Professional Documents
Culture Documents
Computer Physics Communications Volume 180 issue 9 2009 [doi 10.1016_j.cpc.2009.04.009] M.M. Rashidi; E. Erfani -- New analytical method for solving Burgers' and nonlinear heat transfer equations and comparison with .pdf
Computer Physics Communications Volume 180 issue 9 2009 [doi 10.1016_j.cpc.2009.04.009] M.M. Rashidi; E. Erfani -- New analytical method for solving Burgers' and nonlinear heat transfer equations and comparison with .pdf
New analytical method for solving Burgers and nonlinear heat transfer equations
and comparison with HAM
M.M. Rashidi , E. Erfani
Engineering Faculty of Bu-Ali Sina University, PO Box 65175-4161 Hamedan, Iran
a r t i c l e
i n f o
a b s t r a c t
Article history:
Received 22 January 2009
Accepted 15 April 2009
Available online 17 April 2009
Keywords:
Differential transform method (DTM)
Burgers equation
Nonlinear differential equation
Homotopy analysis method (HAM)
Fin
In this study, we present a numerical comparison between the differential transform method (DTM) and
the homotopy analysis method (HAM) for solving Burgers and nonlinear heat transfer problems. The
rst differential equation is the Burgers equation serves as a useful model for many interesting problems
in applied mathematics. The second one is the modeling equation of a straight n with a temperature
dependent thermal conductivity. In order to show the effectiveness of the DTM, the results obtained
from the DTM is compared with available solutions obtained using the HAM [M.M. Rashidi, G. Domairry,
S. Dinarvand, Commun. Nonlinear Sci. Numer. Simul. 14 (2009) 708717; G. Domairry, M. Fazeli, Commun.
Nonlinear Sci. Numer. Simul. 14 (2009) 489499] and whit exact solutions. The method can easily be
applied to many linear and nonlinear problems. It illustrates the validity and the great potential of the
differential transform method in solving nonlinear partial differential equations. The obtained results
reveal that the technique introduced here is very effective and convenient for solving nonlinear partial
differential equations and nonlinear ordinary differential equations that we are found to be in good
agreement with the exact solutions.
2009 Elsevier B.V. All rights reserved.
1. Introduction
Most phenomena in real world are described through nonlinear
equations. Nonlinear phenomena play important roles in applied
mathematics, physics and in engineering problems in which each
parameter varies depending on different factors. The importance
of obtaining the exact or approximate solutions of nonlinear partial differential equations (NLPDEs) in physics and mathematics, it
is still a hot spot to seek new methods to obtain new exact or
approximate solutions. Large class of nonlinear equations does not
have a precise analytic solution, so numerical methods have largely
been used to handle these equations. There are also some analytic techniques for nonlinear equations. Some of the classic analytic methods are the Lyapunovs articial small parameter method,
perturbation techniques and -expansion method. In the recent
years, many authors mainly had paid attention to study solutions
of nonlinear partial differential equations by using various methods. Among these are the Adomian decomposition method (ADM),
tanh method, homotopy perturbation method (HPM), sinhcosh
method, HAM, DTM and variational iteration method (VIM).
In 1992, Liao [1] employed the basic ideas of homotopy in
topology to propose a general analytic method for nonlinear problems, namely the HAM. Based on homotopy of topology, the va-
Corresponding author. Tel.: +98 811 8257409; fax: +98 811 8257400.
E-mail address: mm_rashidi@yahoo.com (M.M. Rashidi).
1540
The basic denitions and fundamental operations of the twodimensional differential transform are dened in [2330]. Consider
a function of two variable w (x, y ) be analytic in the domain and
let (x, y ) = (x0 , y 0 ) in this domain. The function w (x, y ) is then
represented by one series whose centre at located at w (x0 , y 0 ).
The differential transform of the function is the form
W (k, h) =
k+h w (x, y )
k!h!
k x h y
1
,
(x0 , y 0 )
w (x, y ) =
(1)
W (k, h)(x x0 )k ( y y 0 )h .
