sin + Bn cos ] L L L is called n-th mode of vibration or n-th harmonic. m. the coeﬃcients are bn for n = 1..18). 2. x) = n=1 bn sin nπx nπkt cos . Hence.19) from the solution given by (3. we get the solution (3. by formula (3. For example.19).5 sin + 0. L and with zero initial velocity φ1(x) = 0. L L (3.

37

where δn is the phase angle and Rn is the amplitude. Indeed, we have [An sin nπkt nπkt + Bn cos ]= L L An nπkt Bn nπkt 2 A2 + Bn [ sin + cos ]= n 2 2 L L A2 + Bn A2 + Bn n n nπkt nπkδn nπkt nπkδn 2 A2 + Bn [sin sin + cos cos ]= n L L L L nπk 2 A2 + Bn cos (t − δn ) = Rn cos ωn (t − δn ), n L

2 A2 + Bn is the amplitude and δn is n

nπk is frequency, Rn = where ω = L the phase angle. Example 1.

1. By using separation variables and Fourier cosine series, solve the following problem for ﬁnite string with ﬁxed ends for appropriate initial data φ0(x) and φ1 (x) utt = 4uxx , 0 ≤ x ≤ L, 0 < t < ∞, u(t, 0) = 0,

m

u(t, L) = 0, ut(0, x) = 0.

(3.21)

u(0, x) =

1 nπx sin , n L n=1 2

2. Determine the frequency ω, the amplitude Rn and the phase angle δn . 3. Graph the solution u(t, x) for m = 1, L = 2 and t = −1, 0, 1 Solution. To (a): By the formula (3.19), we ﬁnd the solution u(t, x) = 1 nπx 2nπt sin cos . n L L n=1 2

m

2nπ To (b). From the above formula, we ﬁnd that the frequency is ω = , the L 1 amplitude is Rn = n and the phase angle δn = 0. 2 To (c): For m = 1, L = 2, the solution is u(t, x) = 1 πx sin cos πt. 2 2

38

. Example 2. Solve the following initial boundary problem by the method of separation of variables: utt = 36uxx , t > 0, 0 ≤ x ≤ 1, u(0, x) = 0, u(t, 0) = 0, ut (0, x) = 4, 0 ≤ x ≤ 1, u(t, 1) = 0, t ≥ 0. (3.22)

**Solution. We note that φ0(x) = 0. Therefore, by formula (3.17), the coeﬃcint Bn = 0 and we compute the coeﬃcients An = 1 3nπ
**

1 0

4 sin nπx dx =

4 [1 − (−1)n ]. 3n2 π 2

**Hence, by formula (3.15), the solution is: u(t, x) = 4 [1 − (−1)n ] sin(nπx) sin(6nπt). 3n2 π 2 n=1
**

∞

Example 3. What is the solution to the simple supported at the ends beam with initial conditions u(0, x) = sin πx, ut (0, x) = sin πx, , 0 ≤ x ≤ 1.

Solution. We note that the solution u(t, x) satisﬁes the wave equation utt = k 2 uxx, 0 ≤ x ≤ 1,

with the homogeneous boundary conditions u(t, 0) = 0, u(t, 1) = 0, t ≥ 0.

39

and the initial value conditions u(0, x) = sin πx, ut (0, x) = sin πx, 0 ≤ x ≤ 1.

**By method of separation variables, the solution is given by the formula
**

∞

u(t, x) =

n=1

sin

nπkt nπkt nπx [An sin + Bn cos ], L L L

**Hence, for φ0(x) = sin πx, ψ1 (x) = sin πx, and L = 1, we compute the coeﬃcients using formulae (3.17) 2 An = nπk
**

1 1 0

sin πx sin nπxdx = =

Bn = 2

0

sin πx sin nπxdx

1 , n = 1, kπ 0 n = 1, 1, n = 1, 0 n = 1,

Thus, the solution u(t, x) of the supported beam problem is u(t, x) = sinπx[ 1 sin(πkt) + cos(πkt)], πk

for 0 ≤ x ≤ 1 and t ≥ 0. Example 4. A gitar string of length L = 1 is pulled upward at middle so the it reaches heigt 0.5 and satisﬁes the wave equation utt = 9uxx , Assuming the initial position of the string u(0, x) =

0 ≤ x ≤ 1. 0 ≤ x ≤ 0.5,

x,

1 − x, 0.5 ≤ x ≤ 1, 0 ≤ x ≤ 1.

and the initial speed of the string ut(0, x) = 1,

Find the position u(t, x) of the string at time t and point x. Solution. We note that the solution u(t, x) satisﬁes the wave equation utt = 9uxx , 0 ≤ x ≤ 1,

with the homogeneous boundary conditions u(t, 0) = 0, u(t, 1) = 0, t ≥ 0.

−π ≤ x ≤ π. (1 − x) sin πx dx = 3. x) = x. sin 2x. ut (0. x= φ1(x) = 1. x) = cos 3x. x) = 0. x) = ((1 − (−1)n ) sin(3nπt) + kn2 π 2 n2 π 2 n=1 2 sin for 0 ≤ x ≤ 1 and t ≥ 0. u(0.17) An = 2 nπk 0 1 0 1/2 sin nπxdx = 2 1 cos nπ 2 ( − )= [1 − (−1)n ].5 ≤ x ≤ 1. for the given functions φ0 (x). x) = φ0 (x) = x. x) = sin 3x. knπ nπ nπ kn2π 2 1 1/2 Bn = 2 x sin nπxdx + 2 nπ nπ nπ nπ nπ − nπ cos + nπ cos 2 sin 4 sin 2 2 + 2 2 = 2 = n2 π 2 n2 π 2 n2 π 2 Hence. we (icients using formulae (3.40 and the initial value functions u(0. t > 0. 0 ≤ x ≤ 0. x) = cos 3x. u(0. . Exercises Example 3.2 Solve the following initial value problems: utt = 4uxx .23) and for L = 1.5. t > 0. x) of the gitar string problem is nπ ∞ 4 sin 2sin(nπx) 2 cos(3nπt). −∞ < x < ∞. the solution u(t.3 1. utt = uxx . 3. ut (0.23) By method of separation variables the solution is given by the formula u(t. L L L Hence. ut (0. ψ1 (x) by (3. −∞ < x < ∞. |x| > π. x) = t > 0. 0. ∞ ut(0. (3. 0. utt = uxx. 1 − x. u(t. 2. x) = n=1 sin nπx nπkt nπkt [An sin + Bn cos . −∞ < x < ∞. u(0.

0 ≤ x ≤ 2.6 Solve the initial value problem by the D’Alembert method. (c) Transform equation (3. x) = cos 4x. utt − 9uxx = 0.26) −∞ < x.4 For the following equation: uxx + 3uxy + 2uyy = 0.26).24) into canonical form.26) (b) Find the characteristic curves of the equation (3. ut(0. ∞<x<∞ (3. u(t. x) = 1. 0) = 0.26) into its canonical form. . Example 3. x > 0. (3.24) (b) Find characteristic curves of the equation (3. x) = x(4 − x). u(t.25) (b) Find characteristic curves of the equation (3. y∞. −∞ < x < ∞. Example 3. utt − 9uxx = 0.24).28) t ≥ 0. 2) = 0.25).25) into its canonical form. x) = sin 4x.24) (a) Determine the type of the equation (3.27) u(0.3 For the following equation: yutt − 16xuxx = 0. (3. (c) Transform equation (3. (c) Transform the equation (3. t geq0. (3. Example 3. Example 3. t ≥ 0.7 Solve the initial boundary value problem by the method of separation of variables. 0≤x≤4 (3.5 For the following equation: utt − 4utx + 4uxx − ut = 0. u(0. −∞ < x < ∞. t ≥ 0.41 Example 3.25) t > 0. (a) Determine type of the equation (3. ut (0. (a) Determine type of the equation (3.

t ≥ 0.42 Example 3. t ≥ 0.8 Consider the telegraphic equation utt + ut + u = c2 uxx . . and the homogeneous boundary value conditions u(t. L) = 0. u(t. x) = 0. 0 ≤ x ≤ L. Find the solution u(t. 0) = 0. 0 ≤ x ≤ L. x) = x(L − x). ut(0. x) of the telegraphic equation which satisﬁes the initial condition u(0. Hint: Apply the method of separation of the variables t and x.

k 2 is a constant and the given functions f(t. ψL(t) are continuous for 0 ≤ x ≤ L. with the boundary ∂R = (t. L) = ψL (t). 0) = ψ0(t). 0 ≤ t ≤ T }. (4. u(t.Chapter 4 Parabolic Equations 4.3) 0 ≤ x ≤ L. x) = φ0(x). x) : 0 ≤ t < T 43 when x = 0 or x = L . with the initial condition u(0. x) : 0 ≤ x ≤ L. Firstly. x). x). x) : 0 ≤ x ≤ L. 0 ≤ x ≤ L.1 Initial Boundary Value Problem We shall consider the heat equation ut = k 2 uxx + f(t. t ≥ 0. we shall state the weak maximum principle for the heat equation ut = k 2 uxx (4.4) in the closed rectangle R = {(t. t ≥ 0. when t = 0.2) t ≥ 0. φ0(x). and with the boundary conditions u(t. (t. Let us establish some of the properties of the solution u(t. ψ0(t).1) Here. x). (4. (4.

x) attains its maximum M∂R at the closed boundary ∂R. We note that w(t. Proof.4) in the rectangle R. x0) ∈ R interior to R. x) assumes its maximum value on whole closed rectangle R at a point on the boundary ∂R. x) + ≤ M− + 4L2 (x − x0 )2 ≤ M∂R + 4L2 (x − x0 )2 3 L2 = M − < M. therefore 0 < x0 < L and 0 < t0 < T . 4L2 4 In similar way. x) attains its minimum at a point on the boundary ∂R of R. x) = u(t. Then u(t. x) = φ0 (x) E x The initial boundary conditions at ∂R The maximum principle 4. x) + w(0. the maximum on the boundary does not exceeds the maximum on the closed rectangle. then we choose a point (t0 . Deﬁne the auxiliary function (x − x0)2 . We know that u(t. x0) is not on the boundary ∂R. L) < M. Suppose that M − M∂R = . such that u(t0. To proof the thesis of the maximum principle. we shall show that M = M∂R . 0) < M. x) = u(0.1 Let u(t. That is. so that M∂R ≤ M. we arrive at the inequalities w(t. we know that u(t. and w(t. when 0 ≤ x ≤ L. Also. x). x) attains its maximum M at the closed rectangle R.44 Let us note that the points on the interval (T. consider w(t. u(t. 4L2 Then. x) be a solution of the heat equation (4. are not included in the boundary ∂R. . x0) = M. Also. 0 ≤ t ≤ T. Since > 0 and (t0. x) at points on ∂R. t T T u = ψ0 u(t. x) in R u = ψL 0 L u(0.

x1 ) ≤ 0.2). Then. L) = u(t. x1) − k 2 |u(t. x1) ∈ R. x1 ) − k 2 wxx (t1. 0 ≤ x ≤ L.7) Thus. Therefore the maximum of w(t. taking −u(t. we can prove for minimum of u(t. x0) = u(t0. Because 0 < x1 < L.6) t ≥ 0. Proof.2) and (4.(4. 4L2 4 3 L2 = M − < M. x1) − k 2wxx (t1. x).1).x)∈∂R wxx(t1. From the weak maximum principle. 0) = u(t. it follows that every solution u(t.6) against the inequality (4. x). x) on R is at least M and it is attained at a point (t1 . x)|.3) satisﬁes the inequality wt(t1 .3) has at most one continuous solution. x)| ≤ max |u(t. it is easy to show that the diﬀerence w(t. Similarly. (4. 0 < t1 < T. Conclusion.5) (4.1 (Uniqueness) The initial boundary value problem (4. u(t. x) instead of u(t. But k2 k2 =− < 0. we can state the following theorems Theorem 4. x) of the initial boundary value problem ( 4. End of the proof. x) = u1(t. x) and u2(t. L) + ≤ M− + 4L2 (L − x0)2 ≤ M∂R + 4L2 (L − x0 )2 4L2 (0 − x0)2 ≤ M∂R + 4L2 (0 − x0)2 3 L2 = M − < M. x) − u2 (t. the inequality (4. (4. we have w(t. 4L2 4 But w(t0 . x) attains its maximum value on the boundary ∂R of the rectangle R. Hence wt (t1. 2L2 2L (4. Then.(4. 0) + ≤ M− + and w(t. we conclude that M = M∂R .7). Assume that there are two solutions u1 (t. x1 ) = ut (t1.8) As a consequence of the maximum principle. (4. x) is the solution of . (t. we have arrived at the contradiction. not on the boundary ∂R. then wt (t1. x1) ≥ 0. x0 ) = M. x1) = 0. x). x1) − k 2 uxx (t1.1). Therefore.45 Indeed.

