[CleverBoy]

Hello everybody,


I am trying to simulate a case using SU2-Nemo with Mutation++ and I was taking TestCases as reference. In the test cases, they are using "SOLVER= NEMO_NAVIER_STOKES" and "MARKER_EULER= ( Euler)" at the same time.
Shouldn't they be using "MARKER_HEATFLUX" instead, since the viscosity effects are not ignored?


I tried to run my case with both heatflux and Euler BCs and I got completely different results. Which one should I be using in su2-nemo with m++?

[wallym]

Hi Cleverboy,

You are correct with regards to Euler walls compared to Heatflux/Isothermal wall. I believe the test case used both an Euler wall and a heatflux wall. You may chose whatever boundary fits the type of problem you are attempting to simulate.

In the case of the "test cases" provided (especially for NEMO) on the website are used as regression tests. These aren't necessarily production/physical cases, but are used to make sure the developments in the code don't break other functionality.

Long story short, if you are using the the NAVIER_STOKES options, you probably should have Isothermal/Heatflux walls. However, not all need to include the viscous effects. If you have a specific case when using NEMO, id be happy to provide more insight into boundary conditions.


Wally

[CleverBoy]

Originally Posted by wallym

Hi Cleverboy,

You are correct with regards to Euler walls compared to Heatflux/Isothermal wall. I believe the test case used both an Euler wall and a heatflux wall. You may chose whatever boundary fits the type of problem you are attempting to simulate.

In the case of the "test cases" provided (especially for NEMO) on the website are used as regression tests. These aren't necessarily production/physical cases, but are used to make sure the developments in the code don't break other functionality.

Long story short, if you are using the the NAVIER_STOKES options, you probably should have Isothermal/Heatflux walls. However, not all need to include the viscous effects. If you have a specific case when using NEMO, id be happy to provide more insight into boundary conditions.


Wally

Hi Wally,

Thanks for the detailed reply and sorry for my late response.

I am trying to simulate a re-entry capsule at mach 15 with Navier Stokes solver and therefore I am gonna need the heatflux boundary condition.

For now I didn't come across any problems regarding BCs but I always end up with "Warning. The initial solution contains 'xxx' points that are not physical" error; doesn't matter how much I increase the number of cells in the mesh. It only works if I decrease the CFL number. I am using Salome for meshing I gotta say that mesh of the domain is not perfect but it's good enough.

If you have time you can still take look at my cfg file and tell me about any upgrades that I should make.

Thank you for your time.

[wallym]

Hi Cleverboy,

Your config file looks fine. NEMO, at times, struggles for some blunt body cases. Some suggestions would be to try to reduce the CFL (we are working on improving the implicit formulation). I would also suggest trying the MSW numerical scheme...especially for the beginning iterations of the simulation.

I hope this helps,
Wally