(2)
k=0 h=0
1 k+h w (x, y )
(x x0 )k ( y y 0 )h .
k!h!
k x h y
(x0 , y 0 )
(3)
k=0 h=0
In a real application and when (x0 , y 0 ) are taken as (0, 0), then
the function w (x, y ) is expressed by a nite series and Eq. (2) can
be written as
w (x, y )
=
n
m
W (k, h)xk y h ,
(4)
k=0 h=0
k h
in addition, Eq. (4) implies that
k=m+1
h=n+1 W (k, h )x y is
negligibly small. Usually, the values of m and n are decided by
convergences of the series coecients.
3. Application
3.1. The Burgers equation
Consider the Burgers equation [31]
ut + uu x u xx = 0,
x R,
(5)
1
2
1
2
tanh
1
4
1
2
,
(6)
where w (x, y ) is the original function and W (k, h) is the transformed function. The transformation is called T-function and the
lower case and upper case letters represent the original and transformed functions respectively. Then its inverse transform is dened
as
1
2
1
2
tanh
x
.
(7)
Taking the two-dimensional transform of Eq. (5) by using the related denitions in Table 1, we have
(h + 1)U (k, h + 1) +
k
h
(k r + 1)U (r , h s)U (k r + 1, s)
r =0 s=0
(k + 1)(k + 2)U (k + 2, h) = 0,
(8)
by applying the differential transform into Eq. (7), the initial transformation coecients are thus determined by
Table 1
The operations for the two-dimensional differential transform method.
Original function
Transformed function
w (x, y ) = u (x, y )
w (x, y ) =
w (x, y ) =
w (x, y ) =
u ( x, y )
x
u ( x, y )
y
r + s u ( x, y )
xr y s
W (k, h) = (k + 1)U (k + 1, h)
W (k, h) = (h + 1)U (k, h + 1)
W (k, h) = (k + 1)(k + 2) . . . (k + r )(h + 1)(h + 2) . . . (h + s)U (k + r , h + s)
W (k, h) =
u ( x, y ) v ( x, y )
x
x
u ( x, y ) v ( x, y )
w (x, y ) = x
y
v ( x, y )
w (x, y ) = u (x, y ) x
2 v ( x, y )
w (x, y ) = u (x, y ) x
W (k, h) =
w (x, y ) =
W (k, h) =
W (k, h) =
W (k, h) =
k
h
k
h
k
h
k
h
k
h
r =0
r =0
r =0
r =0
r =0
s=0
U (r , h s) V (k r , s)
r + 1)(h s + 1)U (k r , s) V (r , h s + 1)
s=0 (k r + 1)U (r , h s) V (k r + 1, s)
s=0 (k
r + 2)(k r + 1)U (r , h s) V (k r + 2, s)
U (k, 0)xk =
k=0
1
2
x
8
x3
384
31x9
743178240
x5
15360
17x7
u (x, t )
=
10321920
691x11
653996851200
+ .
(9)
U (k, 0) = 0,
1
2
+
(10)
Table 2
Some values of U (k, h) of Example 3.1.
U (1, 1) = 0
U (2, 2) = 0
U (3, 3) = 0
U (2, 3) = 3!1211
U (3, 2) = 3!1210
U (3, 4) =
17
3!217
3!21719 5
U (k, h)xk t h
k=0 h=0
n
m
1541
16
t 3 x2
t3
t5
3072
491520
8
17t 5 x2
x3
t2x
512
t 2 x3
12888
15728640
384
6144
17t 3 x4
31t 5 x4
x5
2359296
17t 2 x5
1966080
188743680
31t 4 x5
94371840
t4x
49152
17t 4 x3
4718592
tx2
256
tx4
6144
15360
+ .
(11)
Our approximation has one more interesting property, if we expand exact solution (6) using Taylors expansion about (0, 0), we
have the series same as the our approximation (11).
In Fig. 1, we study the diagrams of the results obtained by the
DTM for m = 50, n = 50 in comparison with the HAM [19] and
exact solution (6).
