A High-Order Compact Hermite Difference Method for Double-Diffusive Convection

TL;DR

This paper introduces a high-order compact Hermite difference scheme for simulating double-diffusive convection, improving accuracy and stability by directly discretizing fluxes without auxiliary equations.

Contribution

It presents a novel high-order compact Hermite scheme that simplifies the discretization process and enhances stability for double-diffusive convection simulations.

Findings

Numerical results agree well with benchmark solutions.

The scheme effectively handles steady and unsteady double-diffusive convection.

Preliminary applications demonstrate versatility across different parameters.

Abstract

In this paper, a class of high-order compact finite difference Hermite scheme is presented for the simulation of double-diffusive convection. To maintain linear stability, the convective fluxes are split into positive and negative parts, then the compact Hermite difference methods are used to discretize the positive and negative fluxes, respectively. The diffusion fluxes of the governing equations are directly approximated by a high-order finite difference scheme based on the Hermite interpolation. The advantages of the proposed schemes are that the derivative values of the solutions are directly solved by the compact central difference scheme, and the auxiliary derivative equation is no longer required. The third-order Runge-Kutta method is utilized for the temporal discretization. Several numerical tests are presented to assess the numerical capability of the newly proposed algorithm. The numerical results are in great agreement with the benchmark solutions and some of the accurate results available in the literature. Subsequently, we apply the algorithm to solve steady and unsteady problems of double-diffusive convection and a preliminary application to the double-diffusive convection for different Raleigh numbers and aspect ratios is carried out.