Application of Adomian's Approximations To The Navier-Stokes Equations in Cylindrical Coordinates

Pergamon
S0893-9659(96)00061-4
Appl. Math. Lett. Vol. 9, No, 4, pp. 109-113, 1996 Copyright@1996 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0893-9659/96 $15.00 + 0.00
A p p l i c a t i o n of A d o m i a n ' s A p p r o x i m a t i o n s to the Navier-Stokes Equations in Cylindrical Coordinates

K. HALDAR
Physics and Applied Mathematics Unit, Indian Statistical Institute 203 Barrackpore Trunk Road, Calcutta 700 035, India
(Received June 1994; accepted August 1995)

Abstraet--Adomian has applied the decomposition method to the Navier-Stokes equations in the Cartesian coordinate system. The present analysis gives the application of the decomposition method to the Navier-Stokes equations. K e y w o r d s - - N a v i e r - S t o k e s equations, Decomposition method, Adomian's polynomials, Infinite series solution, Rapidly convergent.
1. I N T R O D U C T I O N The basic dynamical equations of fluid mechanics are represented by the Navier-Stokes equations. These equations are nonlinear partial differential equations which govern the flow field of air around aircraft, in ramjet, blood circulation in the cardiovascular system in the human body, and in many other disciplines. Since the Navier-Stokes equations are nonlinear in character, it is not possible to solve these equations analytically. In order to solve these equations we need some simplifications, such as linearization or assumptions of 'weak' nonlinearity, small fluctuations, etc. Otherwise, we use many traditional numerical techniques which result in massive computations. Our objective is to find out continuous analytic solutions without massive outprints and restrictive assumptions which change the physical problem into a mathematically tractable problem. The solution of the reduced problem can be obtained analytically, but this solution is not consistent with the solution of the original problem. Recently, a powerful method which is called Decomposition Methodology developed by Adomian [1-3] can provide analytic approximations to a wide class of nonlinear ordinary differential equations, systems of differential equations, partial differential equations, and systems of partial differential equations, and this method demands to be parallel to any modern supercomputer. Adomian [3] has applied the decomposition method to the Navier-Stokes equations in the Cartesian coordinate system. The present analysis gives the application of the decomposition method to the Navier-Stokes equations in cylindrical coordinates by means of which the steady two-dimensional irrotational fluid flow problems in tubes of nonuniform circular cross-sections can be studied. 2. A N A L Y S I S Consider steady, two-dimensional motion of a viscous fluid in a tube. The equations of motion which govern the flow field in the tube are the Navier-Stokes equations in cylindrical coordinates~
Typeset by A A.,IS-TEX 109
ii0
K . HALDAR
and they are given by _ _1 _ _Op + u ( 0 2 u + - 1 0 u + c92u'~ Ou Ou ~-5-;z + v ~ = p Oz \ Or~ r ~ "5-jz2 / ' (1)
OV+v~_rr = U-~z
lop (02v 10v 02v p Or -- 2 - + v \0r2 + r ~r + Oz
v2) -r2 '
(2)
assuming that the rotational fluid motion is neglected. The continuity equation is
;. Or
1 c9 (rv) + Ou
~=0.
(3)
Here (u, v) are the components of fluid velocity in the axial coordinate x and radial coordinate r, respectively, p is the fluid pressure, p is the fluid density and u is the kinematic viscosity of fluid. If we introduce the stream function defined by u 1 0 r Or'
v = -
1 0 - r Oz'
(4)
the continuity equation is identically satisfied. If we eliminate p between (1) and (2), and then use the relations (4), we have the dynamical equation of motion in terms of as
1 0(L,) 2 0 L 9 = vL2~#, 7 o(r, z----7- - r-~ - f f z

