Conservative discontinuous Galerkin methods with implicit penalty parameters and multiscale hybridizable discontinuous Galerkin methods for PDEs

点击翻译按钮获取中文摘要

This project concentrates on the development of novel computational methods for efficiently solving problems that have conserved physical properties or highly oscillatory wave solutions. The new conservative methods can preserve physically interested quantities and allow accurate and stable simulations over a long time period. They will be useful for applications in various fields, such as fluid dynamics, nonlinear optics, plasma physics, and Bose-Einstein condensates. The new multiscale methods can accurately and efficiently capture highly oscillatory wave solutions. They will have a positive impact in the study of quantum mechanics and great potential in application to the design of ultrafast and low consumption nanoscale electronic devices. The methods developed in the project will help people understand theoretically unresolved issues and provide new frameworks for devising competitive numerical algorithms for solving other complex problems. The project will also involve mentoring and training of undergraduate and graduate students, including the traditionally underrepresented groups. It will provide students great opportunities to integrate research into their educational experience.The project includes the following topics: (1) in-depth investigation of the novel conservative discontinuous Galerkin (DG) method with implicit penalty parameters for the Korteweg-de Vries (KdV) equation, (2) development of conservative DG methods via implicit penalization for more complicated wave models with conservation properties, including the Hirota-Satsuma coupled KdV system, the Schrodinger-KdV system, the abcd-Boussinesq system, and the two-dimensional Zakharov-Kuznetsov (ZK) equation and Kadomtsev-Petviashvili (KP) equation, (3) design, analysis, and implementation of hybridizable discontinuous Galerkin (HDG) methods with multiscale basis for efficiently capturing highly oscillatory solutions of Schrodinger equations on coarse meshes. The novel idea in the first two topics is to enforce conservation properties via implicit penalization, and this can be generalized to other types of problems that feature conservation of physical quantities. The methods in the third topic integrate the efficient HDG framework and the multiscale non-polynomial basis functions, which makes them perform better than traditional finite element methods for Schrodinger equations on both coarse meshes and fine meshes.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.