Next Article in Journal
Fractal Operators and Fractional Dynamics with 1/2 Order in Biological Systems
Next Article in Special Issue
Monotonicity Results for Nabla Riemann–Liouville Fractional Differences
Previous Article in Journal
Hermite–Hadamard Type Inequalities Involving (k-p) Fractional Operator for Various Types of Convex Functions
Previous Article in Special Issue
Asymptotic Regularity and Existence of Time-Dependent Attractors for Second-Order Undamped Evolution Equations with Memory
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Numerical Approximation of the Fractional Rayleigh–Stokes Problem Arising in a Generalised Maxwell Fluid

1
Division of Applied Mathematics, Science and Technology Advanced Institute, Van Lang University, Ho Chi Minh City 700000, Vietnam
2
Faculty of Applied Technology, School of Engineering and Technology, Van Lang University, Ho Chi Minh City 700000, Vietnam
3
School of Mathematics, Iran University of Science and Technology, Narmak, Tehran 16846-13114, Iran
4
Department of Applied Mathematics, Xi’an Jiaotong-Liverpool University, Suzhou 215123, China
5
Institute of Mechanical Engineering, Faculty of Engineering, University of Porto, 4200-465 Porto, Portugal
*
Author to whom correspondence should be addressed.
Submission received: 10 June 2022 / Revised: 28 June 2022 / Accepted: 30 June 2022 / Published: 2 July 2022

Abstract

:
This paper presents a numerical technique to approximate the Rayleigh–Stokes model for a generalised Maxwell fluid formulated in the Riemann–Liouville sense. The proposed method consists of two stages. First, the time discretization of the problem is accomplished by using the finite difference. Second, the space discretization is obtained by means of the predictor–corrector method. The unconditional stability result and convergence analysis are analysed theoretically. Numerical examples are provided to verify the feasibility and accuracy of the proposed method.

1. Introduction

Fractional calculus (FC) generalises the classical integer-order calculus [1,2]. In the last decade, FC has received considerable attention in a variety of scientific areas [3,4,5,6]. In most cases, it is difficult to compute an explicit solution to fractional differential equations, which has attracted researchers to look for accurate and efficient numerical approaches for solving these Equations [7,8,9,10,11,12,13,14].
FC has successfully described viscoelastic fluid constitutive relations [15,16,17]. One usually starts the process of modelling viscoelastic fluids via fractional derivatives by modifying a traditional differential equation. This generalisation involves using the Riemann–Liouville (R–L) fractional derivative operator instead of the standard time derivative. Shen et al. [18] derived the Rayleigh–Stokes Equation (RSE) for a generalised fluid of the second grade flowing within a heated edge and over a heated flat plate. The analytic solutions for the temperature and velocity fields were obtained via the fractional Laplace and the Fourier sine transforms. Mainardi [19] provided a comprehensive review of the relationship between FC and viscoelastic models, linear viscoelasticity, and wave propagation. Qi and Xu [20] studied an unsteady flow of fractional Maxwell fluid in a channel. Xue and Nie [21] addressed the RSE to a heated generalised fluid of the second grade with fractional derivative flows inside a porous half-space.
In this paper, we investigate the numerical solution for the time fractional Rayleigh–Stokes Equation (TFRSE) including the time fractional derivative
v ( x , y , t ) t = 0 D t 1 γ k 1 2 v ( x , y , t ) x 2 + k 2 2 v ( x , y , t ) y 2 + k 3 2 v ( x , y , t ) x 2 + k 4 2 v ( x , y , t ) y 2 + f ( x , y , t ) ,
with initial and boundary conditions (abbreviated as IC and BCs, respectively):
v ( x , y , 0 ) = ϕ ( x , y ) , 0 x , y L ,
v ( x , 0 , t ) = φ 1 ( x , t ) , v ( x , L , t ) = φ 2 ( x , t ) , 0 t T , 0 x L ,
v ( 0 , y , t ) = ψ 1 ( y , t ) , v ( L , y , t ) = ψ 2 ( y , t ) , 0 t T , 0 y L ,
where v ( x , y , t ) represent the velocity field, the coefficients k 1 , k 2 , k 3 and k 4 are positive constants, 0 < γ < 1 , f ( x , y , t ) is a source term, functions ϕ , φ 1 , φ 2 , ψ 1 and ψ 2 are prescribed and 0 D t 1 γ v ( x , y , t ) is the R–L fractional differential derivative of order 1 γ denoted by
0 D t 1 γ v ( x , y , t ) = 1 Γ ( γ ) t 0 t v ( x , y , η ) ( t η ) 1 γ d η .
Some numerical techniques have been used to approximate the TFRSE Equation (1). Chen et al. [22] formulated an implicit finite difference (FD) algorithm. Wu [23] presented an implicit numerical approximation scheme. Jiang and Lin [24] proposed a numerical technique based on the method of reproducing kernel. Mohebbi et al. [25] adopted the compact FD method (CFDM) and radial basis function (RBF) meshless method (RMM). Zaky [26] and Shivanian et al. [27] used the Legendre–Tau and the meshless singular boundary methods, respectively. Hafez et al. [28] applied the Jacobi Spectral Galerkin method for the distributed TFRSE. Safari and HongGuang [29] adopted the improved dual reciprocity and singular boundary schemes for the TFRSE, while Golbabai et al. [30] proposed the local meshless RBF. Khan et al. [31] developed the high-order compact scheme, whereas Naz et al. [32] advanced a modified implicit scheme for the TFRSE.
This paper introduces a numerical method for the TFRSE and is organized as follows. Section 2 describes an algorithm to approximate the time fractional derivative of the TFRSE Equation (1). Section 3 accomplishes the space discretization with the help of the predictor–corrector method. Section 4 studies the unconditional stability result and convergence analysis of the proposed strategy by using Fourier analysis. Section 5 reports the numerical examples of the TFRSE and verifies the efficiency of the proposed scheme. Finally, Section 6 provides a concise conclusion.