(1) (2) (1) (2) (1) (2) (1) (2) t ≥ 0. we conclude that such a homogeneous initial boundary value problem has only trivial solution. x) = φ0 (x) − φ0 (x). u(1)(t. L) = ψL (t). 0) = ψ0 (t).10) for the same sours of energy f(t. x) is the solution of the heat equation vt = k 2 vxx. t ≥ 0. x) ∈ R. also w(t. (4. holds for all t ≥ 0. x) = u(1)(t. x) attains its total minim on R at the boundary ∂R. x) and u(2)(t.2 Let u(1)(t. Hence w(t. x) = φ0 (x). 0 ≤ x ≤ L. x) ≥ 0 for (t. 0 ≤ x ≤ L. |ψ0 (t) − ψ0 (t)| < . Suppose that the distance of the initial boundary conditions is less than > 0. the inequality |u(1)(t. L) = ψL (t). x) ≤ 0 for all (t. So that w(t. 0 ≤ x ≤ L. x) be two solutions of the two initial boundary value problems ut = k 2 u(1) + f(t. x)| < . Proof. so that |φ0 (x) − φ0 (x)| < . Let us note that the diﬀerence v(t. x) − u(2)(t.9) (1) 0 ≤ x ≤ L. u(1)(t. that is w(t. xx u(1)(0. . and (2) (1) (1) (1) t ≥ 0. But. x) ∈ R. x) = φ0 (x). x). x). xx u(2)(0. t ≥ 0. and 0 ≤ x ≤ L. x) ≡ 0 for all (t. x) ∈ R. x) − u(2)(t. therefore w(t.11) Then. Indeed. not greater than zero. x) ∈ R. 0) = ψ0 (t). u(2)(t. End of the proof. x). (1) (4. (4. t ≥ 0.46 the homogeneous heat equation with the homogeneous initial boundary conditions. By the maximum principle. x) ≡ 0 for all (t. x) attains its total maximum on R at the boundary ∂R. which satisﬁes the initial condition v(0. u(2)(t. Theorem 4. (2) (2) 0 ≤ x ≤ L. ut = k 2 u(2) + f(t. |ψL (t) − ψL (t)| < . because w(t. 0 ≤ x ≤ L.

|v(t. Question 1. 1) = 2 + t2 x∈[−∞. L) = ψL (t) − ψL (t). x) = φ0(x). 2x u(0. for suﬃciently large a. ∞) and u(t.47 and the boundary conditions v(t. −∞<x<∞ for all t ≥ 0 and −∞ < x < ∞. x)| ≤ max |φ0(x)|. by the weak maximum principle (see (4. By the assumption u(t. (t. −∞ < x < ∞. t ≥ 0. Example 4. By the assumption. −a)). x)| < . x)| < . by the maximum principle |v(t.a] for all t ≥ 0 and |x| ≥ a. x) of the initial boundary value problem ut = 36uxx . Then. where ∂R = {(0. 0)| < . for all t ≥ 0 and 0 leqx ≤ L. (1) (2) v(t. x∈[−a. x)|.1 Let us apply the maximum principle to give an estimate of the solution u(t. t ≥ 0. u(0. x). (t.12) u(t. x) → 0 when x → ∞. Let x ∈ [−a. 2t u(t. 0 ≤ x ≤ 1. Therefore. x) → 0. Using the maximum principle show that |u(t. Solution. L)| < . (t. (1) (2) t ≥ 0. x)| ≤ max |u(t. we have |u(t. x)| ≤ for all t ≥ 0 and −∞ < x < ∞. Then.∞] max |φ0 (x)|. a)}. Assume that the given function φ0 (x) is continuous and bounded for all x ∈ (−∞. −∞ < x < ∞. . 1 + x2 (4. |v(0. x)| ≤ max |φ0 (x)|. 0) = 0. so that. −a ≤ x ≤ a. when x → ∞. the inequality |u(t. a] for a positive a > 0. 0) = ψ0 (t) − ψ0 (t).8)) |u(t. Consider the following initial value problem: ut = k 2uxx . |v(t.x)∈∂R t ≥ 0. x) = .

Let us ﬁnd the maximum values on the boundary ∂R (t.15) . to the heat equation.. max } 2 t≥0 0≤x≤1 1 + x t≥0 1 + t2 = max{1. L λn = n2π 2 . T + λk 2 T = 0. Let us consider the following initial boundary value problem u(0. 1)} 0≤x≤1 t≥0 t≥0 2x 2t . x) : 0 ≤ x ≤ 1. Hence. L2 n = 1. x) = max{ max u(0. x) : 0 ≤ t < T when x = 0 or x = 1. x) : 0 ≤ x ≤ 1.14) Solving the boundary value problem for the ordinary diﬀerential equation X + λX = 0. (t. X k T where λ is the separation constant. x) ≤ max u(t. we get the equations X + λX = 0. t ≥ 0. (4. (4. 0. 0) = 0.48 By the maximum principle.2 Solution by Separation of Variables ut = k 2 uxx . 0). Substituting u(t. max 0. 0 ≤ x ≤ L. we observe that the maximum of all values of the solution u(t. when t = 0. x) in the domain R do not exceed 1. x) = φ0 (x). the maximum of all values of the solution u(t. x) = 1 (t. u(t. x) = T (t)X(x). u(t. we obtain (4. 0 ≤ x ≤ L.x)∈∂R max u(t. we arrive at the solution Xn (x) = sin nπx . X T = 2 = −λ. t ≥ 0.x)∈∂R 4. t ≥ 0} is on the boundary ∂R = (t.. L) = 0. x). that is (t. .16) X(0) = X(L) = 0. max u(t. (4.x)∈R max u(t. max u(t. 2. 1} = 1 = max{ max Hence. x) in the domain R = {(t.13) .

The function k 2π 2n2 t nπx L2 sin Bn e u(t. In order to determine the coeﬃcients Bn . 0) = 0.. 1 sin 3πx. L nπx dx. L n=1 ∞ − (4.20) 2 L nπx φ0(x) sin dx. x) = φ0(x) = n=1 Bn sin nπx .. we have ∞ u(0. u(t. we expand in the Fourier series of sines the function φ0 (x) given in the initial condition.. 2.19) is also the solution of the heat equation and satisﬁes the homogeneous boundary conditions for any choice of the coeﬃcients Bn . n = 1.. .49 Now.21) u(0.2 Solve the following initial boundary value problem: Bn = ut = k 2 uxx. x) = φ0 (x) = sin πx + . x) = Tn (t)Xn (t) = e L which satisﬁes the homogeneous boundary conditions. the equation for T (t) becomes T + with the solution −k 2n2 π 2t L2 Tn (t) = e .. L bn = 2 L L 0 φ0(x) sin we arrive at the solution of the initial boundary value problem ∞ L u(t.. .. 2 (4. n = 1. L2 n = 1. 0 ≤ x ≤ 1.. . . un (t. L 0 L Example 4.17) k 2 n2 π 2 T = 0. x) = . n = 1.. t ≥ 0. 2.. 2. x) = where Bn = bn and n=1 0 −n2π 2k 2 t nπx L2 Bn e sin dx L (4. we ﬁnd the solution of the heat equation −k 2n2 π 2t nπx L2 sin . Hence. 1) = 0. n = 1..18) (4.. Then. 2. 2.. . 2.. n = 1. L Having Fourier series of the initial function ∞ φ0(x) = n=1 bn sin nπx . t ≥ 0. . u(t. (4.

So. x) satisﬁes the homogeneous boundary conditions v(t. we ﬁnd the solution ∞ v(t. B3 = . x (ψL − ψ0)]. x) = e−π k t sin πx + e−9π k t sin 3πx. we introduce the new unknown function v(t. x) by the formula u(t. x) = v(t.. So that v(t. we consider the following initial boundary value problem: ut = k 2uxx .23) t ≥ 0. 2 4. x) = where L n=1 0 −n2π 2k 2 t nπx L2 Bn e sin dx. where this time ψ0 . Clearly. v(0. 0 ≤ x ≤ L. the unknown v(t.3 Transformation of Non-homogeneous Boundary Conditions to Homogeneous Let us consider the heat equation with constant temperature at the end of a rod. (4. ψL are constants. L (4. L) = 0. 0) = ψ0. We note that the coeﬃcients of the Fourier series of the initial value function 1 φ0 (x) = sin πx + sin 3πx. u(t. L . the solution is 2 2 2 2 1 u(t. 0 ≤ x ≤ L. 2 1 are B1 = 1. L) = 0. In order to transform the non-homogeneous boundary conditions to homogeneous ones. L t ≥ 0. L Bn = 2 L L 0 φ0 (x) sin nπx dx. L (4. t ≥ 0.20). 2 Therefore. x) = φ0(x) − [ψ0 + By the formula (4. = 0. 0) = 0. B2 = 0. 0) = v(t. u(0. x) + ψ0 + x (ψL − ψ0).50 Solution.25) φ0 (x) = φ0 (x) − [ψ0 + x (ψL − ψ0 )].24) 0 ≤ x ≤ L.. x) is the solution of initial boundary problem vt = k 2 vxx. v(t. x) = φ0(x). B4 = B5 = . u(t. (4. t ≥ 0.22) u(t. L) = ψL .

t ≥ 0. 2) = v(t. x) + 1 + x (4 − 1) 2 = v(t. So that t ≥ 0. 2 (4. the new unknown function v(t. 2 v(t.51 Finally. by the formula (4. 0) = v(t. x) = φ0 (x) = φ0 (x) − [ψ0(0) + (ψL(t) − ψ0(t)) 2 x 3x = 1 − [1 + (4 − 1)] = − . (4. (4.20).28) u(t. 2) + 4 = 4. Solution. x) + ψ0 + L (ψL − ψ0) = v(t. u(t. 0) + 1 = 1. 0 ≤ x ≤ 2.30) . x) = [ φ0 (ξ) sin dξ]e sin L n=1 0 L L 2 2 −4n π t ∞ 2 3ξ nπξ nπx 4 = [ − sin dξ]e sin 2 2 2 n=1 0 ∞ nπx 3 −n2π 2t = e sin 2 n=1 nπ ∞ L (4. 0) = 1. x) = v(t. By introducing the new unknown function v(t. u(t. x) is the solution of the heat equation. v(t. x) satisﬁes the initial condition x v(0. L 0 ≤ x ≤ 2. and vxx = uxx (t. u(t. u(0. x).26) Example 4. vt = k 2 vxx. Hence. x) = ut(t. we obtain the solution −n2 π 2k 2 t 2 nπξ nπx L2 v(t. in terms of original unknown u(t. x) = 1. 3x .29) Therefore v(t. x) + [ψ0 + x (ψL − ψ0)].27) (4. 2) = 0. 0 ≤ x ≤ L. t ≥ 0. 0) = 0. x). t ≥ 0. 2) + 1 + 2 (4 − 1) = v(t. x) + 1 + we note that vt(t. x) = v(t. we obtain Then.3 Solve the following initial boundary value problem: ut = 4uxx . x) by the formula x u(t. 2 2 Hence. 2) = 4.