Note that the solution series obtained by the HAM contains the
auxiliary parameter h , which provides us with a simple way to adjust and control the convergence of the solution series. As pointed
Fig. 1. The results obtained by the DTM (m = 50, n = 50) Eq. (11) and the HAM [19] by 9th-order approximate solution, in comparison with the exact solution (6), when
5 t 12, (a) x = 0.25; (b) x = 0.75; (c) x = 1; (d) x = 1.5.
1542
Table 3
Comparison of the exact solution (6) whit the DTM (m = 50, n = 50) and the HAM (h = 0.6) by 9th-order approximate solution in case of x = 0.5, x = 1.5 and 0 t 12.
x = 0.5
1
2
3
4
5
6
7
8
9
10
11
12
x = 1.5
Exact
HAM (h = 0.6)
DTM
Exact
HAM (h = 0.6)
DTM
0.500000000
0.562176500
0.622459331
0.679178699
0.731058578
0.777299861
0.817574476
0.851952801
0.880797077
0.904650535
0.924141819
0.939913349
0.499967668
0.562186654
0.622618884
0.679494353
0.731289069
0.776880106
0.815694315
0.847843276
0.874236662
0.896665720
0.917848146
0.941425185
0.500000000
0.562176500
0.622459331
0.679178699
0.731058578
0.777299861
0.817574476
0.851952801
0.880797060
0.904646851
0.923677794
0.901792839
0.377540668
0.437823499
0.500000000
0.562176500
0.622459331
0.679178699
0.731058578
0.777299861
0.817574476
0.851952801
0.880797077
0.904650535
0.377496222
0.437785782
0.500133672
0.562684851
0.623385326
0.680114881
0.730852749
0.773875815
0.807988448
0.832782589
0.848926292
0.858478394
0.377540668
0.437823499
0.499999999
0.562176500
0.622459331
0.679178699
0.731058578
0.777299861
0.817574479
0.851953323
0.880854270
0.908702365
Fig. 2. The comparison of the errors in answers results by the DTM (m = 50, n = 50)
and the HAM (h = 0.6) by 9th-order approximate solution at t = 10.
in [19], the valid region of h in this case is 1.6 < h < 0.4. When
h = 1, the solution by the HAM is as the same solution series obtained by HPM, which proposed in 1998 by Dr. He [33]. Therefore,
the HAM logically contains the HPM. We can nd that the best
value of h in this case is 0.6.
In Table 3, we present a numerical comparison between the
HAM (h = 0.6) by 9th-order approximate solution, the DTM (m =
50, n = 50), and exact solution (6) in this case of x = 0.5, x = 1.5
and 0 t 12.
Table 3 indicates that the results obtained by the DTM for the
case of 0 t 8 have nine digits precision whit the exact solutions.
In Figs. 2 and 3, we present the comparison of the errors
in answers results by the DTM (m = 50, n = 50) and the HAM
(h = 0.6) by 9th-order approximate solution at t = 10 and t = 11
respectively for the case of 1 x 1. Considering these two gures, we nd out that errors of the DTM are very less than those
of the HAM even for large t.
3.2. Rectangular purely convective n with temperature dependent
thermal conductivity
Consider a straight n with a temperature-dependent thermal
conductivity, arbitrary constant cross-sectional area A c , perimeter
P , and length b (see Fig. 4). The n is attached to a base surface of
temperature T b , extends into a uid of temperature T a and its tip
is insulated. The one-dimensional energy balance equation is given
Ac
d
dx
k( T )
dT
dx
P h( T b T a ) = 0.
(12)
Fig. 3. The comparison of the errors in answers results by the DTM (m = 50, n = 50)
and the HAM (h = 0.6) by 9th-order approximate solution at t = 11.
k( T ) = ka 1 + ( T T a ) ,
(13)
where ka is the thermal conductivity at the ambient uid temperature of the n and k is the parameter describing the variation of
the thermal conductivity.
Employing the following dimensionless parameters
T Ta
Tb Ta
= ,
b
= ( T b T a ),
h P b2
ka A c
1/ 2
.
