Solving PDEs with DUNE-FEM
VT2022, 3.0 ECTS
– Published 4 April 2022
Objective
This course is intended for students with intermediate knowledge of programming in Python and basic knowledge of partial differential equations (PDEs) who would like to learn techniques how to solve such problems numerically.
The course will give a brief introduction to Finite Element methods and then discusses the main concepts needed to get a new user started on solving complex partial differential equations using the DUNE-FEM package.
DUNE-FEM is a module of the Distributed and Unified Numerics Environment (DUNE), a free and open source modular toolbox for solving PDEs with grid- based methods. Easy implementation of methods like Finite Elements (FE) and Finite Volumes (FV) via the Unified Form Language (UFL) ensure user friendliness whereas techniques such as code generation ensure efficiency in scientific computations including high-performance computing applications. The course also covers how to use different solver backends (built-in solvers from DUNE-ISTL, or SciPy and PETSc and also petsc4py), local grid adaptation and parallelization and other techniques.
Dates
- Block course 4th - 8th April 2022
- Project presentations: 30th April 2022
Personnel
- Course responsible: Robert Klöfkorn
- Teachers: Robert Klöfkorn, Andreas Dedner (Warwick, UK)
Pre-requisites
Python (NumPy, SciPy and Matplotlib) and basic knowledge on partial differential equations
Teaching
The lectures are scheduled to be on campus, but will be streamed online via Zoom if restrictions to avoid spreading the Corona virus are in place. A risk assessment will be conducted after the 25th of March. Students will be informed regarding the measures taken before the course starts.
Material
DUNE-FEM is readily available as a module on LUNARC's Aurora system
Registration
Registration is now closed.
Registration deadline 27th March 2022 (firm deadline)