Beginner’s guide: What is Dune?

Dune (Distributed and Unified Numerics Environment) is a modular toolboy for solving partial differential equations (PDEs) with grid-based methods. It supports the easy implementation of methods like Finite Elements (FE), Finite Volume (FV) and also Finite Differences (FD).

Dune is a free software, which is licensed under GPL (Version 2) with a so called “Runtime Exception”. This license is comparable to the one under which the libstdc++ libraries are distributed. So it is possible to use DUNE also in proprietary software.

The basic idea of Dune is to create slim interfaces, which allow an efficient use of existing and/or new libraries. Modern C++ programming techniques allow very different implementations of the same concept over a common interface with very little effort. Thus Dune provides efficiency in scientific calculations and supports applications for high performance computers.

Special highlights are:

• a generic grid interface, which allows to connect a number of very different grid implementations.
• The iterative Solver Template Library, which contains an algebraic multigrid preconditioner.
• High-quality interfaces for experimental and test functions and generic discretization modules.
• Python binding for the full-grid interface and a flexible concept for connecting user modules.

Dune is based on the following basic principles:

• Separation of data structures and algorithms by abstract interfaces. This offers more functionality with less code and also guarantees the maintainability and extensibility of the framework.
• Efficient implementation of these interfaces by generic programming techniques. The static polymorphism allows the compiler to make further optimizations, especially function inlining, which in turn allows the interface to have very small functions (implemented by one or a few machine instructions) without serious performance degradation. Essentially, the algorithms are parameterized with a specific data structure and the interface is removed at compile time. Thus, the resulting code is as efficient as if it had been written for the specific case.
• Reuse of existing finite element packages with great functionality. Especially the finite element codes UG, ALBERTA and ALUGrid were adapted to the Dune framework. Thus, parallel and adaptive grids with multiple element types and refinement rules are available. All these packages can be linked together in an executable file.

Modules.

The framework consists of a number of modules that can be downloaded as separate packages. There are a number of core modules that are used by most other packages. The core module Dune-Grid already contains some grid implementations and more grid managers are available as add-on modules. Main discretization modules, which provide the infrastructure for solving partial differential equations with Dune, are available as separate modules. The modular structure of Dune allows to use only a small part of the modules (e.g. only the solver module Dune-istl or the module Dune-localfunctions with form functions, without using e.g. the Dune-Grid or a Full Discretization module).

Thank you very much for your visit.