search

MatthewGrim / CFD_Codes

4 stars 4 forks source link

Implement 1D VOF method for ideal gas simulation #13

Closed MatthewGrim closed 7 years ago

MatthewGrim commented 7 years ago

This issue documents to the finite volume fluid solver to add the capability to model multiple fluids in a single simulation.

This involves recording the density of each individual fluid component, as well as its material properties, in order to combine ideal gas equations of state for a particular mixture in a single cell.

The expected solution generated by this issue will do nothing to track the interface.

MatthewGrim commented 7 years ago

Locally, I have a version that handles mass fluxes with Godunov's method and random choice, but is failing for HLLC and MUSCL.

I need to fix the failing flux calculators and also account for updating gamma in the update_states function, as it currently is not done...

MatthewGrim commented 7 years ago

The shock tube problems are working in 1D, for all solvers, which I'm taking to mean that this work is done.

There are oscillations in the mass fractions for the strong rarefaction case that I've had to suppress using a cap on the bounds [0.0, 1.0]. This could be because we're breaking the continuity assumption, or for some other reason that I'm currently unaware of to do with high order scheme.

Any future work on VOF high order schemes should try to answer this question.

MatthewGrim commented 5 years ago

An additional note - in some of the Sod problems, I observe oscillations at the shock front that have been introduced by the VOF method. Again I think this is to due with numerics but have not explored the problem in detail. This also needs to be explored, if I continue this work.

Ideally the code would have the option to run with mass fractions to ensure the original solution is still correct.