2 0 ≤ x ≤ 1.e. 0 ≤ x ≤ 2. ). n = 1. . u(t. t ≥ 0. 0 ≤ x ≤ 1. ψL = 1 x x v(t.52 Now. we ﬁnd the solution u(t. First. 0 ≤ x ≤ 2. v(t.31) u(0. 2. x). x) = 1 + ∞ 3x 3 −n2 π 2t nπx + e sin . 1≤x≤2 (4. of the Fourier series of the initial value function φ0 (x) for k = 2. For this purpose we introduce new unknown by formula (4. x) satisﬁes the homogeneous boundary conditions. coming back to the original unknown function u(t.32) We compute the coeﬃcients Bn . Solution. 2) = 0. we transform the equation with non-homogeneous boundary condition to the equation with homogeneous boundary conditions. 2) = 0. 0) = 0.4 . x) = u(t. L = 2 Bn = 2 L L 0 φ0 (x) sin 2 nπx 1 1 πnx x πnx dx = x sin dx + (1 − ) sin dx L 2 0 2 2 2 1 1 πn 2 πn = cos + 2 2 sin nπ 2 nπ 2 2 πn 1 πn 2 + 2 2 sin − cos − 2 2 sin nπ nπ 2 nπ 2 nπ 4 πn = sin n2 π 2 2 . u(t. 0 ≤ x ≤ 2. L 2 Clearly. 0) = 0 and v(t. Solve the initial boundary value problem for the diﬀusion equation: ut = 4uxx . x) = φ0 (x) = x. Now. we solve the initial boundary problem vt = 4vxx . 2 2 n=1 nπ t ≥ 0. ψ0 = 0. t ≥ 0.23) for L = 2. v(t. 2 x . 0) = 0. 1 ≤ x ≤ 2 Write ﬁrst three terms of the solution u(t. the new function v(t. v(t. x) − . x) + ψ0 − (ψL − ψ0) = u(t. x) − φ0 (x) = φ0 (x) = x 1− . 1.. (4. Example 4. x) = u(0. i. t ≥ 0. by the formula (4.. v(0.29). 2) = 1. t ≥ 0.

t ≥ 0 v(0. t ≥ 0. t ≥ 0 t ≥ 0. 0 ≤ x ≤ L. v(t. . uxx = e−βtvxx (t. represents heat ﬂow across the lateral boundary. we get the solution ∞ v(t. 0 ≤ x ≤ L. β > 0. x) = e−βtv(t.20). x) = e−βt Bn e sin (4. vt = k 2 vxx . 0 ≤ x ≤ L. x) = n=1 4 n2 π 2 sin πn −n2π 2t πnx e sin 2 2 (4. we obtain the solution −n2π 2 k 2t ∞ nπx L2 u(t. (4. u(t. 0) = 0. n = 1. x) = ∞ x x 4 πn −n2π 2t πnx + v(t. 0 ≤ x ≤ L.. (4. x). L) = 0. x). L) = 0. t ≥ 0..20). the term −βu. u(0. (4. by the formula (4.35). 0) = 0.34) Let us consider the following initial boundary value problem: Here. 1 sin 3πx. 2. 2 v(t. Let us consider the initial boundary value problem: Bn = ut = uxx − u. Solving the initial boundary value problem (4.37) where .34) into simple one for the new unknown v(t. t ≥ 0.33) Coming back to the original unknown. by the formula (4. x) = e−βtv(t. ut = −βe−βtv(t. u(0.36) L n=1 2 L nπx int0 φ0 (x) sin dx. 0) = 0. x) = + sin e sin 2 2 n=1 n2 π 2 2 2 4. u(t. L) = 0.53 Hence. L L Example 4. Let us note that by the substitution u(t. x) = φ0 (x). 0 ≤ x ≤ L. x) + e−βtvt (t. 0 ≤ x ≤ L. x) we can transform the initial boundary problem (4. x) = sin πx + u(t.5 .4 More Developed Heat Equation ut = k 2 uxx − βu. we ﬁnd the solution u(t.35) v(t.. x). x) = φ0 (x).

L n=1 ∞ − (4. 2 1 −9π2 t e sin 3πx]. Then. t ≥ 0. u(0.19). u(t. 0 ≤ x ≤ L. x) = sin πx + v(t. x) of the non homogeneous heat equation. we ﬁnd the solution u(t. ψ0 (t) ≡ 0. v(t. 2 The solution of the problem (4. x) = 0. 1 sin 3πx. x) = e−t [e−π t sin πx + 2 4. x) = f1 (t) sin πx 2πx nπx + f2 (t) sin + . 0) = 0. 1 −9π2 t e sin 3πx. x) = ϕ(x). we have found the solution k 2π 2n2 t nπx L2 sin u(t. 2 0. coming back to the original unknown.. Assume that the given function as the heat sourse f(t. x). when f(t. 0) = 0. L) = 0. x) = e−π t sin πx + 2 Hence. x). ψl (t) ≡ 0. x) = e−t v(t. x) satisﬁes the heat equation with the initial boundary conditions vt = uxx. L L L (4.. x) ≡ 0.38) v(0. u(t. we shall ﬁnd the solution u(t. x) = Bn e . and f(t. (4. We apply the substitution u(t.54 Solution.39) In the previous section (see (4. L 0 L Now. (4.38) is v(t. t ≥ 0. L) = 0. t ≥ 0.40) where Bn = 2 L nπx ϕ(x) sin dx.5 Non-homogeneous Heat Equation Let us consider the non homogenous heat equation with the initial boundary value conditions ut = k 2uxx + f(t. t ≥ 0.. 0 ≤ x ≤ L. + fn (t) sin + . to eliminate the term −u. 0 ≤ x ≤ L. u(t. 0 ≤ x ≤ L. x) possesses the following series presentation: f(t..41) .

2..48) . and integrate from zero to L with respect to the variable L x. (4. . t ≥ 0.41) by sin . t ≥ 0.. . } we ﬁnd L the coeﬃcients nπx 2 L f(t.43) fn (t) = L 0 L Replacing the heat sourse function f(t. L ∞ Hence k 2n2 π2 nπx Tn (t) − fn (t)] sin = 0..47) Then. L 0 L n = 1. 0) = 0. x) sin ∞ nπx dx = fn (t) L n=1 sin ∞ ut = k 2 uxx + n=1 fn (t) sin u(0. n = 1. L n=1 [T (t) + (4. 0 ≤ x ≤ L. n = 1. (4.. 2. . Then. we have the following equations: ∞ t ≥ 0. u(t. . n = 1. by the orthogonality of the sequence {sin . By substitution (4. where the functions Tn (t). L2 2 L nπx Tn (0) = ϕ(x) sin = Bn .. x) by its decomposition (4. L 0 ≤ x ≤ L.. .44) to the equations (4. 2. (4. Tn (t) sin n=1 ∞ n=1 ∞ ∞ ∞ nπx n2 π 2 nπx nπx = −k 2 Tn (t)sin + fn (t) sin 2 L L L n=1 L n=1 ∞ Tn (t) sin 0 = 0..42) L L 2 0 0 nπx Hence. 2. solving the ordinary diﬀerential equations Tn (t) + k 2 π 2n2 Tn (t) = fn (t).55 In order to ﬁnd the coeﬃcients fn (t). we multiply both sides of mπx (4. 0 ≤ x ≤ L.. x) sin dx. t ≥ 0.. x) = ϕ(x). (4.46) Tn (0) sin n=1 nπx = ϕ(x). L) = 0.... n = 1.45)..we ﬁnd the coeﬃcients Tn (t). we obtain nπx mπx L sin dx = fm (t). are to be determined. 2. n=1 Tn (t) sin nπ = 0. x) = Tn (t) sin (4.45) u(t.. L2 L n=1 ∞ nπx Tn (0) sin = ϕ(x)..41). 2. (4.. nπx .. n = 1. .. we ﬁnd the solution ∞ nπx u(t. .44) L n=1 of the initial boundary value problem L L f(t.

56 By integrating factor method. 0 ≤ x ≤ 1.50) Example 4. x) = sin πx.51) Solution. n = 1. 1) = 0. We compute 1 fn (t) = 2 Bn = 2 1 0 0 [sin πs + sin 2πs] sin nπx ds = 1. t ≥ 0. t ≥ 0. We shall ﬁnd the solution u(t. (4. 2 0 ≤ x ≤ 1. otherewise . x) = L L n=1 n=1 ∞ − t 0 k 2π 2n2 (t − τ ) L2 e fn (τ ) dτ. u(0.52) t ≥ 0. u(t. 2. 2 0. 0 ≤ x ≤ 1. otherewise. t ≥ 0. 0 ≤ x ≤ 1. x) = n=1 Tn (t) sin nπx.. n = 1.39) is given by the following formula: k 2π 2n2 t ∞ nπx nπx L2 sin Bn e + sin u(t. sin πs sin nπs ds 0. 0) = 0. the solution of the initial boundary value problem (4. . − (4. (4. 1. u(t. x) in the series form ∞ u(t.. (4. are determined by the ordinary differential equation T (t) + n2 π 2Tn (t) = 1 1 0 1 [sin πs + sin 2πs] sin nπx ds = fn (t). n = 1. Tn (0) = 2 0 sin πs sin nπs ds = Bn .6 Solve the following initial boundary value problem: ut = uxx + sin πx + sin 2πx..49) Finally. we ﬁnd the solution k π nt + L2 Tn (t) = Bn e − 2 2 2 t 0 k 2 π 2n2 (t − τ ) L2 e − fn (τ ) dτ. where the coeﬃcients Tn (t).

x) and Q = (ξ. 2 2 π 2 + eπ t − 1 T1(0) = 1. T1(0) = 0.. 4π 2 2 (4. (x − ξ)2 (x − ξ) 1 =[ ]e 4((t − η) 5 − 2(t − η)3/2 2 4(t − η) . as the function of the variables ξ and η. n = 2. T2 (t) + 4π T2 (t) = 1.52) for n = 1. Tn (t) = 0. 2 4π 2 π 2 eπ 4. (4. Also.. Tn (t) + n2 π 2Tn (t) = 0. Also. 2. ξ. we ﬁnd ut = [ and uxx (x − ξ)2 4(t − η)5/2 2 (x − ξ)2 1 − ]e 4((t − η) 2(t − η)3/2 − − So that ut = uxx. .. x) = n=1 Tn (t) sin nπx = T1 (t) sin πx + T2(t) sin 2πx 2 2 π 2 + eπ t − 1 1 − e−4π t = sin πx + sin 2πx. we note that U(t. we ﬁnd n = 1. η) given for η < t.57 Solving the ordinary diﬀerential equation (4. T1 (t) + π T1(t) = 1. − Let us consider the heat equation The fundamental solution of the heat equation is given by the formula 1 U(x − ξ. at ﬁxed t and x. x.54) (x − ξ)2 e 4(t − η) . η) satisﬁes the conjugate heat equation uη + uξξ = 0. Finally. one can check that the fundamental solution U(t. t − η) = √ t−η (4.53) n ≥ 3. η) satisﬁes the heat equation as the function of the variables t and x at ﬁxed ξ and η.55) Let us note that the fundamental solution is the function of two points P = (t. we obtain the solution ∞ u(t.6 Fundamental Solution for the Heat Equation ut = uxx. ξ. T1 (t) = 2 π 2 eπ 2 1 − e−4π t T2(0) = 0. Indeed. x. T2 (t) = .

