FEniCS Project


The FEniCS Project is a collection of free and open-source software components with the common goal to enable automated solution of differential equations. The components provide scientific computing tools for working with computational meshes,
finite-element variational formulations of ordinary and partial differential equations, and numerical linear algebra.

Design and components

The FEniCS Project is designed as an umbrella project for a collection of interoperable components. The core components are
DOLFIN, the computational high-performance C++ backend of FEniCS, functions as the main problem-solving environment and user interface. Its functionality integrates the other FEniCS components and handles communication with external libraries such as PETSc, Trilinos and Eigen for numerical linear algebra, ParMETIS and SCOTCH for mesh partitioning, and MPI and OpenMP for distributed computing.
As of May 2022, DOLFINx is the recommended user-interface of the FEniCS project.

History

The FEniCS Project was initiated in 2003 as a research collaboration between the University of Chicago and Chalmers University of Technology. The following institutions are currently, or have been, actively involved in the development of the project

[|DOLFINx]

Since 2019, the core components of the FEniCS project have received a major refactoring. resulting in . DOLFINx supports many new features not available in the old DOLFIN interface, including:
  • Arbitrary degree finite elements on interval, triangle, quadrilateral, tetrahedral and hexahedral cells, including unstructured meshes without special ordering;
  • Meshes with flat or curved cells;
  • Custom partitioning of cells across multiple processes;
  • Parallel IO via Gmsh, VTK,, and ;
  • Assembly and solvers using different floating point scalar types, including complex types;
  • Assembly of custom element kernels written using Numba;
  • Interpolation of functions into arbitrary function spaces;
  • Interpolation between function spaces built on different meshes, including meshes using non-affine geometry;
  • Ability to non-intrusively support different linear algebra backends, e.g., NumPy, PETSc, Trilinos, and Eigen;
  • User-defined finite elements.