2. The Time Discretization

Let us define t k = k τ , where τ = T N represents the time step size for k = 0 , 1 , 2 , , N . We suppose that the solution u ( x , y , t ) has a continuous partial derivative u ( x , y , t ) t for t 0 , and that the R–L fractional derivative 0 D t 1 γ u ( x , y , t ) can be evaluated using the Grünwald–Letnikov (G–L) formulation [33], described as
0 D t 1 γ u ( x , y , t ) = 1 τ 1 γ j = 0 [ t / τ ] λ j ( 1 γ ) u ( x , y , t j τ ) + O ( τ q ) , 0 < γ < 1 , 0 < q < 1 ,
so that λ j ( 1 γ ) correspond to the coefficients of the generating function
λ ( z , 1 γ ) = j = 0 λ j ( 1 γ ) z j ,
where the coefficients λ 0 ( 1 γ ) = 1 and λ j ( 1 γ ) = ( 1 ) j 1 γ j are the normalized G–L weights. The coefficients can be obtained recursively as:
λ 0 ( 1 γ ) = 1 , λ j ( 1 γ ) = 1 2 γ j λ j 1 ( 1 γ ) , j 1 .
The λ j 1 γ have some useful properties, as given by Lemma 1 ([33]).
Lemma 1.
The coefficients λ j 1 γ introduced in Equation (6) satisfy
  • λ j ( 1 γ ) < 0 , j = 1 , 2 , .
  • j = 0 λ j ( 1 γ ) = 0 , n N , j = 0 n λ j ( 1 γ ) < 1 and j = 0 n 1 λ j ( 1 γ ) > 0 .
  • j = n λ j ( 1 γ ) 1 n 1 γ Γ ( γ ) ·
Let us consider the following notations:
0 D t 1 γ v ( x , y , t ) t = t k = τ γ 1 l = 0 k λ l v ( x , y , t k l ) + O ( τ ) ,
v ( x , y , t k ) t = v ( x , y , t k ) v ( x , y , t k 1 ) τ + O ( τ ) .
Now, we can formulate the semi-time discretization scheme for Equation (1) by means of the aforementioned relations
v ( x , y , t k ) v ( x , y , t k 1 ) τ = τ γ 1 l = 0 k λ l k 1 2 v ( x , y , t k l ) x 2 + k 2 2 v ( x , y , t k l ) y 2 + k 3 2 v ( x , y , t k l ) x 2 + k 4 2 v ( x , y , t k l ) y 2 + f ( x , y , t k ) .

3. The Space Discretization

Let Ω = { ( x i , y j ) | 1 i M 1 , 1 j M 2 } with x i = i h x , y j = j h y , so that h x = L M 1 , h y = L M 2 represent the space steps in the x and y directions, respectively, and also M 1 and M 2 denote the total number of space steps in the x and y directions, respectively. Discretizing Equation (1) at the above grid points ( x i , x j , t k ) and by using
2 v ( x i , y j , t k ) x 2 = δ x 2 v ( x i , y j , t k ) h x 2 + O ( h x 2 ) ,
2 v ( x i , y j , t k ) y 2 = δ y 2 v ( x i , y j , t k ) h y 2 + O ( h y 2 ) ,
where
δ x 2 v ( x i , y j , t k ) = v ( x i 1 , y j , t k ) 2 v ( x i , y j , t k ) + v ( x i + 1 , y j , t k ) ,
δ y 2 v ( x i , y j , t k ) = v ( x i , y j 1 , t k ) 2 v ( x i , y j , t k ) + v ( x i , y j + 1 , t k ) ,
we obtain the following corrector formula as
v i , j k = 1 τ [ v i , j k 1 + μ 1 l = 1 k λ l δ x 2 v i , j k l + μ 2 l = 1 k λ l δ y 2 v i , j k l + ( μ 1 + μ 3 ) ( v i + 1 , j k + v i 1 , j k ) + ( μ 2 + μ 4 ) ( v i , j + 1 k + v i , j 1 k ) + τ f i , j k ] ,
1 i K 1 1 , 1 j K 2 1 , 1 k N .
such that
τ = 1 + 2 ( μ 1 + μ 2 + μ 3 + μ 4 ) ,
μ 1 = k 1 τ γ h x 2 , μ 2 = k 2 τ γ h y 2 , μ 3 = k 3 τ h x 2 , μ 4 = k 4 τ h y 2 ,
with f i , j k representing the value of function f at ( x i , y j , t k ) . The truncation error [22] is obtained as
R = O ( h x 2 + h y 2 ) τ γ l = 0 k λ l + O ( τ h x 2 + τ h y 2 + τ 2 ) ,
and the predictor formula is denoted by
v i , j k = v i , j k 1 ϵ v i , j k 1 ,
1 i K 1 1 , 1 j K 2 1 , 1 k N ,
with the IC and BCs
v i , j 0 = ϕ ( x i , y j ) , i = 0 , 1 , , K 1 , j = 0 , 1 , , K 2 ,
v i , 0 k = φ 1 ( x i , t k ) , v i , K 2 k = φ 2 ( x i , t k ) , i = 1 , 2 , , K 1 1 , k = 1 , 2 , , N ,
v 0 , j k = ψ 1 ( y j , t k ) , v K 1 , j k = ψ 2 ( y j , t k ) , j = 1 , 2 , , K 2 1 , k = 1 , 2 , , N ,
respectively. We can adopt the following iterative procedure to approximate Equation (1) with Equations (2)–(4).
P: Predict some value [ v i , j k ] 0 for v i , j k with
[ v i , j k ] 0 = v i , j k 1 ϵ v i , j k 1
1 i K 1 1 , 1 j K 2 1 , 1 k N ,
where ϵ is a very small number.
E: Evaluate the implicit part in Equation (12) with [ v i , j k ] 0 .
C: Correct [ v i , j k ] 0 to obtain a new [ v i , j k ] 1 for v i , j k with
[ v i , j k ] 1 = 1 τ [ v i , j k 1 + μ 1 l = 1 k λ l δ x 2 v i , j k l + μ 2 l = 1 k λ l δ y 2 v i , j k l + ( μ 1 + μ 3 ) ( [ v i + 1 , j k ] 0 + [ v i 1 , j k ] 0 ) + ( μ 2 + μ 4 ) ( [ v i , j + 1 k ] 0 + [ v i , j 1 k ] 0 ) + τ f i , j k ] ,
1 i K 1 1 , 1 j K 2 1 , 1 k N .
E: Evaluate the implicit part in Equation (12) with [ v i , j k ] 1 .
C: Correct [ v i , j k ] 1 with
[ v i , j k ] 2 = 1 τ [ v i , j k 1 + μ 1 l = 1 k λ l δ x 2 v i , j k l + μ 2 l = 1 k λ l δ y 2 v i , j k l + ( μ 1 + μ 3 ) ( [ v i + 1 , j k ] 1 + [ v i 1 , j k ] 1 ) + ( μ 2 + μ 4 ) ( [ v i , j + 1 k ] 1 + [ v i , j 1 k ] 1 ) + τ f i , j k ] ,
1 i K 1 1 , 1 j K 2 1 , 1 k N .
The sequence of operations
PECECECEC…
determines v i , j k for a sequence of values as
[ v i , j k ] 0 , [ v i , j k ] 1 , [ v i , j k ] 2 ,
An appropriate stop condition is
| [ v i , j k ] n + 1 [ v i , j k ] n | < ε , n = 0 , 1 , 2 ,
where n is the number of iterations. Hence, the relation Equation (12) can be rewritten in following form:
[ v i , j k ] P = v i , j k 1 ϵ v i , j k 1 [ v i , j k ] C = 1 τ [ v i , j k 1 + μ 1 l = 1 k λ l δ x 2 v i , j k l + μ 2 l = 1 k λ l δ y 2 v i , j k l + ( μ 1 + μ 3 ) ( [ v i + 1 , j k ] P + [ v i 1 , j k ] P ) + ( μ 2 + μ 4 ) ( [ v i , j + 1 k ] P [ v i , j 1 k ] P ) + τ f i , j k ] ,
1 i K 1 1 , 1 j K 2 1 , 1 k N ,
with the IC and BCs Equations (16)–(18).