x2) − Q(c. x2)dx2 − Q(c. c)]dx1 ∂R = − R P (x1. we compute (see picture) R ∂P dx1 dx2 = ∂x2 b a dx1 b a d c ∂P dx2 = ∂x2 a b a b [P (x1. d) − P (x1 .56) where the rectangle R = {(x1.56) . x2) R ∂Q ∂P − ]dx1 dx2 = ∂x1 ∂x2 ∂R P dx1 + Q dx2 . (4.7 Fundamental Formulae Green’s Formula. c)dx1 = − d P dx1 ∂Q dx1 dx2 = ∂x1 = d c d c dx2 ∂Q dx1 = ∂x1 d c c [Q(b. c ≤ x2 ≤ d}. d)dx1 − b a P (x1 . Below. Proof. x2) and Q(x1. we present Green’s formula in its simplest form for two continuosly diﬀerentiable functions P (x1 . with the boundary ∂R. x2)]dx2 ∂R Q(b.58 4. x2)dx2 = Q dx2 Hence. x2) : a ≤ x1 ≤ b. We shall show that R R ∂P dx1 dx2 = − P dx1 ∂x2 ∂R ∂Q dx1 dx2 = Q dx2 ∂x1 ∂R Indeed. we obtain Green’s formula (4.

x2(ξ))dξ The ﬁrst fundamental formula for the heat equation Let us note that the following identity holds: vF (u) − uG(v) = ∂ ∂u ∂v ∂(uv) [v −u ]− . x2(ξ))v(a. Green’s formula. ∂x ∂x ∂x ∂t . d)v(x1 (ξ).57) uv ds R R ∂R R R ∂R we obtain the formula of integration by parts in two variables R u R ∂v = ∂x1 ∂v u = ∂x2 ∂R uv ds − uv ds − R ∂R R ∂u dx1 dx2 ∂x1 ∂u v dx1 dx2 ∂x2 v (4. c)v(x1(ξ). c)dξ + u(x1(ξ). x2 (ξ))v(x1(ξ).58) Here the line integral along the boundary ∂R of the rectangle R is ∂R u v ds = ∂R u(x1(ξ). d)dξ + u(b.59 t d T ' ∂R T c R E c 0 a b E x The rectangle R with the boundary ∂R For Q = u v and P = 0. x2(ξ))v(b. x2 (ξ)) (x1 (ξ))2 + (x2(ξ))2 dξ d b a d which becomes ∂R u v ds = + b a c b u(x1(ξ). from. x2(ξ))dξ u(a. by the identities ∂uv dx1 dx2 = ∂x1 ∂uv dx1 dx2 = ∂x2 ∂v ∂u +v dx1 dx2 = ∂x1 ∂x1 ∂v ∂u u +v dx1 dx2 = ∂x2 ∂x2 u uv ds (4.

η) − u(ξ. (4. ξ.63) v(t. η)dξ = 0. 4) = 2. η) for δ < η. (4. 0) = 1.62) 4. x. η)]dη+ √ 2 πu(t.60) The second fundamental formula for the heat equation Substituting into the ﬁrst fundamental formula. in the limit when δ− > 0. t − δ)e [uξ (ξ. ξ. . η)U(t. u(t. x. ξ. G(v) = + ∂x2 ∂t ∂x2 ∂t Integrating by parts both sides of the above identity (see formula integration by parts (4. η = t − δ. (t. η)]dη ∂R ∂R δ + u(ξ.57)). η)Uξ (t. x) : 0 ≤ t ≤ T. x. if u and v satisfy the equations ∂u ∂ 2 u − 2 = 0. x. we obtain (x − ξ)2 dξ 4δ √ = u(ξ. η)U(t. ξ. 0 ≤ x ≤ 4. 0 ≤ x ≤ L} we obtain the ﬁrst fundamental formula for the heat equation R for [vF (u) − uG(v)]dt dx = ∂R u v dx + (v ∂u ∂v − u ) dt ∂x ∂x (4. x. ∂x ∂x then the ﬁrst fundamental formula for the heat equation becomes ∂R u v dx + (v (4. (t. x) = sin πx + 2 sin 5πx. v(ξ.59) Hence. +u(ξ. η)Uξ t.8 Exercises Question 1. η) − u(ξ. t ≥ 0. x). Solve the initial boundary problem: ut = 9uxx . x) ∈ R. η) dξ. ξ. η)U(t. ∂t ∂x2 ∂u ∂v − u ) dt = 0. we obtain the second fundamental formula − ∂R [uξ (ξ. η) = U(t. x. x. ξ.61) Hence. t ≥ 0 u(0.60 ∂ 2u ∂u ∂ 2v ∂v F (u) = − . η)U(t. 0 ≤ x ≤ 4. in the rectangle R = {(t. ξ. ∂t ∂x ∂v ∂ 2 v + = 0. x) out of closed R (4.

Solve the initial boundary problem: ut = 16uxx − 3u.61 Question 2. (4. 1 < x ≤ 2. u(0. 2) = 0. t ≥ 0 u(t. t ≥ 0. .64) x. 1. u(t. x) = 0 ≤ x ≤ 2. 0 ≤ x ≤ 1. 0) = 1.

62 .

r r y π where r2 = x2 + y 2 and θ = arctan . Laplace’s equation takes the following form: 1. ∂x2 ∂y 2 ∆u = uxx + uyy = 0.1) where the Laplace’s operator ∆ ≡ 2. x = 0.1 Laplace Equation .2) (r. x = r cos θ. θ) θ 6 E 0 x Polar Coordinates x = r cos θ. y = r sin θ Indeed. y = r sin θ. Laplace’s equation in the polar coordinates (r. x = 0.Chapter 5 Elliptic Equations 5. θ = . ∂x2 ∂y (5. In two variables x. 1 1 ∆u = urr + ur + 2 uθθ = 0. we can transform Laplace’s equation from Cartesian coordinates 63 . r = 0. ∂2 ∂2 + 2. x 2 yT r s (5. y ∆u = ∂ 2u ∂ 2u + = 0. θ).

θ. r r . x = r cos θ.3) 4. y = sin θ. Laplace’s equation in the three cylindrical coordinates r. we ﬁnd 1 1 ∂ 2u ∂ 2u ∂ 2u urr + ur + 2 uθθ = ( 2 cos2 θ + 2 sin2 θ + 2 sin θ cos θ) r r ∂x ∂y ∂x∂y 1 ∂u ∂u ( cos θ + sin θ) + r ∂x ∂y 1 ∂u ∂ 2u ∂ 2u + 2 (−r sin θ + r2 2 cos2 θ − r2 sin θ cos θ) r ∂y ∂y ∂x∂y 1 ∂u ∂ 2u ∂ 2u + 2 (−r cos θ + r2 2 sin2 θ − r2 sin θ cos θ) r ∂x ∂x ∂x∂y ∂ 2u ∂ 2u = + ∂x2 ∂y 2 3. (5. with r = 0. z = z 1 1 ∆u = urr + ur + 2 uθθ + uzz = 0. ∂x2 ∂y 2 ∂z 2 ∆u = uxx + uyy + uzz = 0. y) = u(r cos θ. y. u(x. (5.64 to Laplace’s equation in polar coordinates by the following computations: x = r cos θ y = r sin θ. z. z ∆u = ∂ 2u ∂ 2u ∂ 2u + + = 0.4) . r sin θ) ∂u ∂u ∂u = cos θ + sin θ ∂r ∂x ∂y ∂ 2u ∂ 2u ∂ 2u ∂ 2u = cos2 θ + 2 sin2 θ + 2 sin θ cos θ ∂r2 ∂x2 ∂y ∂x∂y ∂u ∂u ∂u = r( cos θ − sin θ) ∂θ ∂y ∂x ∂u ∂ 2u ∂ 2u ∂ 2u = −r sin θ + r2 2 cos2 θ − r2 sin θ cos θ ∂θ2 ∂y ∂y ∂x∂y ∂u ∂ 2u ∂ 2u −r cos θ + r2 2 sin2 θ − r2 sin θ cos θ ∂x ∂x ∂x∂y Now. Laplace’s equation in three variables x.

x 5. y) of Laplace equation uxx + uyy = 0. (5.65 z T s (r. x = 0.6) . θ. r = 0. x © Spherical coordinates x = r sin ϕ cos θ. y = r sin ϕ sin θ. Cylindrical coordinates x = r cos θ. θ. z = r cos ϕ. y = r sin ϕ sin θ. z y where r2 = x2 + y 2 + z 2 . ϕ. Dirichlet boundary value problem Find the solution u(x. θ. cos ϕ = . (x. ϕ) 0 y d d d E .5) z T r ϕ d θ d s (r. z = z y where r2 = x2 + y 2. y = r sin θ.2 Boundary Value Problems for Laplace Equation The following three types of the boundary value problems are considered: 1. 1 cotϕ 1 2 ∆u = urr + ur + 2 uϕϕ + 2 uφ + 2 2 uθθ = 0. with x = r sin ϕ cos θ. θ = arctan . r r r r sin ϕ . tan θ = r x 5. z) x © d θ d 0 d d r y d E . z = z. z = r cos ϕ. Laplace’s equation in the spherical coordinates r. y) ∈ Ω (5.

Third kind boundary problem Find the solution u(x. t)u = φ(x. we note that u(x. we shall denote by Ω a bounded domain with the boundary ∂Ω. y) is at some interior point . (5.66 in the domain Ω. (x. y) is not on the boundary ∂Ω. y) ∈ Ω in the domain Ω. ∂n (x. Let u(x. y) ∈ ∂Ω. By contradiction to the thesis. y) attains its maximum in the bounded and closed domain Ω as a continuous function. y). y) + B(x. (x. then the function u(x. If u(x. y) is a given function on the boundary ∂Ω. Here φ(x. y) > 0 and µ(x. y) ∈ Ω in the domain Ω. y) ∂u(x. where A2(x. which satisﬁes the third kind boundary condition A(x. 2. ∂n du where denotes normal derivative internal to the boundary ∂Ω of the dn domain Ω. Firstly. 5. Below. Then. y) = ψ(x. which satisﬁes the Neumann’s boundary condition (BNC) ∂u(x. y) = φ(x. y) attains its maximum and minimum values on the boundary ∂Ω of Ω. Here ψ(x. y) of Laplace equation uxx + uyy = 0. y) is a given function on the boundary ∂Ω of Ω. y) ∈ ∂Ω. the maximum value of u(x. assume that the maximum value of u(x. 3. y) is a harmonic function in the domain Ω. 1.7) Throughout this chapter. y) ∈ ∂Ω. y) + B 2(x.3 The Maximum Principle for Laplace Equation Every solution of Laplace’s equation is called harmonic function. y). (x. we shall give some of properties of the harmonic functions. Proof. y) are given functions on the boundary ∂Ω. The maximum principle. y). y) be a continuous function in the bounded and closed domain Ω. (x. Neumann boundary problem Find the solution u(x. y) of Laplace equation uxx + uyy = 0. which satisﬁes the Dirichlet’s condition (BDC) u(x.