where L = ~ r - ( l / r ) . ~r + ~)-~"
o02 o02
(5)
02 o0~ CASE 1. Let L1 -- ~ - ( l / r ) o . Then the operator L becomes L = L1 + ~ and equation (5) takes the form 02 L~ = u - l g 040z 4 2 ~Z2 (L1), (6)
where N is the nonlinear term denoted by
N =- 1 . 0 ( L ,
0(r, z)
) _ 2.0_____.L. r 2 Oz
(7)
Operating with the appropriate inverse L~-2 on (6), we have [1,2] 02 (04~ ~p(z, r) = ~Po(Z, r) + u - I L ~ 2 ( N ) - 2L~-20-~z2(LI) - L~-2 \ Oz 4 ] ,
(s)
where ~P0 is the solution of the homogeneous equation L2g)0 -- 0 subject to certain boundary conditions. In order to determine the inverse operator of L1, we consider the equation [3] L1 = 0.
o02
(9)
If we define L~r --- ~ and L~ = ~-;,o0then the operator L1 becomes L1 = Lr~ - (1/r)Lr and equation (9) takes the form 1 Lrr~# - - " L ~ = 0. (10)
r
Solving for the linear terms LrrW and LrF, we have L r r = _1. L~, r L ~ = rL~r. (11) (12)
Navier-Stokes Equations
Operating with Lr~ on (11) and L~-1 on (12), we can write = 1 + L~-~ ( 1 L ~ ) , = 2 +
1ll
(13) (14)
L~I(rL~),
where 1 and 2 are the solutions of the homogeneous equations L r r = 0 and L~ = 0. The inverse operators L~-1 and L~-1 are defined by
L~# = f f [.] dr dr,

JJ
L~-1
= / [ . ] dr.
J
/,
(15)
Adding (13) and (14), and then dividing by 2, we have
= o +-~ L-~
where 0 = (1/2)(1 + 2). Then
Lr
+L;I(rLr~) ,
(16)
1/~1=~ [L~J (l'Lr) ~-L~-l(rLrr)] o,

(17)
If the quantity within the bracket is denoted by S, then we can write
1 = ~So,
2 = ~S20,
(18) 1
~n+~ = 2n+l Sn+1o,

and
: ~
n:O
Cn = ~
r',,:O
---- ( L r r -
" SnO -~
n:O
-~ L-rJ
(1)
Lr
Q- L < l ( r L r r
o,
(19)
so that the inverse L{-1
(1/r)Lr) -1 has been identified as
(20)
n=O
Now we can return to equation (8) and write the parametrized form of it as [2]
~p(z,r)
[ o(z,r) + A v-tLT2(N~p)
2LT2b--~z2(Ltt~) - L12 \-b-~z4) ] .
02
(04~'~I
(21)
112
K. HALDAR
Let and the nonlinear term N be decomposed into the following forms:
oo
(22)
n=0 oo
g = E
n-----O
Andn'
(23/
where An are special polynomials of Adomian [1,2] to be discussed later. The parameter A introduced here is not a perturbation parameter, and it is used for grouping the terms only. Then equation (21) takes the form EAnCn=~bo+A
n=O
u-IL~ 2
n=O
Ang,~-2n7 ~
LIEAnCn
n=O
-L12
/
~z4EAnCn
n=O
(24)
Equating the terms of like power of A, we have 02 1 = v - l L 1 2 A o - 2L1-2 ~ 2 = v - t L t 2 A 1 - 2 L t 2 ~

CO2
(Lie0)
coao L12 COz4 ,

cO41
(L11) - L~ 2 cOz4 ,
(25)
cO~ cO4n n+1 = v - I L ~ 2 A n - 2L12 -ff~z2 (Lien) - L~-2 cOz4 The polynomials A0, A 1 , . . . , An are Adomian's polynomials (see [2]). They are defined in such a way that each An depends only on 0 , 1 , . . . , n . Thus A0 = Ao(o), A1 = A1(0,1), A2 --- A2(0, 1, 2), etc. From (6) and (23) we have
oo
E A'~An 1 n=0 =r"
cO(L, ~b) 20~b cO(r,z) - r ' 2 " ~ z ' L "
(26)
Substituting (22) into (26) and then equating like powers of A from both sides of (26), we obtain Ao = 1 . 0 ( L e o , G o ) _ 2 . aG0 . L0, r a(r, z) r 2 Oz
A1 = A2 = + " " O(L~, ~o) 2 0o. Let O(r, z) - -~ r " Oz cO(L2, o) 2 0 0 . L~ cO(r,z) - -'~ r " cOz " + + . . . . . cO(r, z) . . . cO(r, z) r2 r2
(27)
01 . L0] Oz J
020z "L0]
(2s)
[1
cO(LI,I) 2 cO1.L1], cO(r,z) - r --~' cOz
(29)
If ~0 is once obtained, then we can obtain ~Pl in terms of 0. Similarly, 2 can be obtained in oo terms of ~Pt, etc. So, all components of ~p are calculable and tp = ~'~n=o Cn. 02 CASE 2. When L = Lrr - (1/r)Lr + Lzz, where Lzz = ~',x, equation (5) becomes
L2~ = v - i N C .
(30)
Navier-Stokes Equations
113
Operating with the inverse operator L -2 on (30), we get

