A variable mesh C-SPLAGE method of accuracy for 1D nonlinear parabolic equations
Introduction
High accuracy numerical approximation of scalar nonlinear parabolic initial boundary value problems has received a lot of attention during the last three decades. It is well known that the main difficulty in these problems is the presence of singular terms in the differential equation, whose high order approximation on non-uniform meshes have been the main focus of numerous research endeavors. Problems of this type arise in numerous applications from science and engineering, such as fluid flow at high Reynolds number, heat transfer with small diffusion parameters, just to name a few. In the context of the cubic spline method, the robust approximation of differential equation requires the use of the non-uniform meshes in the solution space. Jain and Aziz [1] have first used cubic spline technique for the solution of two point nonlinear boundary value problems. Later, Mohanty et al. [2] have extended their method to a variable mesh. The numerical solution of two point nonlinear boundary value problem using a variable mesh has been studied by Jain et al. [3], Mohanty [4], [5] and Bieniasz [6]. Recently, Mohanty and Singh [7] and Mohanty [8] have developed new two-level implicit variable mesh schemes for the solution of nonlinear parabolic equations. First, Papamichael and Whiteman [9] have used cubic spline technique for the solution of one-dimensional heat conduction equation. Later, using second order accurate cubic spline approximations, Rubin and Graves [10], and Jain and Lohar [11] have discussed numerical solution of nonlinear viscous flow problems. Recently, using three spatial grid points, Mohanty and Jain [12] have developed a cubic spline alternating group explicit (C-SPLAGE) method of order 2 in time and 4 in space for the solution of one-dimensional nonlinear parabolic equation on a uniform mesh. In this paper, using a variable mesh, we discuss a new two-level implicit highly accurate C-SPLAGE method for the solution of one-dimensional nonlinear parabolic equation. However, for constant mesh case, the proposed method reduces to the method discussed by Mohanty and Jain [12]. Further, recently Mohanty [13] has discussed a two-level implicit variable mesh cubic spline method for singularly perturbed non-linear parabolic initial boundary value problems. Different numerical methods both for the initial value and the initial boundary value problems for ordinary and partial differential equations have been already discussed in the literature [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29]. Difficulties were experienced in the past for the numerical solution of parabolic equations with singular coefficients using cubic spline approach, especially on a non-uniform mesh. The solutions become unbounded in the vicinity of the singular point. We modify our approximation in such a way that the solutions become bounded in the vicinity of the singularity.
In Section 2, we discuss the derivation procedure of the proposed cubic spline method. In Section 3, we discuss the application of the proposed cubic spline method to heat equation with singular coefficients and present C-SPLAGE and Newton-C-SPLAGE iterative methods for the solution of one-space dimensional linear and nonlinear parabolic equations. Since these methods are coupled compactly and explicit in nature, they are suitable for the use on parallel computers. Formation of variable mesh grid points is discussed briefly and comparative results are given in Section 4. Finally, concluding remarks and scope for further research work are given in Section 5.
Section snippets
Derivation procedure for cubic spline approximation
Consider the second order non-linear parabolic equation of the formsubject to the initial and boundary conditions are given by
Assume that the functions ϕ(x, t, u, ux, ut), u0(x), g0(t) and g1(t) are sufficiently smooth and their required high-order derivatives exist.
For the approximate solution of the above initial-boundary value problems, we discretize the space interval [0, 1] as 0 = x0 < x1 < ⋯ < xN < xN+1 = 1, where N is a
Application of the C-SPLAGE iterative method
First, we discuss the application of cubic spline method (18) to the linear singular parabolic equationsubject to appropriate initial and Dirichlet boundary conditions given by (2a), (2b), respectively. For α = 0, the equation above represents heat conduction equation. Replacing the variable x by r and for α = 1 and 2, Eq. (21) represents heat conduction equation in cylindrical and spherical coordinates, respectively.