So that.e. y) on the boundary ∂Ω.y)∈∂Ω (x. say M = u(x0 . which. the assumption that u(x. must be in Ω. y) on the boundary ∂Ω of Ω is equal at most Mb + d2 . and the maximum of v(x. by the deﬁnition of v(x. since the graph of v(x. y)| ≤ u(x.y)∈∂Ω Also.y)∈∂Ω min φ(x. . where d is the diameter of Ω. Conclusion From the maximum principle it follows that every harmonic function u(x. we repeat the proof for the maximum of −u(x. y). y). we have |u(x. for some > 0.67 (x0. (x. y) at (x0. the minimum of u(x. since the value M of v(x. (x. There is also strong maximum principle for harmonic function which we present below. we have vxx + vyy ≤ 0. y) ∈ Ω. y0) is larger than the value of v(x. where Mb is maximum of u(x. be points in Ω. 0 < < (M − Mb )/d2 . Let u(x. y1). y0) > Mb . however. So that. Then v(x0. However. y1). (5. leads to the contradiction. The Strong maximum principle. Suppose that the function u(x. y) is larger than M. y) = u(x. y0) ∈ Ω. we must have vxx ≤ 0 and vyy ≤< 0. (x. y)|. Then u(x. y) + [(x − x0 )2 + (y − y0)2 ]. y)|.y)∈∂Ω . y) ≤ u(x. y0) = u(x0. y) must be also attainable on the boundary ∂Ω of Ω. y) cannot be concave up in the x or y direction at (x1 . i. y) at an interior point.y)∈∂Ω (x. we have vxx + vyy + 2 + 2 = 4 > 0. y) is harmonic on Ω.8) or − max |φ(x. y) be a harmonic function on the domain Ω. y) attains its maximum of u(x. Let the maximum of v(x. as we have seen. y) ≤ max |φ(x. y)| ≤ max |φ(x. where v(x. In order to prove that minimum value of u(x. y0) = M. we have M > Mb + d. the maximum of v(x. y1). y1). Thus. y) attains its maximum or minimum at some interior point of Ω. y) satisﬁes the inequality (x. For suﬃciently small > 0. y) is attanable at the boundary ∂Ω. y) ∈ Ω. y) be attained at (x1 . we have used the assumption that u(x. There may. y) ≤ max φ(x. The above two inequalities contradict one the other. At (x1. For such . y). Here. y) at any boundary point. y) cannot occur on the boundary of Ω. y) must be constant throughout Ω. and not on the boundary. at (x1 . (x. Let us introduce the auxiliary function v(x.

min φ(x. . (x. y) ∈ Ω = {(x. 0 < y < 1}. Consider the following boundary value problem: uxx + uyy = 0. y) = sin πx + cos πy are at points when the partial derivatives are equal to zero. y) ∈ Ω ∪ ∂Ω. y) ≤ max φ(x. y) = sin πx + cos πy. and (x. y) = π sin πy = 0. y) ∈ Ω. Question 1. y) = sin πx + cos πy. u) is determined by the inequality (x. y). . Namely. y) ≤ u(x. ∂Ω 3 and the minimum is at the point ( . we compute (x. 1) Thus. so that φx (x. y) = π cos πx = 0. (5. y) on the boundary ∂Ω of Ω. 0) equal 2 max[sin πx + cos πy] = 2. Solution. By the maximum principle. y) are in the range from −2 to 2. we shall prove the following theorems: .y)∈∂Ω where φ(x. the values of the solution u(x. y) ∈ ∂Ω. (x.9) Find the range of the values of the solution u(x. 1 3 We ﬁnd two boundary points ( . (x. the maximum is at 2 2 1 the point ( . and φy (x. for all points (x. 1) equal 2 min[sin πx + cos πy] = −2.y)∈∂Ω min [sin πx + cos πy]. the range of the values of the solution u(x. ∂Ω Hence.68 where u(x. the minimum and maximum of φ(x. y). for (x. The maximum principle implies uniqueness and continuous dependence of solutions on boundary data. so that the following inequality holds: −2 ≤ u(x. Now. 0) and ( .y)∈∂Ω u(x. y) = φ(x. y) : 0 < x < 2.y)∈∂Ω max [sin πx + cos πy] Clearly. y) ≤ 2.

y)| ≤ M. Let v = u1 − u2 . ∂Ω in Ω. u2 = φ2 on on ∂Ω ∂Ω |u1(x. y) ∈ Ω ∪ ∂Ω.1). Assume that there are two solutions u1 and u2 of the Dirichlet boundary problem (5. y) − u2 (x. for all x2 + y 2 ≤ 1. sin θ) = sin θ + cos 2θ. the maximum principle implies that v ≤ 0 and v ≥ 0 on Ω.2 Let u1 and u2 be the solutions of the Dirichlet boundary value problems ∆u1 = 0 ∆u2 = 0 Then where M = max |φ1 (x. the function v = u1 − u2 is continuous in the closed domain Ω and harmonic in the open domain Ω. we obtain the inequality |u1(x. and harmonic in the open disk r < 1.1 (Uniqueness). y) − u2 (x. Then. y) ≤ max |φ1 (x. Proof. Note that v(x. so that v ≡ 0 on Ω and u1 ≡ u2. Now. The maximum principle then implies that u − v ≤ 0 throughout the disk. Solution. Then. y) − φ2(x. Then.69 Theorem 5. Proof. u ≤ v on the boundary of the disk r ≤ 1. y)|. u1 = φ1 in Ω. the maximum of the harmonic function u − v on the boundary r = 1 must be less than or equal to zero. for all (x. y)| ≤ v(x. y) − φ2(x. . then show that u(x. y)| ≤ M. Example 1. Suppose that u(x. y) ≤ y + x2 − y 2. Since v = 0 on ∂Ω. we have − max |φ1(x. y) is a continuous function on the closed disk r ≤ 1. Theorem 5. y)|. If u(cos θ. ∂Ω ∂Ω Hence. The Dirichlet boundary value problem has at most one solution. . we shall state the theorem on continuous dependence of a harmonic function on its boundary values. y) − φ2 (x. y) = y + x2 − y 2 is a harmonic function with v(cos θ. sin θ) ≤ sin θ + cos 2θ. By the assumption.

y) = φ(x. y) ∈ Ω ∪ ∂Ω . y) (x. If the given function f(x. y).y)∈Ω (x. for (x. The following maximum principle holds: Maximum principle. is stable in the maximum norm.70 5. y). y) ∈ Ω. y). y) satisﬁes the following inequality: |u(x. Proof. y)|]. we can state the theorem on stability of the boundary problem for equation (5. and if there exists a regular function g(x. (5. then the solution u(x. (x.12) where M = ea − 1 for 0 ≤ x ≤ a. y) attains non-negative maximum M at a boundary point.y)∈∂Ω (x. (x.1 . Theorem 5. or ∂ 2u ∂ 2u + = f(x. which satisﬁes the following conditions: 1. y) is continuous in the closed bounded domain Ω. y) ∈ Ω. If f(x.3 (Stability Theorem). 0 ≤ x ≤ a. y)|. the solution u(x. y) ∈ Ω. y) attains its non-positive minimum m at a boundary point. y)| ≤ max |u(x. or if f(x. then the the boundary value problem ∆u = f(x. we can assume that the domain Ω is on the right side of y axis.4 The Maximum Principle for Poisson Equation We shall state the maximum Principe for Poisson’s equation ∆u = f(x. that is. y). y) ∈ Ω.10) with Dirichlet’s boundary condition.10) u(x. that is. g(x. (5. y) is given continuous functions in the bounded domain Ω. y) is a regular solution of the Poisson’s equation. We shall prove the theorem on stability using the following lemma: Lemma 5. X∈∂Ω (x. This maximum principle can be proved in the same way as maximum principle for harmonic functions. y)| + M max |f(x.(Comparison Principle. (5. y) ∈ Ω. ∂x2 ∂y 2 where f(x. y) ≤ 0 then u(x. Without any restriction. As a consequence of the above maximum principle. y) ∈ Ω. (x. y) ≥ 0. y) ≥ max |v(x.) If a function v(x.11) (x.

y). hold for (x. Thus. y) ∈ Ω. y) = v(x. y). y) ∈ Ω. y) = v(x. (x. y) for all (x. y) + maxΩ |f(x. (x. y) + g(x. we have the inequalities −g(x. y) ≤ 0. We shall show that the function g(x. y) − ∆g(x.y)∈Ω (x. y) = ∆v(x. for all (x. y) ≥ f(x. y) − g(x. y) ≥ 0. This assumption can be satisﬁed by a linear translation of Ω in x direction. x is a ﬁxed value on x axis. we assume that the domain Ω lies in the right side of x axis. z1(x. y)|. y) Proof of lemma. where x ≤ x. ∆z1(x. End of lemma proof. x ≥ 0. We consider the function g(x. by the maximum principle z1(x. z2(x. y) + [ex − ex ] max |f(x. satisfy the inequalities z1(x.71 2. Hence. y) ∈ Ω. y) ∈ Ω ∪ ∂Ω. By assumption 2. We note that. y)| ≤ 0 z2(x. y) ≥ max |f(x. y) ≥ f(x. |v(x. we shall show that the functions z1(x. y) − g(x. y) ∈ Ω ∪ ∂Ω. y) = v(x. and z2(x. y) + ∆g(x. y) ≤ v(x. y) ≥ 0. then for all (x. y) = v(x. y) ∈ Ω ∪ ∂Ω. y) ≤ 0. by the deﬁnition. or |v(x. y) ≤ 0 and z2(x. on the boundary ∂Ω. y) + g(x. y)| ≤ g(x.y)∈∂Ω Ω (x. we estimate . Here. y). y) ≥ 0 for all (x. y)|. y) ∈ Ω ∪ ∂Ω. that is. In order to prove the lemma. y)| ≤ g(x. y)| ≥ 0 ∆z2(x. −∆g(x. y) satisﬁes assumptions 1 and 2 of the lemma. y) = ∆v(x. Indeed. y) − maxΩ |f(x. y) = max |u(x. In order to prove the theorem. y) ≤ g(x.

y)| ≥ max |f(x. (5. y) satisﬁes the boundary condition u(x. y)| + (ea − 1) max f(x. Hence. y) ≥ max |u(x. we obtain the required inequality |u(x. (x. y < 1. y): |u(x. y) ∈ Ω. . (x. From the deﬁnition of g(x. y) ∈ Ω.72 1. y) ∈ Ω ∪ ∂Ω. By the thesis. for all (x. y) ∈ ∂Ω. y). y) = φ0(x.13) Show that the boundary value problem is stable and estimate the solution u(x. −∆g(x.when 0 ≤ x ≤ a. y)| + M max |f(x. y)| ≤ max |φ(x. it is clear that g(x. y) Solution. y = ex max |f(x. (x. ∂x2 ∂y 2 u(x. Consider the following boundary value problem: ∂ 2u ∂ 2u + = sin πx + cos πy. the constant M = e − 1 and |u(x.y)∈Ω This inequality means stability of the Dirichlet boundary value problem for the Poisson’s equation. Ω Ω for 0 ≤ x ≤ x. where dsM = ea − 1 is the upper bound of the expression ex − ex . Question 1. y)| ≤ 1 + 2(e − 1). 2.y)∈∂Ω (x. y)|].13) satisﬁes the assumptions of the theorem on stability. y)| (x. (x. y). y) ∈ Ω ∪ ∂Ω. y)| ≤ max |u(x. y) ∈ ∂Ω.y)∈∂Ω for (x. y) = sin πxy. y)| + M max |f(x. we obtain the following estimate of the solution u(x. y)| ≤ max |φ(x. We note that the Poisson’s equation (5. y)| ∂Ω Ω = max | sin πxy| + M max | sin πx + cos πy)| ∂Ω Ω ≤ 1+2M Because. 0 ≤ x ≤ 1 therefore. (x. y) : 0 < x. y)|. y)|. Let us note that if the solution u(x.y)∈∂Ω Ω (x. then we get the following the estimate of the solution |u(x. y) ∈ Ω = {(x. (x.