4. Theoretical Analysis of the Proposed Method

4.1. Stability Analysis

We now examine the stability of the proposed method Equation (19) by using Fourier analysis. Let v i , j k and v ¯ i , j k be the exact and the approximated solutions for [ v i , j k ] C in Equation (19). Then, the error can be defined as:
E i , j k = v i , j k v ¯ i , j k .
We can obtain for corrector formula
E i , j k = 1 τ [ E i , j k 1 + μ 1 l = 1 k λ l δ x 2 E i , j k l + μ 2 l = 1 k λ l δ y 2 E i , j k l + ( μ 1 + μ 3 ) ( E i + 1 , j k + E i 1 , j k ) + ( μ 2 + μ 4 ) ( E i , j + 1 k + E i , j 1 k ) ] .
Let us introduce the following function:
E k ( x , y ) = E i , j k ( x , y ) Ω 1 , 0 ( x , y ) Ω 2 ,
such that
Ω 1 = ( x , y ) | x i 1 2 < x x i + 1 2 , y j 1 2 < y y j + 1 2 ,
Ω 2 = ( x , y ) | 0 x h x 2 or L h x 2 < x L o r 0 y h y 2 o r L h y 2 < y L .
Then, E k ( x , y ) can be approximated by using the Fourier series as
E k ( x , y ) = l 1 = + l 2 = + ζ k ( l 1 , l 2 ) e 2 π I ( l 1 x + l 2 y ) L 1 k N ,
where
ζ k ( l 1 , l 2 ) = 1 L 2 0 L 0 L E k ( x , y ) e 2 π I l 1 x + l 2 y L d x d y , I = 1 .
Applying Parseval equality for k = 0 , 1 , , N
| | E k | | 2 = i = 1 K 1 1 j = 1 K 2 1 h x h y | E i , j k | 2 1 2 = l 1 = l 2 = | ζ k ( l 1 , l 2 ) | 2 1 2 .
Suppose that the difference Equation (20) has the following solution
E i , j k = ζ k e I ( σ 1 i h x + σ 2 j h y ) ,
where σ 1 = 2 π l 1 L and σ 2 = 2 π l 2 L · Substituting Equation (22) into Equation (20) and simplifying leads to
ζ k = 1 τ ζ k 1 ψ l = 1 k λ l ζ k l + ϕ ζ k ,
such that
ψ = 4 μ 1 sin 2 σ 1 h x 2 + 4 μ 2 sin 2 σ 2 h y 2 , ϕ = 2 ( μ 1 + μ 3 ) ( 1 2 sin 2 σ 1 h x 2 ) + 2 ( μ 2 + μ 4 ) ( 1 2 sin 2 σ 2 h y 2 ) .
Now, we investigate the stability of the method Equation (19).
Theorem 1.
The predictor–corrector method Equation (19) is stable if and only if ϕ 0 .
Proof. 
The proof will be obtained by using mathematical induction. For k = 1 , we get
| ζ 1 | = 1 τ | 1 ψ ( γ 1 ) + ϕ | | ζ 0 | 1 τ 1 ψ | ( γ 1 ) | + ϕ | ζ 0 | | ζ 0 | .
Assume that
| ζ k | | ζ 0 | , k = 1 , 2 , , N 1 .
Then, by using [Lemma 2 in [25]], we arrive at
| ζ k | = 1 τ ζ k 1 ψ l = 1 k λ l ζ k l + ϕ ζ k , 1 τ 1 ψ l = 1 k λ l + ϕ | ζ 0 | , = 1 τ 1 + ψ l = 1 k λ l + ϕ | ζ 0 | , 1 τ ( 1 ψ + ϕ ) | ζ 0 | , | ζ 0 | .
Using the Parseval equality Equation (21), the error in the solution of the difference Equation (23) satisfies
| | E k | | 2 | | E 0 | | 2 , k = 1 , 2 , , N .
The proof of the theorem is completed. □

