Modern Wave Propagation - Discontinuous Galerkin & Julia, Summer 2020
In the past few years, the Discontinuous Galerkin (DG) method has made waves. Wave propagation is everywhere: It is used to simulate earthquakes, tsunamis and even plasma physics. With the advent of many-core supercomputers it has become crucial to use communication avoiding methods. DG is currently one of the most promising.
The goal of this lab course is to write a numerical solver for hyperbolic partial differential equations using the DG method with the modern programming language Julia. We provide a framework which handles input/output, mesh construction and other boilerplate tasks. Missing in this framework is the implementation of the numerical method. This is your task.
We introduce (only) the necessary theory in lecture-style presentations. You then have to apply this in practice by modifying the provided framework. The expected result is a DG-solver that can be used to simulate various wave problems at a reasonable speed.
The lab course ends with a project, in which you investigate further aspects. Examples for this could be performance optimizations, modifications of the numerical method and the implementation of example scenarios (e.g. tsunamis).