y)| ≤ max |u(x. and f(x. This inequality means stability of the Dirichlet boundary value problem for the Helmholz-Poisson equation. y)|]. y)| ≤ max |u(x. y)u = f(x. y). y) ≥ 0 and c(x. y < 1.y)∈Ω max |u(x. y) ≤ 0 and c(x. We can prove this theorem in a similar way as theorem on stability for Poisson’s equation. (x. then we get the following priori estimate of the solution u(x. y).73 5. (5. Theorem 5. y) are continuous in the closed bounded domain Ω. y) satisﬁes the boundary condition u(x. y) ∈ Ω. Let us note that if the solution u(x.y)∈∂Ω (x. y) ∈ Ω = {(x. or (x. y)| + M max f(x. then u(x. As a consequence of the above maximum principle. y) attains its non-positive minimum m at a boundary point. y) ∈ Ω. If the given functions c(x. y) ≤ 0 and f(x. ∂x2 ∂y 2 u(x.15) where M = ea − 1 for 0 ≤ x ≤ a.16) . y) = sin πxy. then the solution u(x.4 (Stability Theorem). If f(x. y) ≤ 0. y) attains non-negative maximum M at a boundary point. y) = φ0(x. y) ∈ ∂Ω. (x. (x. X∈∂Ω (x. y)|]. y) ≤ 0 for (x. or if f(x. for 0 ≤ x ≤ a. y) ∈ ∂Ω. Question 1. y) : 0 < x. Consider the following boundary value problem: ∂ 2u ∂ 2u + − 2u = sin πx + cos πy.5 The Maximum Principle for Helmholz-Poisson Equation ∆u + c(x. y) ≤ 0. y) are given continuous functions in the bounded domain Ω. (x. y). y) satisﬁes the following inequality: (x.14) We shall state the maximum Principe for Helmholz-Poisson equation ∂ 2u ∂ 2u + + c(x. (5. (x. where M = ea − 1. ∂x2 ∂y 2 where c(x. The following maximum principle holds: Maximum principle. we state the theorem on stability of the solution of Dirichlet’s boundary problem for equation (5.y)∈Ω (5. y) ∈ Ω. y)| + M max |f(x.14). y) ∈ Ω. then the solution u(x. y)u = f(x. y) |u(x.y)∈Ω (x.

y)| ∂Ω Ω = max | sin πxy| + M max | sin πx + cos πy)| ∂Ω Ω ≤ 1+2M Because. y)| ≤ 1 + 2(e − 1). or X (x) Y (y) =− = −λ.17) This boundary value problem can be solved by the method of separation of variables. (5. Then. y)| ≤ max |φ(x. we obtain X (x)Y (x) + X(x)Y (y) = 0. y) Solution. L2) = ϕ(x). y) = 0. 0 ≤ x ≤ 1. y) = −2 ≤ 0.74 Show that the boundary value problem is stable and estimate the solution u(x. then the constant M = e − 1.16) satisﬁes the assumptions of the theorem on stability. for all (x. y) = X(x)Y (y). and the solution u(x. Indeed. 0) = X(x)Y (0) = 0. By the thesis. u(0. Namely. Hence. 5. y): |u(x. y) satisﬁes the inequality |u(x. X(x) Y (y) X (x) + λX(x) = 0.6 Boundary Value Problem for Laplace’s Equation in a Rectangle Let us consider the following boundary value problem: ∂ 2u ∂ 2u + = 0. 0 < y < L2 }. the coeﬃcient c(x. we have two equations . We note that the Helmholz-Poisson equation (5. 0 ≤ y ≤ L2 . y) ∈ Ω ∪ ∂Ω. by substitution to Laplace’s equation. y)| + M max |f(x. 0) = 0. we obtain the following estimate of the solution u(x. ∂x2 ∂y 2 u(x. From the boundary conditions u(x. y) ∈ Ω = {(x. u(x. y) = u(L1 . (x. let u(x. 0 ≤ x ≤ L1 0 ≤ x ≤ L1 . Y (y) − λY (y) = 0. y) : 0 < x < L1 .

. .18) where the coeﬃcients Bn . we ﬁnd C1 + C2 = 0 and C2 = −C1 .. we observe that the terms of the sequence un (x. y) = X(0)Y (y) = 0 and u(L1. Then. .. that is when y = 0 or x = 0. 2. y) = Bn sin nπx nπy sinh . L1 L1 n = 1. y) = X(L1 )Y (y) = 0 therefore X(0) = X(L1 ) = 0.. X(0) = X(L1 ) = 0 is known and the eigenvalues and eigenfunctions are given by the formulae λn = n2 π 2 L2 1 X (n) (x) = sin nπx . 2. so that the solution nπy nπy − nπy . y) = n=1 Bn sin nπx nπy sinh .. which has the general solution Y (y) = C1 e λn y + C2 e − λn y for arbitrary constants C1 and C2 .. are determined by the boundary condition ∞ nπx nπL2 u(x.. L1 n = 1... for arbitrary constants Bn .. Also. . 2.. 2.. we have u(0. we solve the corresponding equation for Y (y) Y (y) − λn Y (y) = 0. Y (n)(y) = C1(e L1 − e L1 ) = 2C1 sinh L1 Now. we consider the solution in the form of the following series: ∞ u(x. or x = L1. By the condition Y (0) = 0. Let us note the solution of the eigenvalue problem X (x) + λX(x) = 0.. . Then. are harmonic functions and satisfy the homogeneous boundary value condition at the three sides of the rectangle Ω.75 So that Y (0) = 0. . n = 1.. 2. L1 L1 (5.. L2) = Bn sin sinh = φ0 (x). L1 L1 n=1 . n = 1. n = 1..

−π < θ < π. 2 n π sinh 2nπ Hence. θ) = φ(θ). we obtain the following formula for the coeﬃcients Bn ..19) or Bn = n = 1. −π ≤ θ ≤ π. y) = 16 ∞ (1 − (−1)n ) nπx sin sinh 2nπy.76 Multiplying the above identity by sin mπx and integrating from 0 to L1 . (5. . 0) = 0. 0 < y < 4}. we ﬁnd the solution in the form of the following series u(x. and L1 using orthogonality property. 0 < r < R. . 2.22) . rn sin nθ..20) Solution. 0≤x≤2 0 ≤ x ≤ 2.. y) : 0 < x < 2. It is easy to check that the functions 1.. 2 φ0 (s) sin nπs ds. n = 1.21) with the boundary condition u(R. (5. 0 ≤ y ≤ 4. y) ∈ Ω = {(x. Bn sinh 2 nπL2 = L1 L1 L1 L1 0 φ0 (s) sin nπs ds L1 (5. 5. 0 ≤ y ≤ 4.. we compute the coeﬃcients Bn = 2 2 sinh nπ4 2 0 2 u(x.19). rn cos nθ. 4) = x(2 − x).. r r Let us consider the Laplace’s equation in the polar coordinates (5. s(2 − s) sin nπs 16(1 − (−1)n ) ds = 3 3 . nπL2 0 L1 sinh L1 Example.7 Boundary Value Problem for Laplace’s Equation in a Disk 1 1 urr + ur + 2 uθθ = 0. ∂x2 ∂y 2 u(x.. By the formulae (5. u(0. 2. y) = 0. y) = u(2. 3 3 sinh 2nπ π n−1 n 2 0 ≤ x ≤ 2.18) and (5. Consider the following boundary value problem: ∂ 2u ∂ 2u + = 0. L1 (x.

25) θ u(2.24).. 4 Solution. θ) given by formula (5. θ) = a0 + an Rn cos nθ + bn Rn sin nθ = φ(θ).. 4 π −π √ π 1 s 4(−1)n 2 an = n cos cos ns ds = − n . the solution u(r. So. −π < θ < π. These coeﬃcients are given by the formulae 1 π φ(s) ds. We shall ﬁnd the solution u(r. θ). Example. Therefore. 0 < r < 2. for n = 1.. Let us consider the following boundary value problem: 1 1 urr + ur + 2 uθθ = 0.. 2. . .23) satisﬁes the Laplace’s equation in the poplar coordinates (r. 2 π −π 4 Hence. So that. θ) = a0 + an rn cos nθ + bn rn sin nθ. θ) in the form of the following trigonometric series: ∞ 1 u(r. 2 π −π 4 2 π(16n2 − 1) π 1 s bn = n cos sin ns ds = 0.. 2. we apply the boundary condition ∞ 1 u(R.. the solution √ √ 2 2 4 2 ∞ (−1)n u(r. θ) of the boundary value problem for Laplace’s equation in a disk is given by the formulae (5. θ) = − rn cos nθ. 2. we apply the formulae (5. In order to ﬁnd the coeﬃcients an and bn . bn = n φ(s) sin ns ds. we compute the Fourier coeﬃcients of the function φ(θ) = cos πθ given in the boundary condition. n = 1. an and bn are the coeﬃcients of the Fourier series of the given function φ(θ) for −π ≤ θ ≤ π. .24). −π ≤ θ ≤ π.24) π π 1 1 an = n φ(s) cos ns ds.23) and (5. In order to ﬁnd the solution u(r. . (5. By formulae (5. θ).24)..77 are harmonic in the disk with radius R and the center at the origin. 2 n=1 −π ≤ θ ≤ π. π π n=1 2n (16n2 − 1) 1 a0 = π π (5. r r (5.26) . R π −π R π −π for n = 1. a0 = π −π (5. we ﬁnd 4 √ s 4 2 cos ds = ..23) and (5.23) 2 n=1 Clearly u(r. θ) = cos ..

Y ) = We note that (x1 − y1)2 + (x2 − y2 )2. 2 ∂x1 r ∂x2 r2 1 1 2 ∂ 2 ln 1 2(x1 − y1)2 ∂ ln r 1 2(x2 − y2)2 r = − . . y plane except the focus point Y = (y1. Here. and X = (x1 . Now. ∂x2 r 1 Let r(X. ∂x2 ∂x2 1 2 1 . we compute r(X. Y ) 1 1 ∂ ln x1 − y1 r =− r = − x2 − y2 . ∂x1 r ∂r x2 − y2 = . Then the distance of the point Y = (y1. Y ) = ln r(X. = 2− .7. ∂r x2 − y2 = . Then the distance of the point Y from the point X = (x1.78 5. y3) be a ﬁxed point on the R3 space. ∂x2 r ∂r x3 − y3 = . Now. n=2 ∂x2 ∂x2 1 2 ∂ 2u ∂ 2u ∂ 2u + + = 0. ∂ 2u ∂ 2u + = 0. y2 . Y ) = ln . x2. Y )) the whole x. Y ) > 0 and let U(X. x2 ). y2). x1 − y1 ∂r = . ∂x2 ∂x2 ∂x2 1 2 3 (5.27) Let n = 2 and let Y = (y1. let n = 3 and Y = (y1. x2−plane. This function is called fundamental solution of the two dimensional Laplace equation. U(X.1 Fundamental Solution of Laplace Equation We shall ﬁnd the fundamental solution of Laplace’s equation. ∂x2 r2 r4 ∂x2 r r4 11 2 ∂ ln Hence. ∂x3 r (x1 − y1)2 + (x2 − y2 )2 + (x3 − y3)2 . x3). Y ) > 0 is the harmonic function on r(X. y2) be a ﬁxed point on x1 . y2) from a point X = (x1. Y ) = We note that ∂r x1 − y1 = . x2. x3 ) is Therefore. we ﬁnd ∂ 2U ∂ 2U + = 0. ∂x1 r . r(X. we write the Laplace equations in the variables X = (x1. n = 3. x2 ) is r(X. .