4.2. Convergence Analysis

Let us define the truncation error in the proposed method to satisfy the following form
R i , j k = v ( x i , y j , t k ) v ( x i , y j , t k 1 ) μ 1 l = 0 k λ l δ x 2 v ( x i , y j , t k l ) μ 2 l = 0 k λ l δ y 2 v ( x i , y j , t k l ) μ 3 δ x 2 v ( x i , y j , t k ) μ 4 δ x 2 v ( x i , y j , t k ) τ f ( x i , y j , t k ) ,
1 i K 1 1 , 1 j K 2 1 , 1 k N .
From Lemma 3 in [25], we have
R i , j k = O ( τ h x 2 + τ h y 2 + τ 2 ) ,
or
| R i , j k | C 1 ( τ h x 2 + τ h y 2 + τ 2 ) .
1 i K 1 1 , 1 j K 2 1 , 1 k N ,
where C 1 R + . Subtracting Equation (12) from Equation (26), we get
e i , j k = 1 τ [ e i , j k 1 + μ 1 l = 1 k λ l δ x 2 e i , j k l + μ 2 l = 1 k λ l δ y 2 e i , j k l + ( μ 1 + μ 3 ) ( e i + 1 , j k + e i 1 , j k ) + ( μ 2 + μ 4 ) ( e i , j + 1 k + e i , j 1 k ) + R i , j k ] ,
where e i , j k = v ( x i , y j , t k ) v i , j k , and
e 0 , j k = e K 1 , j k = 0 , e i , 0 k = e i , K 2 k = 0 , e i , j 0 = 0 ,
1 i K 1 , 1 j K 2 , 1 k N .
Let us consider the following two functions:
e k ( x , y ) = e i , j k , ( x , y ) Ω 1 , 0 , ( x , y ) Ω 2 ,
and
R k ( x , y ) = R i , j k , ( x , y ) Ω 1 , 0 , ( x , y ) Ω 2 .
Then, e i , j k and R i , j k can by approximated by the Fourier series
e k ( x , y ) = l 1 = + l 2 = + α k ( l 1 , l 2 ) e 2 π I ( l 1 x + l 2 y ) L 0 k N , R k ( x , y ) = l 1 = + l 2 = + β k ( l 1 , l 2 ) e 2 π I ( l 1 x + l 2 y ) L 0 k N ,
in which
α k ( l 1 , l 2 ) = 1 L 2 0 L 0 L e k ( x , y ) e 2 π I l 1 x + l 2 y L d x d y
and
β k ( l 1 , l 2 ) = 1 L 2 0 L 0 L R k ( x , y ) e 2 π I l 1 x + l 2 y L d x d y .
Moreover, we can obtain the values of | | e k | | 2 and | | R k | | 2 for k = 0 , 1 , , N as
| | e k | | 2 = i = 1 K 1 1 j = 1 K 2 1 h x h y | e i , j k | 2 1 2 = l 1 = l 2 = | α k ( l 1 , l 2 ) | 2 1 2 ,
and
| | R k | | 2 = i = 1 K 1 1 j = 1 K 2 1 h x h y | R i , j k | 2 1 2 = l 1 = l 2 = | β k ( l 1 , l 2 ) | 2 1 2 .
Suppose that e i , j k and R i , j k have the following form
e i , j k = α k e I ( σ 1 i h x + σ 2 j h y ) , R i , j k = β k e I ( σ 1 i h x + σ 2 j h y ) ,
where
σ 1 = 2 π l 1 L , σ 2 = 2 π l 2 L ·
Substituting Equation (31) into Equation (28) gives
α k = 1 τ α k 1 ψ l = 1 k λ l α k l + ϕ α k + β k ,
where τ , ψ and ϕ are as defined before. By virtue of the relations Equations (27) and (30), we get
| | R k | | 2 K 1 h x K 2 h y C 1 ( τ h x 2 + τ h y 2 + τ 2 ) , = C 1 L ( τ h x 2 + τ h y 2 + τ 2 ) .
Based on [22], it holds that
| β k | | β k ( l 1 , l 2 ) | C 2 | β 1 | C 2 | β 1 ( l 1 , l 2 ) | k = 1 , 2 , , N ,
where C 2 R + .
Proposition 1.
Let α k denote the solutions of Equation (32). Then, we have
| α k | C 2 k | β 1 | , 1 k N ,
where C 2 R + .
Proof. 
The principle of mathematical induction is applied by considering e 0 = 0 , which leads to
α 0 α 0 ( l 1 , l 2 ) = 0 .
For k = 1 , we obtain
| α 1 | | 1 τ | | β 1 | | β 1 | C 2 | β 1 | .
Let us assume that
| α n | n C 2 | β 1 | 1 n k 1 .
Then, based on Lemma in [25], we can conclude that
| α k | 1 τ [ ( k 1 ) ψ ( k 1 ) l = 1 k | λ l | + ϕ ( k 1 ) + 1 ] C 2 | β 1 | , 1 τ [ k ψ ( k 1 ) + ϕ ( k 1 ) ] C 2 | β 1 | , k C 2 | β 1 | ,
which completes the proof. □
Theorem 2.
The predictor–corrector method Equation (19) is convergent with the order O ( h x 2 + h y 2 + τ ) .
Proof. 
By considering Proposition 1 and using relations Equations (32) and (33), we can obtain
| | e k | | 2 k C 2 | | R 1 | | 2 k C 1 C 2 L ( τ h x 2 + τ h y 2 + τ ) ,
and noticing that k τ T , then
| | e k | | 2 C ( h x 2 + h y 2 + τ ) ,
in which
C = C 1 C 2 T L ,
and this completes the proof. □