(14)
d2
d 2
+
d
d
2
2 = 0,
(15a)
1543
Table 4
Comparison of the dimensionless temperature errors in answers results by the DTM (k = 15) and the HAM by 12th-order approximate solution for the case of constant
thermal conductivity, i.e. = 0.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.
d
d
=1
= 0.5
DTM
HAM (h = 1)
HAM (h = 0.92)
DTM
HAM (h = 1)
HAM (h = 0.791)
0.000000000
0.000000000
0.000000000
0.000000000
1.1102 1016
0.000000000
0.000000000
1.1102 1016
1.1102 1016
1.1102 1016
0.000000000
1.3600 1012
1.3433 1012
1.2936 1012
1.2121 1012
1.1013 1012
9.6378 1013
8.0147 1013
6.2105 1013
4.2377 1013
2.1505 1013
6.9944 1015
4.8849 1015
4.6629 1015
4.4408 1015
3.1086 1015
2.9976 1015
2.3314 1015
5.2180 1015
6.7723 1015
1.2212 1014
3.9968 1015
7.7715 1015
2.0095 1014
2.0317 1014
2.0539 1014
2.0983 1014
2.1649 1014
2.2648 1014
2.3869 1014
2.5091 1014
2.5979 1014
2.3092 1014
0.0000000000
1.7793 105
1.7574 105
1.6922 105
1.5854 105
1.4395 105
1.2582 105
1.0458 105
8.0782 106
5.4985 106
2.7835 106
0.000000000
4.5732 1011
4.6543 1011
4.7738 1011
4.5748 1011
3.5230 1011
1.0373 1011
3.2824 1011
9.1056 1011
1.4602 1010
1.4952 1010
5.3290 1015
= 0 at = 0,
=1
at = 1.
(15b)
(15c)
(k + 1)(k + 2)(k + 2)
+
k
(k r + 2)(k r + 1)(r )(k r + 2)
r =0
k
+
(r + 1)(k r + 1)(r + 1)(k r + 1)
r =0
2 (k) = 0,
(16)
( ) =
(k) .
(17)
k=0
(1) = 0,
( )
=
m
(k) k = +
k=0
(2 1)(14 1) 6 6
6!(1 + )5
(2 1)(1 + 4 (112 19)) 8 8
+
8!(1 + )7
(18)
by assuming
(0) = ,
(19)
moreover, substituting Eqs. (18) and (19) into Eq. (16) and by recursive method we can calculating another values of (k), some
results are listed as follows
(k) = 0,
if k = 3, 5, 7, . . . ,
2
,
2!(1 + )
(1 2 ) 4
,
(4) =
4!(1 + )3
(2 1)(14 1) 6
,
(6) =
6!(1 + )5
(2 1)(1 + 4 (112 19)) 8
,
(8) =
8!(1 + )7
(2) =
..
..
+ .
(21)
( )analytical =
(20)
2
(1 2 ) 4 4
2 +
2!(1 + )
4!(1 + )3
e + e
e + e
(22)
In Table 4, numerical comparison of the dimensionless temperature errors in answers results by the DTM and the HAM [22] were
compared for = 0.5 and = 1. Note that the solution series
obtained by the HAM contains the auxiliary parameter h , which
inuence its convergence region and rate. We should therefore focus on the choice of h by plotting of errors in answers results by
the HAM for some values of h .
As pointed in [22], the valid region of h for the case of = 0.5
and constant thermal conductivity is 1.3 < h < 0 and for the case
of = 1 and constant thermal conductivity is 1.6 < h < 0.1.
In Fig. 5, we present the comparison of the errors in answers
results by the DTM (m = 15) and the HAM (h = 0.78, h = 0.785
1544
[8] H. Bateman, Some recent researches on the motion of uids, Mon. Weather
Rev. 43 (1915) 163170.
[9] J.M. Burgers, Mathematical examples illustrating relations occurring in the theory of turbulent uid motion, Trans. R. Neth. Acad. Sci. Amst. 17 (1939) 153.
[10] J.M. Burgers, A Mathematical Model Illustrating the Theory of Turbulence,
in: Advances in Applied Mechanics, vol. I, Academic Press, New York, 1948,
pp. 171199.