y3). x1)ds + ∂x1 ∂Ω ∂x1 ∂u ∂v [v −u ] cos(n. X = (x1. Y = (y1. ∂x2 ∂x2 ∂x2 1 2 3 1 . Then. y3) and let U(X. Y ) = n = 2. ∂Ω is a curve on x1. and for n = 3. dn dn dv du − v )ds. Note that for Here dn n = 2. Y ) = Then. dn dn Ω ∂Ω . r(X. ∂Ω is a surface in the space R3 . the following second Green’s identity holds: Therefore. Let Ω be a bounded domain in R2 with the piecewise smooth closed boundary ∂Ω. n = 3.79 Let r(X.28) du denotes normal inner derivative to the boundary ∂Ω. Green’s identities. Y )) > 0. we have ∂ 2u v 2 dx1 = − Ω ∂x1 ∂ 2u v 2 dx2 = − Ω ∂x2 ∂u ∂v [v −u ] cos(n. x3 ). x2)ds + ∂x2 ∂Ω ∂x2 ∂ 2v u 2 dx1 . U(X. we ﬁnd ∂ 2U ∂ 2U ∂ 2U + + = 0. by adding both sides and moving the term with Laplacian from right to left side. = − 3. Y )) > 0 is the harmonic function on the r(X. Y ) ∂U x1 − y1 ∂U x2 − y2 ∂U x3 − y3 =− . =− 3 2 ∂x1 r ∂x2 r ∂x3 r3 ∂ 2U 3(x1 − y1 )2 1 ∂ 2U 3(x2 − y2 )2 1 ∂ 2u 3(x3 − y3)2 1 = − 3. x3 in the whole space except the focus point Y = (y1 . y2 . Ω ∂x1 ∂ 2v u 2 dx2 . This function is called fundamental solution of the three dimensional Laplace equation. So. =− . x2. (u dn dn ∂Ω (5. The prove of both Green’s identities are similar. we compute 1 . . x2 plane. Proof. x2. Integrating by parts. r(X. y2. (v∆u − u∆v)dσ Ω Ω = ∂Ω (u (v∆u − u∆v)dσ = dv du − v )ds. we obtain the Green’s identity [v∆u − u∆v]dσ = [u dv du − v ]ds. Ω ∂x2 Hence. v ∈ C 2(Ω) be twice continuously diﬀerentiable functions in Ω. = − 3 2 2 2 5 5 5 ∂x2 r r ∂x2 r r ∂x3 r r Hence. Y )) variables x1 . Let u. let us prove the Green’s identity for n = 2. Note that the both functions u and v are not assumed to be harmonic. .

x2). Y ) du(Y ) − U(X. dn dn ∂Ω (5. Every harmonic function u(X) in the bounded domain Ω with a smooth boundary ∂Ω satisﬁes the following formula u(X) = u(X) = 1 2π 1 4π [u(Y ) dU(X. n = 3. Then. Y ) ]dsY . x1) + cos(n.29) ∂Ω du dv [v − u ]ds = 0. ∂Ω dn n = 2. n = 2 ∂Ω dU(X. respectively. Y ) du(Y ) − U(X. dn dn n=2 (5. n = 3. (5.s identities. n = 3. dn dn X ∈ Ω. X ∈ Ω. since the fundamental solution U(X.30) du ds = 0. in Ω0 . dn ∂x1 ∂x2 Here cos(n. corollary 1.32) ∂Ω0 dU du [u − U ]ds = 0.2 Theorem on representation of harmonic functions Representation Theorem. we cannot apply the formula (5.7. For harmonic functions u and v in Ω. where K is a disk when n = 2 or a ball when n = 3. we have [u(Y ) [u dU du − U ]ds = 0.29) to these functions. dn dn n=2 (5. dn dn ∂Ω0 . However.80 where the normal derivative ∂u ∂u du = cos(n. x1) cos(n. x2) are directive cosines between normal vector n and x1 and x2 axes. we obtain next corollary corollary 2. from Green. we can apply the formula to the domain Ω0 = Ω − K. ∂Ω dn 5. . So that. we obtain the following formulae: [v du dv − u ]ds = 0. n = 3. dn ∂Ω dn In particular. Every harmonic function u satisﬁes the following identity: du ds = 0. when v = 1. Y ) has the singular point X = Y ∈ Ω. Y ) ]dsY .31) Proof. Let u be a harmonic function and v = U be the fundamental solution of Laplace equation.

dn dn ∂K Next. r U(X. let us write the above integrals along the boundary ∂Ω of Ω and the boundary ∂K of K.33). x2 . X = (x1. y3) ∈ ∂K r Then. du dU 1 [U − u ]ds = 2 dn dn r ∂Ω Now. y3 ) ∈ ∂K r2 Substituting the above formulae for the normal derivatives into (5.81 Now. n = 3.33) ∂Ω ∂K dU du [u − U ]ds = dn dn ∂Ω dU du [u − U ]ds n = 3. x2. we compute dU(X. Y ) = dn 1 . n = 2. we have [U 1 u ds − r ∂K du ds n = 3. dn dn n=2 (5. Y ) = 1 . Y = (y1. n = 3. x3) ∈ Ω0 .32). X = (x1 . x3) ∈ Ω0 . we obtain ∂Ω [u dU du 1 − U ] ds = dn dn r ∂K u ds − ln 1 r du ds. y2) ∈ ∂K. we compute the normal derivative of the fundamental solution U(X.36) Because of uniform continuity of the harmonic function u(X). ∂K dn n=2 (5. we apply the identity ∂K ∂K u(Y )dsY = ∂K [u(Y ) − u(X)]dsY + u(X) ∂K dsY . Y = (y1. x2 ) ∈ Ω0 . we have [u dU du − U ]ds = dn dn [u dU du − U ]ds. Taking into consideration negative orientation of ∂K with respect Ω. Y = (y1. y2 ) ∈ ∂K. by corollary 2. u(Y )dsY n=2 (5. Y = (y1. we have |u(Y ) − u(X)| < .35) n = 3. X = (x1 . X = (x1. n = 2. Y ) to the boundary ∂K 1 ln . (5.34) dU du 1 [u − U ]ds = 2 dn dn r ∂Ω Hence. ∂K dn ∂Ω ∂u dU 1 − u ]ds = dn dn r ∂K u(Y )dsY . x2) ∈ Ω0 . r 1 . from (5. y2 . y2.

= 2. . Y ) dsY dn ∂Ω dU(X. n = 3 R dn R Hence. ∂K Using (5. = .31). Then. X = (x1. we have | | [u(Y ) − u(X)]dsY | < 2π r [u(Y ) − u(X)]dsY | < 4π r 2 ∂K for n = 2 (5. X ∈ Ω. lim 1 ) ∂K u(Y )dsY = 4πu(X).38) Gauss Mean Value Formula Let Ω = K be the disk (n = 2) or the ball (n = 3) with the radius R and the center at X. we obtain equality (5. we obtain Gauss Mean Value Formula for harmonic functions u(X) = 1 2πR u(Y )dsY . we conclude the following important formula concerning boundary value problems for harmonic functions.31).82 for suﬃciently small r(X. n = 2. Namely. R dn R (5.35).39) 1 dU 1 U(X. Y ). From the representative theorem. let us substitute to (5.37) for n = 3. x3 ) ∈ K.37). Y ) = . n = 3. X ∈ Ω. x2) ∈ K. (5. This ends the proof. Y ∂K u(Y )dsY = 2πu(X) for n = 2. X = (x1. u(X) ≡ 1. Hence by formula (5.40) ∂K 1 u(X) = 4πR2 ∂K u(Y )dsY . n = 3. x2. n = 2. on the surface of K. (5. we compute the limits 1 r−>0 r(X. for n = 3. we have 1 dU 1 U(X. Y ) lim r−>0 r 2 (X. by the representative theorem. we obtain the formula dU(X. Y ) = ln .Y ) dsY dn ∂Ω = 2π. Then. n = 2 = 4π.36) and (5. Then.

83

5.7.3

Green’s Function

. We note that every harmonic function H(X, Y ) of the variable Y , and at ﬁxed X, satisﬁes the identity (see formula (5.29)) [u(Y ) dH(X, Y ) du(Y ) − H(X, Y ) ]ds = 0, dn dn n=2 (5.41)

∂Ω

dH(X, Y ) du(Y ) [u(Y ) − H(X, Y ) ]ds = 0, n = 3. dn dn ∂Ω

Let us choose the harmonic function H(X, Y ) = U(X, Y ) for Y ∈ ∂Ω at ﬁxed X ∈ Ω. Then, the Green’s function is G(X, Y ) = U(X, Y ) − H(X, Y ) By the representative theorem (see formula (5.31)), Green’s function satisﬁes identity u(X) = u(X) = 1 2π 1 4π [u(Y ) dG(X, Y ) du(Y ) − G(X, Y ) ]dsY , dn dn X ∈ Ω, n = 2

∂Ω

dG(X, Y ) du(Y ) − G(X, Y ) ]dsY , X ∈ Ω, n = 3. dn dn ∂Ω (5.42) Because G(X, Y ) = 0, for Y ∈ ∂Ω, X ∈ Ω, therefore, every harmonic function u(X) in Ω, satisﬁes the identity [u(Y ) u(X) = 1 2π u(Y ) dG(X, Y ) dsY , dn X ∈ Ω, n = 2

∂Ω

1 u(X) = 4π

dG(X, Y ) u(Y ) dsY , X ∈ Ω, n = 3. dn ∂Ω

(5.43)

Hence, by the formula (5.43), we arrive at the following theorem: Theorem 5.5 If G(X, Y ) is the Green’s function for Laplace’s equation, then the solution u(X) of the Dirichlet’s problem ∆u(X) = 0, is given by the formula u(X) = 1 2π φ(Y ) dG(X, Y ) dsY , dn X ∈ Ω, n = 2 X ∈ Ω, u(X) = φ(X), X ∈ ∂Ω

∂Ω

1 u(X) = 4π

dG(X, Y ) φ(Y ) dsY , X ∈ Ω, n = 3. dn ∂Ω

(5.44)

84

Green’s function for a disk with Dirichlet’s condition. Let r(0, Y ) < R, be the disk with the radius R and the center at the origin. Let us denote by ρ = 0X, ρ = 0X,

X ρ ¡ ¡

r = Y X,

r = Y X.

96 X ¡ ρs s r ¡r 0 t ¡ rt r R ¡s Y 87

**Disk K One can check that the Green’s function for the disk K is R 1 rρ 1 − ln( . ) = ln , X = 0, r ρ r Rr G(X, Y ) = ln 1 − ln 1 , X=0 r R
**

ln

(5.45)

Green’s function for a ball with Dirichlet’s conditions. Similarly, we construct Green’s function for a ball K(0, R) = {X = (x1, x2, x3 ) : r(0, X) < R}. Then, following the notations given above, we write the Green’s function for the ball R 1 1 − . , X = 0, r ρ r (5.46) G(X, Y ) = 1 1 − , X=0 r R Poisson’s Integral. Let us note that in the case when the domain Ω is a disk K or a ball K, the Green’s function is given by formulae (5.45) and (5.46). Then, the solution u(X) of the Dirichlet’s boundary value problem, in the case , is given by the Poisson’s integral (see Representative Theorem, formulae (5.44)). u(X) = 1 2π φ(Y ) dG(X, Y ) dsY , dn X ∈ K, n = 2

∂K

1 u(X) = 4π

dG(X, Y ) φ(Y ) dsY , X ∈ K, n = 3. dn ∂K

(5.47)

In order to express the solution u(X) in a transparent form, we shall evaluate ∂G(X, Y ) the kernel . Let n = 3 and X = 0. Then, we ﬁnd ∂nY dG(X, Y ) 1 dr R 1 dr =− 2 + . 2. dnY r dnY ρ r dnY (5.48)

85

Because

3

r=

i=1

(yi − xi )2

therefore dr = dnY yi But cos(n, xi ) = − , so that R

3 i=1

(yi − xi )2 cos(n, xi ).

