C0 continuous data

General
1

Hello all,

Is there a “filter” that can render C0 continuous data from PyFR output (that is by definition discontinuous at the elements).

Why? I’m trying to calculate PSD in space for Taylor-Green Vortex case and there occur peaks on wave-numbers corresponding to the element size.
Kindly suggest if someone has some other solution. I’ve already tried Paraview’s CleantoGrid filter, but no help in this case.

BR,
Vishal

2

Hi,

Is there a "filter" that can render C0 continuous data from PyFR output (that is by definition discontinuous at the elements).

Why? I'm trying to calculate PSD in space for Taylor-Green Vortex case and there occur peaks on wave-numbers corresponding to the element size.
Kindly suggest if someone has some other solution. I've already tried Paraview's CleantoGrid filter, but no help in this case.

So the clean to grid filter will eliminate -- albeit somewhat arbitrarily -- the multi-valued nature of the solution at the interfaces between elements. The result should be C0 continuous.

However, this will probably not resolve your issue. Indeed, it is perhaps worth asking if those peaks you observe in the PSD are really spurious. After all, when the solver 'runs' it is not working to actively eliminate these effects and so -- at some level -- they are really there in the simulation. Of course, they're numerical artefacts, albeit ones which can have a meaningful effect on the subsequent dynamics of the simulation.

If you do want to get a 'clean' result your best bet is probably to construct a global basis (which given the test case is relatively simple) and then project the solution onto this basis.

Regards, Freddie.

3

Thanks Freddie for the input.

I was able to project the results onto a globally continuous basis and the PSD results look good.

BR,
Vishal