Applying the method (18) to the differential
Formation of grid points and numerical results
We have solved the following three-benchmark problems using the proposed cubic spline method (18) by the help of C-SPLAGE and Newton-C-SPLAGE iterative methods on a variable mesh. The exact solutions are provided in each case. The right hand side homogeneous functions, initial and boundary conditions may be obtained by using the exact solution as a test procedure. In all cases, we have obtained the optimal values of the relaxation parameter (ρopt) and compared the performance of the proposed
Concluding remarks
Using three spatial variable mesh points (see Fig. 1), we have developed a new two-level implicit stable C-SPLAGE method based on cubic spline polynomial approximation for the solution of non-linear parabolic partial differential Eq. (1). Although the proposed variable mesh C-SPLAGE method involves more algebra, we do not require any fictitious points near the boundaries to solve singular parabolic equations. The numerical results indicate that the proposed variable mesh C-SPLAGE and
Acknowledgements
This research was supported by ‘The University of Delhi’ under Research Grant No. Dean(R)/R&D/2007/Ph-III/386.
The author thanks the reviewers for their valuable suggestions, which substantially improved the standard of the paper.
References (33)
- et al.
Cubic spline solution of two point boundary value problems with significant first derivatives
Comput. Meth. Appl. Mech. Eng.
(1983) - et al.
Variable mesh methods for the numerical solution of two point singular perturbation problems
Comput. Meth. Appl. Mech. Eng.
(1984) A family of variable mesh methods for the estimates of (du/dr) and the solution of nonlinear two point boundary value problems with singularity
J. Comput. Appl. Math.
(2005)A class of non-uniform mesh three point arithmetic average discretization for y″ = f(x, y, y′) and the estimates of y′
Appl. Math. Comput.
(2006)An implicit high accuracy variable mesh scheme for 1-D nonlinear singular parabolic partial differential equations
Appl. Math. Comput.
(2007)An explicit almost P-stable two-step method with phase-lag of order infinity for the numerical integration of second order periodic initial-value problems
Appl. Math. Computat.
(1992)Modified Runge–Kutta methods for the numerical solution of ordinary differential equations with oscillating solutions
Appl. Math. Comput.
(1997)- et al.
A P-stable exponentially fitted method for the numerical integration of the Schrodinger equation
Appl. Math. Comput.
(2000) A high order predictor-corrector method for periodic initial value problems
Appl. Math. Lett.
(1993)A new variable step method for the numerical integration of special second order initial value problems and their applications to the one-dimensional Schrodinger equation
Appl. Math. Lett.
(1993)
A predictor-corrector phase-fitted method for y″ = f(x, y)
Math. Comput. Simulat.
A Runge–Kutta Fehlberg method with phase-lag of order infinity for initial value problems with oscillating solution
Comput. Math. Appl.
Runge–Kutta–Nystrom interpolants for the numerical integration of special second-order periodic initial-value problems
Comput. Math. Appl.
Runge–Kutta interpolants with minimal phase-lag
Comput. Math. Appl.
An non-uniform mesh cubic spline TAGE method for nonlinear singular two point boundary value problems
Int. J. Comput. Math.
A set of compact finite difference approximations to first and second derivatives related to the extended Numerov method of Chawla on nonuniform grids
Computing
Cited by (7)
Fourth-Order Numerical Scheme Based on Half-Step Non-Polynomial Spline Approximations for 1D Quasi-Linear Parabolic Equations
2020, Numerical Analysis and ApplicationsA new high accuracy off-step cubic spline approximations on a quasi-variable mesh for the system of nonlinear parabolic equations in one space dimension
2020, International Journal for Computational Methods in Engineering Science and MechanicsA New Two-Level Implicit Scheme for the System of 1D Quasi-Linear Parabolic Partial Differential Equations Using Spline in Compression Approximations
2019, Differential Equations and Dynamical SystemsA finite element method by using bivariate splines for one dimensional heat equations
2013, Journal of Information and Computational Science