5. Results and Discussion

This section presents two numerical examples for illustrating the stability and accuracy of the proposed method with several values of h x , h y , τ and γ . For this aim, we compute the following maximum-norm error L :
L = max 1 j N 1 | v ( x j , T ) V ( x j , T ) | ,
where v and V are the approximate and exact solutions, respectively. In addition, we evaluate the computational order in the time direction for the proposed method as:
C τ = log ( E 1 E 2 ) log ( τ 1 τ 2 ) ,
where E 1 and E 2 represent the errors corresponding to time step with sizes τ 1 and τ 2 , respectively.
Example 1.
Let us consider the following TFRSE:
v ( x , y , t ) t = 0 D t 1 γ 2 v ( x , y , t ) x 2 + 2 v ( x , y , t ) y 2 + 2 v ( x , y , t ) x 2 + 2 v ( x , y , t ) y 2 + f ( x , y , t ) , ( x , y ) Ω = [ 0 , 1 ] 2 .
The IC and BCs as well as f ( x , y , t ) are determined from an exact solution v ( x , y , t ) = exp ( x + y ) t 1 + γ .
The proposed method is implemented for solving this problem when the total time T = 1 with various values of h x , h y , τ and γ . Table 1 presents the maximum absolute errors L , time convergence orders C τ and execution times (in seconds) for h x = h y = 1 8 , γ { 0.55 , 0.85 } and various values of τ . We see that the obtained computational orders C τ in time direction are close to the theoretical convergence rate, that is, O ( τ ) .  Table 2 lists the maximum absolute errors L in the solution for γ { 0.5 , 0.75 } , and several values of h x = h y and τ .  Table 3 compares the maximum absolute errors L in the solution with those resulting from the method described in [25] for h x = h y = 1 16 , γ = 0.15 and various values of τ .  Table 4 makes the comparison of L errors in the solution with those resulting from the method in [25] for different values of h x = h y , γ = 0.2 , and τ .  Table 5 compares the maximum absolute errors L in the solution and execution times (in seconds) with those obtained with other schemes [22,25] for h x = h y = 1 4 , τ = 1 900 and different values of γ . Figure 1 shows the approximate solution and the associated computational error L of the proposed method with γ = 0.85 , K 1 = K 2 = 16 , and N = 256 .
Example 2.
Consider the following TFRSE:
v ( x , y , t ) t = 0 D t 1 γ 2 v ( x , y , t ) x 2 + 2 v ( x , y , t ) y 2 + 2 v ( x , y , t ) x 2 + 2 v ( x , y , t ) y 2 + f ( x , y , t ) , ( x , y ) Ω = [ 0 , 1 ] 2 .
The IC and BCs as well as source term f ( x , y , t ) are achieved from an exact solution v ( x , y , t ) = exp ( x + y ) t 2 .
The proposed method is adopted for solving this problem when the total time T = 1 with various values of h x , h y , τ and γ . Table 6 reports the maximum absolute errors L and time convergence orders C τ for h x = h y = 1 8 , γ { 0.65 , 0.95 } and various values of τ . As shown in Table 6, the computational orders C τ agree with the theoretical convergence order. Table 7 compares the maximum absolute errors L in the solution for γ { 0.35 , 0.75 } and several values of h x = h y and τ .  Table 8 lists the maximum absolute errors L in the solution for γ = 0.2 and several values of h x = h y and τ .  Table 9 makes the comparison of the maximum absolute errors L in the solution and execution times (in seconds) with those obtained with other schemes [22,25] for h x = h y = 1 4 and τ = 1 256 and various values of γ . Figure 2 depicts the approximate solution and the associated computational error L of proposed method with γ = 0.95 , K 1 = K 2 = 16 , and N = 256 .

6. Concluding Remarks

We adopted the predictor–corrector method for approximating the TFRSE. First, the time discretization of the problem was accomplished by using the finite difference. Second, the space discretization was obtained with the help of the predictor–corrector method. The convergence and unconditional stability properties of the approach were discussed theoretically. Numerical experiments compared the results obtained with the proposed method and those produced using existing alternative schemes. The superiority of the new approach was thus verified and illustrated.

Author Contributions

Conceptualization, L.D.L. and O.N.; data curation, B.M. and L.D.L.; formal analysis, B.M., O.N. and Z.A.; investigation, L.D.L. and O.N.; methodology, B.M., L.D.L. and A.M.L.; software, B.M. and O.N.; supervision, Z.A. and A.M.L.; validation, B.M., O.N. and Z.A.; visualization, B.M. and O.N.; writing–original draft, B.M. and O.N.; writing–review and editing, Z.A. and A.M.L. All authors have read and agreed to the published version of the manuscript.

