cpraveen / fem

Codes for finite element method
23 stars 13 forks source link

Isentropic vortex on unstructured grid #1

Open cpraveen opened 3 weeks ago

cpraveen commented 3 weeks ago

On unstructured grid, there is some bug which causes vortex to distort when it crosses the boundary.

This does not happen with cartesian grid, whether generated inside deal.II or in gmsh.

cpraveen commented 3 weeks ago

@JR-Khan can you test this. Try both cartesian and unstructured grid. See inside grid.geo file for how to get these two types of grids.

JR-Khan commented 3 weeks ago

Here are some outputs from different case. Please have a look

https://www.dropbox.com/scl/fo/mz1q30iccotzmxgvr2ia8/ACsmk1UecXqbzwOoAJ1G1QA?rlkey=nhylkpb2ydnp82ap9tsazo5mr&st=pi0otjog&dl=0

cpraveen commented 3 weeks ago

If you are computing vorticity within visit, then it is not very accurate and at partition boundaries, it will be still less accurate. This is because visit will do some finite differencing.

Do you observe errors in primary quantities when crossing partition boundaries ?

JR-Khan commented 3 weeks ago

No.