3 1 3 dr 2 = ( xi yi − yi ). dnY rR i=1 i=1 3 3 3 yi , and the expression i=1 i=1

(5.49)

We note that R2 =

xiyi is the inner product of

**the vectors 0X and 0Y . Hence, it implies the identity
**

3

xi yi = Rρ,

i=1

where ω is the angle between the vectors 0X and 0Y . Then, we can write formula (5.49) in the form dr 1 = (ρ cos ω − R). dnY r In a similar way, we obtain the derivative Taking into consideration the proportion dr . dnY (5.50)

R r ρ = = , ρ r R we obtain dr 1 ρ R2 1 = (ρ cos ω − R) = ( cos ω − R) = (R cos ω − ρ). dnY r Rr ρ r From the formulae (5.48), (5.50) and (5.51), we ﬁnd dG(X, Y ) 1 R = − 2 (ρ cos ω − R) + (R cos ω − ρ) dnY r rrρ 1 ρ = 3 [−(ρ cos ω − R) + (R cos ω − ρ)] r R Hence, w obtain the equality dG(X, Y ) R2 − ρ2 = , dnY Rr3 X ∈ K, m = 3. (5.51)

(5.52)

(5.53)

at ﬁxed pointX. y2 = R sin ϕ . the kernel of the Poisson’s integral R2 − ρ2 R2 − ρ2 = 2 . R.54) X ∈ K. ρ. φ and θ of the Green’s function for the disk K. n = 3. the angle ϕ ∈ [0. From the ﬁgure and by cosine formula. we ﬁnd r2 = R2 + ρ2 − 2Rρ cos(θ − ϕ).55) On the ﬁgure. Poisson’s integral takes the following form: u(X) = and u(X) = 1 4πR ∂K 1 2πR ∂K R2 − ρ2 φ(Y )dsY . 2π]. (5. Therefore.86 Therefore. y2) ∈ ∂K. Let us write the Poisson’s integral in the polar coordinates y1 = R cos ϕ. n = 2. (5. r3 X ∈ K. Changing the variable of integration Y = (y1. . . r2 R2 − ρ2 φ(Y )dsY . we present the parameters r. r2 R + ρ2 − 2Rρ cos(θ − ϕ) (5.56) When point Y is moving along the circle K with the radius R.

4 Solution. that is 2π 0 R2 − ρ2 dϕ = 2π.58) Example 1. we get 1 2π R2 − ρ2 1= dϕ. Thus. y) ∈ K = {x2 + y 2 < 4}.57) We can interpret the Poisson integral solution (5.56). (5. 2π 0 R2 + ρ2 − 2ρR cos(θ − ϕ) Hence 2π 0 Example 3. u(x. we compute u(0. Then. Show that the integral of Poisson kernel is equal to 2π. 0 ≤ ϕ ≤ 2π. θ) as a weighted average of the boundary potential φ(θ) weighted by the Poisson’s kernel (5. 1 2π 2π 0 R2 − ρ2 dϕ = 2π.57). (x. The Poisson’s kernel tells just how much weight to assign to each point. θ) = 4 − ρ2 2 dϕ. Let us note that the potential at the center of the circle is given by the formula u(0. y) ∈ ∂K = {x2 + y 2 = 4}.57) as ﬁnding the potential u at (ρ. R2 + ρ2 − 2ρR cos(θ − ϕ) Solution . by formula (5. Evaluate the potential at the center of the circle by Poisson ϕ integral for the boundary given function φ(ϕ) = cos . θ) = 1 2π 2π 0 R2 − ρ2 φ(ϕ)dϕ. we obtain the solution of the boundary value problem given by the Poisson’s integral u(ρ.57). This tells us something about physical systems: namely that the potential at a point is the weighted average of neighboring potentials. y2 = R sin ϕ by (5. 0) = 1 2π 2π 0 φ(ϕ) dϕ. . 4 π Example 2. By formula (5. R2 + ρ2 − 2Rρ cos(θ − ϕ) (5. 0) = 1 2π 2π 0 φ(ϕ)dϕ = 1 2π 2π 0 cos ϕ 2 dϕ = .56).87 to polar coordinates y1 = R cos ϕ. 4 + ρ2 − 2 ∗ 4 cos(θ − ϕ) (x.59) Solution. Solve the following boundary value problem using Poisson integral uxx + uyy = 0. By the Poisson formula. y) = 2. the solution in the polar coordinates is u(ρ.54) and (5. R2 + ρ2 − 2Rρ cos(θ − ϕ) (5. the solution of the boundary problem u(X) = 1 for all X ∈ K. Let the boundary given function φ(ϕ) ≡ 1.

n-th order Bessel function of the ﬁrst kind. θ) = 2 2π 2π 0 4 − ρ2 dϕ = 2. Bessel’s Equation Now. physical condition. y) = 2 for all (x. we shall solve the ordinary diﬀerential Bessel’s equation r2 R + rR + (λ2 r2 − n2 ))R = 0. Θ + n2 Θ = 0. R2(r) = BYn (λr). R(1) < ∞. r r Applying the method of separation of variables let us substitute to Helmholz equation u(r. (5. θ) = 0. (5. y = r sin θ with homogeneous boundary conditions: ∆u(r.59). R(1) = 0. Θ + n2 Θ = 0. θ) = 0. y) ∈ K. we obtain r2 R + rR + (λ2 r2 − n2 ))R = 0.88 Hence. R(1) = 0. that Bessel’s equation has two linearly independent solutions 1. we have also the constant solution u(x. where 0 ≤ θ2π. 4 + ρ2 − 2 ∗ 4 cos(θ − ϕ) In the Cartesian coordinates. 5. 0 < r < 1. 1 1 ∆u(r. n-th order Bessel function of the second kind.62) Bessel s equation (5. θ) ≡ urr + ur + 2 uθθ . We shall solve the following Helmholz equation in the polar coordinates (r. θ). Then. θ) = R(r)Θ(θ). x = r cos θ. As we know from the theory of ordinary diﬀerential equations.8 Helmholz Equation and Eigenvalue Problem. 2.61) . R1(r) = AJn (λr). by formula (5. θ) + λ2 u(r.60) u(1. we compute u(ρ.

y). where knm is the m-th root of Jn (r) = 0. y) = sin πx + sin πy. (x. y) = cos πx y. y). −1 < y < 1}. Finally. (x. y) ∈ ∂Ω. (d) u(x. y) = φ(x. θ) = AJn (knm r). substituting R(1) = 0 into AJn (λr). y) ∈ ∂Ω. y) ∈ ∂Ω. y) ∈ Ω = {(x. (b) u(x. y) ∈ ∂Ω. In other words. (x.9 Exercises Question 1. y) : 0 < x < 1. we must pick the separation constant λ to be one of roots of the equation Jn (r) = 0. y) = cos πx + cos πy.63) Find the range of the values of the solution u(x. y) = sinπxy. we obtain Jn (λ) = 0. (x. Show that the following boundary value problem is stable uxx + uyy − u = f(x. Consider the following boundary value problem: uxx + uyy = 0. y) ∈ Ω = {(x. π 2 π 2 k! 4 k=0 Jn (λr) = ( Since Yn (λr) is unbounded at r = 0. (c) u(x. Namely. (5. in order to be R(r) = 0 on the boundary ∂K of the circle K. y) ∈ ∂Ω. ((x. Question 2. that is λ = knm .64) . we obtain the solution unm (r. there are power series representation of the Bessel functions λr n ∞ (−1)m λr ) ( )2m . which satisﬁes one of the boundary condition (a) u(x. u(x. 2 m=0 m!(n + m)! 2 n−1 2 λr 1 λr (n − k − 1)! (λr)2 k Yn (λr) = ln Jn (λr) − ( )−n [ ] . Next. we choose as our solution R(r) = AJn (λr). 5. (5. 0 < y < 1}.89 Here. (x. y) : −1 < x < 1. (x. we ﬁnd R(r) and λ using the boundary condition R(1) = 1. y).

θ) = cos .90 for given continuous functions f(x.65) Question 4. 0 ≤ y ≤ 2. when f(x. y) ∈ ∂Ω. y) : 0 < x < 1. 0≤x≤1 0 ≤ x ≤ 1. ∂x2 ∂y 2 u(x.66) . Give an priori estimate of 2 2 the solution u(x. (x. u(x. (x. Question 3. (5. 0 < y < 2}. y) = e−x −y . 2 (5. Solve the following boundary value problem: 1 1 urr + ur + 2 uθθ = 0. (x. y) and φ(x. y) ∈ Ω = {(x. y). −π < θ < π. u(0. y) = 0. 2) = sin πx. −π ≤ θ ≤ π. Solve the following boundary value problem: ∂ 2u ∂ 2u + = 0. y) ∈ Ω and φ(x. y) = 0. 0 < r < 4. y). r r θ u(4. 0) = 0. y) = u(1.

65 harmonic functions. 69 Green’s identity. 52 canonical form. 68 general solution. 22 Jacobian. 52 stability theorem. 19 initial boundary problem. 8. 28 fundamental solution. 10. 58 hyperbolic equation. 53 method of separation of variables. 8. 8 Poisson equation. 12 cylindrical coordinates. 51 maximum principle. 8 operator hyperbolic. 19 Dirichlet conditions. 1 Fourier series. 70 polar coordinates. 74 boundary conditions. 11 characteristics equations. 59 Poisson’s integral. 53 heat equation. 8 parabolic equation. 53 91 operator elliptic. 52 D’Alembert solution. 23 Neumann conditions. 8 operator parabolic. 53 elliptic equation. 59 Hopf’s theorem. 57 string equation. 1 Green function. 51 representation theorem. 64 Gauss mean value theorem. 15 Laplace equation. 8 Helmholtz equation. 65 Green’s function. 8 . 8 ﬁrst order PDE. 27 third kind boundary conditions. 9 characteristic equation. 53 wave equation.Index Bessel’s equation. 66 spherical coordinates. 25 boundary value problem.

Sign up to vote on this title

UsefulNot useful- Applied Partial Differential Equations 4E - Haberman
- Shiv Das
- MA1506CHAP8
- review_m3073_2009w
- app_PDE
- ChewMA1506-14 Ch8
- formule-en
- CH5
- 2 Wave Equation
- L-PDE-2011
- MATH 2930 - Worksheet 13 Solutions
- Partial Differential Equation Notes
- hw12s PDE
- 25_3_soln_sep_vars.pdf
- Apps of Pde Chapter3maths3-110725025341-Phpapp02
- Solution Sap de 3
- Problem Set 5 Partial Differential Equations
- Calculus of Two Variables Lecture Notes 2009 Part II p63to97
- 25 3 Solutn Using Sepratn Variabls
- Heat Equation in 1D
- solutions21.3
- PDE Homework Solutions
- Chapter 4 PDE
- Assignment 1
- MA 2211 Unit 4
- Alberto Bressan and Yuxi Zheng- Conservative Solutions to a Nonlinear Variational Wave Equation
- MATH 2930 - Worksheet 12 Solutions
- Solutions partial differential equations
- mmch7
- 25 1 Partial Diffrntl Equatns
- Partial Differential Equations

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd