Version 1.0 as published on zenodo

These examples reproduce the results shown in

Henry von Wahl, Thomas Richter, Stefan Frei, Thomas Hagemeier:
Falling balls in a viscous fluid with contact: Comparing numerical simulations with experimental data
arXiv, 2020

This version is compatible with Gascoigne 3d, Version 1.0, as published on

www.gascoigne.de

The complete compilation of test cases in this archive is also
available as one of the "Gascoigne 3D - Documented Examples", called
"Rigid-Body-Interactions". Again, see www.gascoigne.de for updates.

*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*

This archive builds four executables

- StationaryBall to reproduce the Stationary flow test case. Run this
  example in the directory 'StationaryRubber' by issuing
  '../build/StationaryBall'

- NonstationaryBall to reproduce the test case with a prescribed
  motion. Run it in 'NonstationaryRubber' by issuing
  '../build/NonstationaryBall'

- FallingBall to reproduce the 2d benchmark problem. In 'FallingBall'
  start them by issuing '../build/FallingBall falling-rubber22.param'
  or '../build/FallingBall falling-ptfe6.param'

- FallingBall3d to reproduce the 3d benchmark problem. In
  'FallingBall3d' start by issuing '../build/FallingBall3d
  falling-rubber3d-22.param' 

The most important parameters in the corresponding param-files are

- prerefine to set the initial steps of global refinement. The
  FallingBall 2d benchmark problems are using 3,4 and 5 levels of
  refinement. The 3d test case uses 2,3 and 4 levels.

- dt to indicate the time step size. For each step of mesh refinement
  we performed 2 steps of temporal refinement such that fourth order
  convergence (in h) gets possible.

For more information we refer to www.gascoigne.de and the
corresponding Documented Example.


