A domain- specific language and code generation for Riemann solvers

A domain- specific language and code generation for Riemann solvers

Internship Description

Riemann solvers are the core algorithms in numerical methods for wave propagation. They are also the most complex part of developing code for new applications. The design and implementation of an effective approximate Riemann solver typically requires substantial expertise and time. However, there exist generic approaches that utilize relatively straightforward properties of the hyperbolic system and are reasonably efficient. The goal of this project is to develop a domain- specific language for hyperbolic conservation laws and Riemann solvers, and implement automatic generation of efficient solvers based on a purely symbolic representation of the problem.  ​​​

Faculty Name

David Ketcheson

Field of Study

​Mathematics, computer science or engineering