Funding

The author Le Dinh Long is supported by Van Lang University.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Milici, C.; Drăgănescu, G.; Machado, J.T. Introduction to Fractional Differential Equations; Springer: Berlin/Heidelberg, Germany, 2019; Volume 25. [Google Scholar]
  2. Biswas, K.; Bohannan, G.; Caponetto, R.; Lopes, A.M.; Machado, J.A.T. Fractional-Order Devices; Springer: Berlin/Heidelberg, Germany, 2017. [Google Scholar]
  3. Qi, F.; Qu, J.; Chai, Y.; Chen, L.; Lopes, A.M. Synchronization of incommensurate fractional-order chaotic systems based on linear feedback control. Fractal Fract. 2022, 6, 221. [Google Scholar] [CrossRef]
  4. Chen, L.; Li, X.; Chen, Y.; Wu, R.; Lopes, A.M.; Ge, S. Leader-follower non-fragile consensus of delayed fractional-order nonlinear multi-agent systems. Appl. Math. Comput. 2022, 414, 126688. [Google Scholar] [CrossRef]
  5. Sabatier, J.; Agrawal, O.P.; Machado, J.T. Advances in Fractional Calculus; Springer: Berlin/Heidelberg, Germany, 2007; Volume 4. [Google Scholar]
  6. Gutiérrez, R.E.; Rosário, J.M.; Tenreiro Machado, J. Fractional order calculus: Basic concepts and engineering applications. Math. Probl. Eng. 2010, 2010, 375858. [Google Scholar] [CrossRef]
  7. Nikan, O.; Avazzadeh, Z. Numerical simulation of fractional evolution model arising in viscoelastic mechanics. Appl. Numer. Math. 2021, 169, 303–320. [Google Scholar] [CrossRef]
  8. Qiu, W.; Xu, D.; Guo, J.; Zhou, J. A time two-grid algorithm based on finite difference method for the two-dimensional nonlinear time-fractional mobile/immobile transport model. Numer. Algorithms 2020, 85, 39–58. [Google Scholar] [CrossRef]
  9. Xu, D.; Qiu, W.; Guo, J. A compact finite difference scheme for the fourth-order time-fractional integro-differential equation with a weakly singular kernel. Numer. Meth. Part Differ. Equ. 2020, 36, 439–458. [Google Scholar] [CrossRef]
  10. Farnam, B.; Esmaeelzade Aghdam, Y.; Nikan, O. Numerical investigation of the two-dimensional space-time fractional diffusion equation in porous media. Math. Sci. 2021, 15, 153–160. [Google Scholar] [CrossRef]
  11. Qiu, W.; Xu, D.; Guo, J. The Crank-Nicolson-type Sinc-Galerkin method for the fourth-order partial integro-differential equation with a weakly singular kernel. Appl. Numer. Math. 2021, 159, 239–258. [Google Scholar] [CrossRef]
  12. Aghdam, Y.E.; Mesgrani, H.; Javidi, M.; Nikan, O. A computational approach for the space-time fractional advection–diffusion equation arising in contaminant transport through porous media. Eng. Comput. 2021, 37, 3615–3627. [Google Scholar] [CrossRef]
  13. Qiu, W.; Xu, D.; Guo, J. Numerical solution of the fourth-order partial integro-differential equation with multi-term kernels by the Sinc-collocation method based on the double exponential transformation. Appl. Math. Comput. 2021, 392, 125693. [Google Scholar] [CrossRef]
  14. Nikan, O.; Avazzadeh, Z.; Machado, J.T. Numerical approach for modeling fractional heat conduction in porous medium with the generalized Cattaneo model. Appl. Math. Model. 2021, 100, 107–124. [Google Scholar] [CrossRef]
  15. Wenchang, T.; Wenxiao, P.; Mingyu, X. A note on unsteady flows of a viscoelastic fluid with the fractional Maxwell model between two parallel plates. Int. J. Non-Linear Mech. 2003, 38, 645–650. [Google Scholar] [CrossRef]
  16. Hayat, T.; Nadeem, S.; Asghar, S. Periodic unidirectional flows of a viscoelastic fluid with the fractional Maxwell model. Appl. Math. Comput. 2004, 151, 153–161. [Google Scholar] [CrossRef]
  17. Heymans, N. Fractional calculus description of non-linear viscoelastic behaviour of polymers. Nonlinear Dyn. 2004, 38, 221–231. [Google Scholar] [CrossRef]
  18. Shen, F.; Tan, W.; Zhao, Y.; Masuoka, T. The Rayleigh–Stokes problem for a heated generalized second grade fluid with fractional derivative model. Nonlinear Anal. Real World Appl. 2006, 7, 1072–1080. [Google Scholar] [CrossRef]
  19. Mainardi, F. Fractional Calculus and Waves in Linear Viscoelasticity: An Introduction to Mathematical Models; World Scientific: Singapore, 2010. [Google Scholar]
  20. Qi, H.; Xu, M. Unsteady flow of viscoelastic fluid with fractional Maxwell model in a channel. Mech. Res. Commun. 2007, 34, 210–212. [Google Scholar] [CrossRef]
  21. Xue, C.; Nie, J. Exact solutions of the Rayleigh–Stokes problem for a heated generalized second grade fluid in a porous half-space. Appl. Math. Model. 2009, 33, 524–531. [Google Scholar] [CrossRef]
  22. Chen, C.M.; Liu, F.; Anh, V. Numerical analysis of the Rayleigh–Stokes problem for a heated generalized second grade fluid with fractional derivatives. Appl. Math. Comput. 2008, 204, 340–351. [Google Scholar] [CrossRef] [Green Version]
