User login


You are here

Validation and benchmarking problems for 3D fluid simulation with free boundaries

I have a hard time finding problem sets for validation and benchmarking of methods and codes for 3d fluid simulation with free boundaries. Please guide me to (standardized) problem sets or key litterture that is a suitable starting point. If there are not many standardized tests I would like to initiate a dsicussion about what a good test collection should be. I understand that iMechanica is focused more on solid mechanics than fluid dynamics but the overlap is big.

The best collection I have found so far is the one from SPHERIC ( but there should be more. Meeting the 2D benchmark tests is not a guarantee that a 3D code is working properly

I am missing standardized tests - with established experimental results and/or exact solutions - on

  • 3D objects (sphere, wedge, cone etc) entering fluid surface
  • 3D fluid interacting with flexible solid or jointed rigid bodies
  • wave scattering on static and dynamic objetcs
  • time evolution of large amplitude waves
  • surface wake fields of standard geometries - static and dynamic (standardized ship geoemtry)
  • transients and instabilities of varios 3D flows with free surfaces (shear flow, vortices etc)
  • special tests for high viscosity, viscoplasticity, non-newtonian fluids in 3D and with free boundaries

I am currently collecting litterature on such tests but I can't help thinking that problem sets should already be compiled somewhere "out there". Is it that free surfaces is a too difficult problem (or model dependent) and all methods and software are bound not to agree with experiments? Well, we need the tests to ever resolve this. Thera is an ever increasing number of approaches for these systems - meshfree and mesh based - and it is impossible to judge what approach to use on which type of problem and what artefacts to watch out for.

Subscribe to Comments for "Validation and benchmarking problems for 3D fluid simulation with free boundaries"

Recent comments

More comments


Subscribe to Syndicate