APES Overview

The APES Suite contains several tools that allow for a flexible computation of mesh based simulations. The central component is the TreElm library that implements the mesh handling.

In general the workflow in APES is to first generate a mesh (with Seeder), then run the simulation (for example with Ateles), and finally post-process the results (harvesting, for Ateles with atl_harvester. Mesh generation can be done by the solver itself for simple meshes. In this case there is no need to run Seeder beforehand. Similarily, the post-processing can also already be done online by the solver. However, including the post-processing in the solver run itself is only advisable for small setups or smaller subsets from the overall data, as each process will generate a separate visualization file, which is not suitable for large scale runs with many processes.

All tools are configured with the help of the Aotus library, which provides access to Lua scripts. Lua is a complete scripting language, which allows you to use variables and functions in the configuration. This enables us, for example, to explicitly state relations between quantitites.

The components of APES relevant to Ateles:

  • Aotus: Library to interact with Lua scripts.
  • Treelm: Library to manage the octree mesh it also includes the libharvesting for post-processing. Treelm incorporates Aotus as a subrepository.
  • Polynomials: A library providing functions for Legendre polynomials.
  • Seeder: Mesh generation in the Treelm octree format.
  • Ateles, the Discontinuous Galerkin Solver implemented within APES.