  23. Wu, C. Numerical solution for Stokes’ first problem for a heated generalized second grade fluid with fractional derivative. Appl. Numer. Math. 2009, 59, 2571–2583. [Google Scholar] [CrossRef]
  24. Lin, Y.; Jiang, W. Numerical method for Stokes’ first problem for a heated generalized second grade fluid with fractional derivative. Numer. Methods Partial. Differ. Equations 2011, 27, 1599–1609. [Google Scholar] [CrossRef]
  25. Mohebbi, A.; Abbaszadeh, M.; Dehghan, M. Compact finite difference scheme and RBF meshless approach for solving 2D Rayleigh–Stokes problem for a heated generalized second grade fluid with fractional derivatives. Comput. Methods Appl. Mech. Eng. 2013, 264, 163–177. [Google Scholar] [CrossRef]
  26. Zaky, M.A. An improved tau method for the multi-dimensional fractional Rayleigh–Stokes problem for a heated generalized second grade fluid. Comput. Math. Appl. 2018, 75, 2243–2258. [Google Scholar] [CrossRef]
  27. Shivanian, E.; Jafarabadi, A. Rayleigh–Stokes problem for a heated generalized second grade fluid with fractional derivatives: A stable scheme based on spectral meshless radial point interpolation. Eng. Comput. 2018, 34, 77–90. [Google Scholar] [CrossRef]
  28. Hafez, R.M.; Zaky, M.A.; Abdelkawy, M.A. Jacobi Spectral Galerkin method for Distributed-Order Fractional Rayleigh-Stokes problem for a Generalized Second Grade Fluid. Front. Phys 2020, 7, 240. [Google Scholar] [CrossRef]
  29. Safari, F.; Sun, H. Improved singular boundary method and dual reciprocity method for fractional derivative Rayleigh–Stokes problem. Eng. Comput. 2020, 37, 3151–3166. [Google Scholar] [CrossRef]
  30. Nikan, O.; Golbabai, A.; Machado, J.T.; Nikazad, T. Numerical solution of the fractional Rayleigh–Stokes model arising in a heated generalized second-grade fluid. Eng. Comput. 2021, 37, 1751–1764. [Google Scholar] [CrossRef]
  31. Khan, M.A.; Ali, N.H.M. High-order compact scheme for the two-dimensional fractional Rayleigh–Stokes problem for a heated generalized second-grade fluid. Adv. Differ. Equ. 2020, 2020, 233. [Google Scholar] [CrossRef]
  32. Naz, A.; Ali, U.; Elfasakhany, A.; Ismail, K.A.; Al-Sehemi, A.G.; Al-Ghamdi, A.A. An Implicit Numerical Approach for 2D Rayleigh Stokes Problem for a Heated Generalized Second Grade Fluid with Fractional Derivative. Fractal Fract. 2021, 5, 283. [Google Scholar] [CrossRef]
  33. Yuste, S.B. Weighted average finite difference methods for fractional diffusion equations. J. Comput. Phys. 2006, 216, 264–274. [Google Scholar] [CrossRef] [Green Version]
Figure 1. Approximate solution v ( x , y , T ) and the associated computational error L of the proposed method with γ = 0.85 and K 1 = K 2 = 16 and N = 256 , n = 263 (left and right panels, respectively).
Figure 1. Approximate solution v ( x , y , T ) and the associated computational error L of the proposed method with γ = 0.85 and K 1 = K 2 = 16 and N = 256 , n = 263 (left and right panels, respectively).
Fractalfract 06 00377 g001
Figure 2. Approximate solution v ( x , y , T ) and the associated computational error L of the proposed method with γ = 0.95 and K 1 = K 2 = 16 and N = 256 , n = 263 (left and right panels, respectively).
Figure 2. Approximate solution v ( x , y , T ) and the associated computational error L of the proposed method with γ = 0.95 and K 1 = K 2 = 16 and N = 256 , n = 263 (left and right panels, respectively).
Fractalfract 06 00377 g002
Table 1. The maximum absolute errors L , time convergence orders C τ and execution times (in seconds) of Example 1 for h x = h y = 1 8 and γ { 0.55 , 0.85 } and different values of τ .
Table 1. The maximum absolute errors L , time convergence orders C τ and execution times (in seconds) of Example 1 for h x = h y = 1 8 and γ { 0.55 , 0.85 } and different values of τ .
τ γ = 0.55 γ = 0.85
n L C τ CPU Time (s) n L C τ CPU Time (s)
1/865 6.3090 × 10 3 0.1285 47 9.3000 × 10 3 0.1689
1/1063 5.3982 × 10 3 0.69871 0.1493 47 8.9975 × 10 3 0.14819 0.1093
1/1665 3.7302 × 10 3 0.78639 0.1558 48 4.8582 × 10 3 1.31122 1.7073
1/4066 1.6951 × 10 3 0.86077 0.5390 44 2.3980 × 10 3 0.77053 0.6569
1/6464 1.1170 × 10 3 0.88743 6.5077 40 1.5138 × 10 3 0.97874 1.3609
1/12870 5.6189 × 10 4 0.99127 11.508 32 7.6491 × 10 4 0.98481 2.1372
Table 2. The maximum absolute errors L in the solution of Example 1 for γ { 0.5 , 0.75 } , and different values of h x = h y and τ .
Table 2. The maximum absolute errors L in the solution of Example 1 for γ { 0.5 , 0.75 } , and different values of h x = h y and τ .
h x τ γ = 0.5 γ = 0.75
n L n L
1/41/48 7.9397 × 10 3 7 1.6266 × 10 2
