Niklas Neher comments

Results 61 comments of


                                            Niklas Neher

Reopen #363: Hydrostatic water column (FSI)

It's a validation example from Sun et al. (2019), as well as from O'Connor and Rogers (2021).

Plot quantities like pressure

I'm fine with doing this in 0.2. IMO, plotting of quantities is usually done in paraview.

Resolve conflicts between Trixi.jl and TrixiParticles.jl

Hm maybe we need an abstract type `SemidiscretizationTrixiParticles` ?

Fix `BoundaryModelDummyParticles()` for FSI

see #262 Note, that the simulation `falling_spheres_2d.jl` needs about twice the number of time steps with #262 @svchb @efaulhaber

Calculation of the continuity density

Adami et al produces very good results with an inter particle averaged pressure: ![image](https://user-images.githubusercontent.com/73897120/236232692-84f5e1df-9a26-4ad9-b43e-74f3e0ae21c1.png) with the number density formulation. They do not use any artificial viscosity or empirical parameters but...

Calculation of the continuity density

Indeed. The TVF (transport velocity formulation) of Adami is only for internal flows. But Ramachandra promised a solution for every purpose. I'm very curious.

Calculation of the continuity density

Ramachandran uses also this average pressure, but claims that "that this results in signiﬁcantly improved results that outperform the traditional TVF scheme."

Use mandatory keyword arguments instead of positional arguments?

In some cases, it makes sense to have kwargs. But having a hard rule to only allow kwargs in the API might be too restrictive, and in some cases, it...

Use mandatory keyword arguments instead of positional arguments?

Ah , I see. Then it makes sense for viscosity, eos and setups

Do we need `current_coordinates` in the solid container?

I'm not 100% sure about this, but in `current_coordinates` the moving and also the fixed particle coordinates are stored whereas in `u` only the moving particle coordinates are stored.