[11] N. Su, J.P.C. Watt, K.W. Vincent, M.E. Close, R. Mao, Analysis of turbulent
ow patterns of soil water under eld conditions using Burgers equation and
porous suction-cup samplers, Aust. J. Soil Res. 42 (1) (2004) 916.
[12] N.J. Zabusky, M.D. Kruskal, Interaction of solitons in a collisionless plasma and
the recurrence of initial states, Phys. Rev. 15 (1965) 240243.
[13] P.F. Zhao, M.Z. Qin, Multisymplectic geometry and multisymplectic Preissmann
scheme for the KdV equation, J. Phys. A 33 (18) (2000) 36133626.
[14] H. Brezis, F. Browder, Partial differential equations in the 20th century, Adv.
Math. 135 (1998) 76144.
[15] P.C. Jain, D.N. Holla, Numerical solution of coupled Burgers 1 equations, Int. J.
Numer. Methods Eng. 12 (1978) 213222.
[16] F.W. Wubs, E.D. de Goede, An explicitimplicit method for a class of timedependent partial differential equations, Appl. Numer. Math. 9 (1992) 157181.
[17] C.A.J. Fletcher, Generating exact solutions of the two-dimensional Burgers
equation, Int. J. Numer. Methods Fluids 3 (1983) 213216.
[18] M.A. Abdou, A.A. Soliman, Variational iteration method for solving Burgers and
coupled Burgers equations, J. Comput. Appl. Math. 181 (2) (2005) 245251.
[19] M.M. Rashidi, G. Domairry, S. Dinarvand, Approximate solutions for the Burgers and regularized long wave equations by means of the homotopy analysis
method, Communications in Nonlinear Science and Numerical Simulation 14
(2009) 708717.
[20] D.Q. Kern, D.A. Kraus, Extended Surface Heat Transfer, McGraw-Hill, New York,
1972.
[21] A.A. Aziz, E. Hug, Perturbation solution for convecting n with variable thermal
conductivity, J. Heat Transfer 97 (1995) 300310.
[22] G. Domairry, M. Fazeli, Homotopy analysis method to determine the n efciency of convective straight ns with temperature-dependent thermal conductivity, Communications in Nonlinear Science and Numerical Simulation 14
(2009) 489499.
[23] C.K. Chen, S.H. Ho, Solving partial differential equations by two-dimensional
differential transform method, Appl. Math. Comput. 106 (1999) 171179.
[24] M.J. Jang, C.L. Chen, Y.C. Liu, Two-dimensional differential transform for partial
differential equations, Appl. Math. Comput. 121 (2001) 261270.
[25] A.H. Hassan, Different applications for the differential transformation in the
differential equations, Appl. Math. Comput. 129 (2002) 183201.
[26] F. Ayaz, On two-dimensional differential transform method, Appl. Math. Comput. 143 (2003) 361374.
[27] F. Ayaz, Solution of the system of differential equations by differential transform method, Appl. Math. Comput. 147 (2004) 547567.
[28] A. Kurnaz, G. Oturnaz, M.E. Kiris, n-Dimensional differential transformation
method for solving linear and nonlinear PDEs, Int. J. Comput. Math. 82 (2005)
369380.
[29] A.H. Hassan, Comparison differential transform technique with Adomian decomposition method for linear and nonlinear initial value problems, Chaos
Solitons & Fractals 36 (2008) 5365.
[30] F. Kangalgil, F. Ayaz, Solitary wave solutions for the KdV and mKdV
equations by differential transform method, Chaos Solitons & Fractals,
doi:10.1016/j.chaos.2008.02.009.
[31] E. Hopf, The partial differential equation ut + uux = uxx, Commun. Pure. Appl.
Math. 3 (1950) 201230.
[32] P.G. Drazin, R.S. Johnson, Soliton: an Introduction, Cambridge University Press,
Cambridge, 1989.
[33] J.H. He, Homotopy perturbation technique, Comput. Meth. Appl. Mech.
Eng. 178 (3/4) (1999) 257262.