1/101/64117 1.2258 × 10 3 80 1.5186 × 10 3
1/161/128419 6.7026 × 10 5 200 7.8371 × 10 4
1/81/210387 9.6004 × 10 4 40 8.2789 × 10 4
Table 3. The maximum absolute errors L of Example 1 for h x = h y = 1 16 , γ = 0.15 and different values of τ .
Table 3. The maximum absolute errors L of Example 1 for h x = h y = 1 16 , γ = 0.15 and different values of τ .
τ Proposed Method RMM [25]CFDM [25]
n L L L
1/10672 4.7049 × 10 3 6.8422 × 10 3 4.3318 × 10 3
1/20775 2.4089 × 10 3 4.9548 × 10 3 2.2982 × 10 3
1/30791 1.6230 × 10 3 3.3632 × 10 3 1.5709 × 10 3
Table 4. The maximum absolute errors L and execution times (in seconds) of Example 1 for γ = 0.2 , and different values of h x = h y and τ .
Table 4. The maximum absolute errors L and execution times (in seconds) of Example 1 for γ = 0.2 , and different values of h x = h y and τ .
h x τ Proposed Method CFDM [25]
n L CPU Time (s) L CPU Time (s)
1/41/5013 8.8037 × 10 4 0.8212 9.6999 × 10 4 0.0470
1/81/128116 2.0750 × 10 4 0.9606 3.8649 × 10 4 0.5000
1/181/28900 1.7396 × 10 3 6.3644 1.7554 × 10 3 0.0460
Table 5. The maximum absolute errors L of Example 1 for h x = h y = 1 4 and τ = 1 900 and different values of γ .
Table 5. The maximum absolute errors L of Example 1 for h x = h y = 1 4 and τ = 1 900 and different values of γ .
γ Proposed Method RMM [25]CFDM [25]
n L L L
0.7 5 8.3469 × 10 4 5.8055 × 10 4 1.8231 × 10 3
0.8 4 7.0423 × 10 4 7.3493 × 10 4 1.8250 × 10 3
0.9 4 9.8265 × 10 4 9.0422 × 10 4 1.8265 × 10 3
Table 6. The maximum absolute errors L and time convergence orders C τ of Example 2 for h x = h y = 1 8 , γ { 0.65 , 0.95 } and different values of τ .
Table 6. The maximum absolute errors L and time convergence orders C τ of Example 2 for h x = h y = 1 8 , γ { 0.65 , 0.95 } and different values of τ .
τ γ = 0.55 γ = 0.85
n L C τ n L C τ
1/1054 6.7051 × 10 3 -117 9.9156 × 10 3
1/4057 1.5503 × 10 3 1.0563 34 8.5430 × 10 3 0.1075
1/9052 6.9213 × 10 4 0.9944 31 1.0942 × 10 3 2.5342
1/19045 3.3634 × 10 4 0.9658 22 5.1960 × 10 4 0.9967
1/120021 7.7520 × 10 5 0.7963 25 8.3012 × 10 5 0.9951
Table 7. The maximum absolute errors L of Example 2 for γ { 0.35 , 0.75 } and different values of h x = h y and τ .
Table 7. The maximum absolute errors L of Example 2 for γ { 0.35 , 0.75 } and different values of h x = h y and τ .
h x τ γ = 0.35 γ = 0.75
n L n L
1/41/8160 9.9757 × 10 3 8 8.0374 × 10 3
1/81/8151 3.8090 × 10 3 73 7.0374 × 10 3
1/161/8486 4.4059 × 10 3 290 3.0374 × 10 3
1/81/6488 3.1647 × 10 4 46 1.1573 × 10 3
1/121/64283 3.1281 × 10 4 116 1.1583 × 10 3
1/41/1211 7.5942 × 10 4 7 4.8539 × 10 4
1/151/128947 1.6066 × 10 5 160 1.8297 × 10 4
1/81/3095 6.4711 × 10 4 51 2.4751 × 10 3
1/601/16740 1.9589 × 10 3 243 2.8112 × 10 3
Table 8. The maximum absolute errors L of Example 2 for γ = 0.2 , and several values of h x = h y and τ .
Table 8. The maximum absolute errors L of Example 2 for γ = 0.2 , and several values of h x = h y and τ .
h x τ Proposed Method Implicit Method [25]
n L L
1/41/509 7.0500 × 10 4 2.0154 × 10 3
1/81/12884 8.8460 × 10 5 1.3776 × 10 4
1/161/3255 2.0142 × 10 3 2.1391 × 10 3
Table 9. The maximum absolute errors L and execution times (in seconds) of Example 2 for h x = h y = 1 4 and τ = 1 256 and different values of γ .
Table 9. The maximum absolute errors L and execution times (in seconds) of Example 2 for h x = h y = 1 4 and τ = 1 256 and different values of γ .
γ Proposed Method Implicit Method [25]
n L CPU Time (s) L CPU Time (s)
0.18 8 9.2879 × 10 4 1.4274 3.2350 × 10 3 0.5373
0.35 6 4.9674 × 10 4 1.5169 3.2306 × 10 3 0.5397
0.70 3 7.8135 × 10 4 1.5070 3.2302 × 10 3 0.5719
0.80 4 2.3660 × 10 3 1.4007 3.2368 × 10 3 0.5651
0.90 3 3.0465 × 10 3 1.3906 3.2457 × 10 3 0.5102
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Long, L.D.; Moradi, B.; Nikan, O.; Avazzadeh, Z.; Lopes, A.M. Numerical Approximation of the Fractional Rayleigh–Stokes Problem Arising in a Generalised Maxwell Fluid. Fractal Fract. 2022, 6, 377. https://0-doi-org.brum.beds.ac.uk/10.3390/fractalfract6070377

AMA Style

Long LD, Moradi B, Nikan O, Avazzadeh Z, Lopes AM. Numerical Approximation of the Fractional Rayleigh–Stokes Problem Arising in a Generalised Maxwell Fluid. Fractal and Fractional. 2022; 6(7):377. https://0-doi-org.brum.beds.ac.uk/10.3390/fractalfract6070377

Chicago/Turabian Style

Long, Le Dinh, Bahman Moradi, Omid Nikan, Zakieh Avazzadeh, and António M. Lopes. 2022. "Numerical Approximation of the Fractional Rayleigh–Stokes Problem Arising in a Generalised Maxwell Fluid" Fractal and Fractional 6, no. 7: 377. https://0-doi-org.brum.beds.ac.uk/10.3390/fractalfract6070377

Article Metrics

Back to TopTop