E.O. Asante-Asamani,A. Kleefeld,B.A. Wade
Abstract:A fourth-order exponential time differencing (ETD) Runge-Kutta scheme with dimensional splitting is developed to solve multidimensional non-linear systems of reaction-diffusion equations (RDE). By approximating the matrix exponential in the scheme with the A-acceptable Padé (2,2) rational function, the resulting scheme (ETDRK4P22-IF) is verified empirically to be fourth-order accurate for several RDE. The scheme is shown to be more efficient than competing fourth-order ETD and IMEX schemes, achieving up to 20 times speed in CPU time. Inclusion of up to three pre-smoothing steps of a lower order L-stable scheme facilitates efficient damping of spurious oscillations arising from problems with non-smooth initial/boundary conditions.
What problem does this paper attempt to address?
### What problem does this paper attempt to solve?
This paper aims to develop an efficient numerical method for solving multi - dimensional nonlinear reaction - diffusion equations (Reaction - Diffusion Equations, RDE). Specifically, the paper proposes a fourth - order exponential time differencing (Exponential Time Differencing, ETD) Runge - Kutta scheme and introduces the dimension - splitting technique. The following are the main objectives and contributions of this research:
1. **Improve computational efficiency**: By using A - acceptable Padé(2, 2) rational function to approximate the matrix exponential, an efficient and accurate fourth - order ETD Runge - Kutta scheme (called ETDRK4P22 - IF) is created. Experimental verification shows that this scheme can achieve up to 20 times CPU time acceleration in some cases.
2. **Solve high - dimensional problems**: For reaction - diffusion equations in high - dimensional spaces, the dimension - splitting technique is used, enabling the new scheme to handle high - dimensional problems more effectively while maintaining high precision.
3. **Handle non - smooth initial - boundary value conditions**: By introducing a pre - smoothing step of the low - order L - stable ETD Runge - Kutta scheme, the spurious oscillations caused by non - smooth initial - boundary value conditions are effectively suppressed, thereby improving the stability of the numerical solution.
4. **Compare with existing methods**: Compared with the existing fourth - order ETD and IMEX schemes, the new scheme not only improves in accuracy but also shows significant advantages in computational efficiency.
### Mathematical model
Consider the reaction - diffusion equation in the following form:
\[
\frac{\partial u}{\partial t}=D\Delta u + f(u, t),\quad\text{in }\Omega_T
\]
\[
u(x, t)=g(x, t),\quad\text{on }\Gamma_T.
\]
Here, $\Omega$ is a bounded open set, $\Omega_T:=\Omega\times(0, T]$, $\Gamma_T:=\Omega_T\setminus\Omega_T$. The solution $u:\Omega_T\rightarrow\mathbb{R}^s$ describes the concentrations of $s$ substances, $D = \text{diag}(d_1, d_2,\ldots, d_s)$ is the diffusion coefficient matrix, and $f:\Omega_T\rightarrow\mathbb{R}^s$ describes the interactions between various species.
### Method overview
1. **Discretization**: First, discretize the spatial derivative to obtain a system of ordinary differential equations (ODE).
\[
\frac{dU}{dt}+AU = F(U(t), t),\quad U\in\mathbb{R}^{s\times m_d}
\]
where $U(t)=(u_1(t), u_2(t),\ldots, u_s(t),\ldots, u_s(t)_{s\times m_d})^T$, $A$ is a $(s\times m_d)\times(s\times 2m_d)$ matrix, and $F\in\mathbb{R}^{s\times m_d}$ is a nonlinear mapping.
2. **Time integration**: Approximate the integral and matrix exponential by the exponential time differencing Runge - Kutta method to solve the above ODE system.
3. **Dimension splitting**: Use the dimension - splitting technique to decompose the two - dimensional Laplacian operator into the tensor product of two one - dimensional Laplacian operators, thereby simplifying the calculation and improving efficiency.
4. **Pre - smoothing**: For problems with non - smooth initial - boundary value conditions, introduce a pre - smoothing step to suppress spurious oscillations and ensure the stability and accuracy of the numerical solution.
Through these improvements, the new scheme proposed in the paper not only improves the computational efficiency but also ensures the accuracy and stability of the numerical solution when solving multi - dimensional nonlinear reaction - diffusion equations.