~b ----u - I L - 2 ( N ) . T h e L -2 involved in (31) can be obtained in the same way as L~-2, and it is given by 1 [ L r J ( 1 L r L-2 = E 3---~ - Lzz)+L;l{r(Lrr
(31)
- Lzz)} + L ; 1 ( ! L r
- L~)]n
(32)
The parametrized form of equation (31) is

~b = u - I A L - 2 ( N ) . (33)
If we substitute (22) and (23) into (33), and then equate like powers of A on b o t h sides of (33), we obtain ~)1 ~
u-IL-2Ao,
~2 = u - I L - 2 A 1 ,
~3 = ~-IL-2A2,
~P~+I = u - I L - 2 A ~ ,
where Ao, A1 . . . . , An are A d o m i a n ' s special polynomials [2] and can be obtained from (27)-(29), etc. T h e use of the operator L in this analysis increases the speed of convergence, b u t the inverse of L and consequent integrations are more difficult [4]. T h e convergence of the series solution obtained by the decomposition m e t h o d has been established [2,5-7] and it is also observed in [4,5] t h a t a rapid stabilization to an acceptable accuracy is evident when numerical c o m p u t a t i o n of the analytic approximation is carried out.
REFERENCES
1. G. Adomian, Nonlinear Stochastic Operator Equations, Academic Press, (1986). 2. G. Adomian, Nonlinear Stochastic Systems Theory and Applications to Physics, Kluwer Academic, (1989). 3. G. Adomian, Application of the decomposition method to the Navier-Stokes equations, Jour. Math. Anal. Appl. 119, 340-360 (1986). 4. G. Adomian, Nonlinear transport in moving fluids, Appl. Math. Lett. 6 (5), 35-38 (1993). 5. G. Adomian and R. Rach, Analytic solution of nonlinear boundary-value problems to several dimensions, Jour. Math. Anal. Appl. 174, 118-137 (1993). 6. Y. Cherruault, Convergence of Adomian's method, Kybernetes 18 (2), 31-39 (1989). 7. Y. Cherruault, Some new results for convergence of Adomian's method applied to integral equations, Mathl~ Comput. Modelling 16 (2), 85-93 (1992).

Application of Adomian's Approximations To The Navier-Stokes Equations in Cylindrical Coordinates

Uploaded by

Copyright:

Available Formats

Application of Adomian's Approximations To The Navier-Stokes Equations in Cylindrical Coordinates

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Application of Adomian's Approximations To The Navier-Stokes Equations in Cylindrical Coordinates

Uploaded by

Copyright:

Available Formats

Pergamon

A p p l i c a t i o n of A d o m i a n ' s A p p r o x i m a t i o n s to the Navier-Stokes Equations in Cylindrical Coordinates

(Received June 1994; accepted August 1995)

lop (02v 10v 02v p Or -- 2 - + v \0r2 + r ~r + Oz

v2) -r2 '

1 0(L,) 2 0 L 9 = vL2~#, 7 o(r, z----7- - r-~ - f f z

where N is the nonlinear term denoted by

L~# = f f [.] dr dr,

Adding (13) and (14), and then dividing by 2, we have

1/~1=~ [L~J (l'Lr) ~-L~-l(rLrr)] o,

If the quantity within the bracket is denoted by S, then we can write

~n+~ = 2n+l Sn+1o,

so that the inverse L{-1

(1/r)Lr) -1 has been identified as

2LT2b--~z2(Ltt~) - L12 \-b-~z4) ] .

Equating the terms of like power of A, we have 02 1 = v - l L 1 2 A o - 2L1-2 ~ 2 = v - t L t 2 A 1 - 2 L t 2 ~

coao L12 COz4 ,

E A'~An 1 n=0 =r"

cO(L, ~b) 20~b cO(r,z) - r ' 2 " ~ z ' L "

cO(LI,I) 2 cO1.L1], cO(r,z) - r --~' cOz

Operating with the inverse operator L -2 on (30), we get

The parametrized form of equation (31) is

You might also like