Challenge 6 Simulation Error

I was able to prevent a crash by applying residual smoothing, but this was not one of the suggested settings so there still might be an error somewhere in my setup. Atleast I am able to already analyze some results right now, even if they might not be 100% correct(mass flow converged well though this time).

Hi Mick,

good you mentioned the point that you are aware of the meaning of the warning message. Good also you tried the setting with the residual smoothing yourself! I double checked the settings and the domain setting in the numerical tab should be:

I was wondering whether we missed that setting when we created the hand-out.

In the hand-out we only specify the settings, which need to be changed compared to the default settings (with only a few exceptions). After investigating the case a bit further, I noticed that the default settings potentially can vary based on the mesh input. In this case you would have these settings, which I assume you got:

Explanation for the different default settings

When you create a new simulation setup, you choose a solver and add a domain. At this point of time the recommended default settings are set based on:

• Chosen solver
• Provided grid if present

If no grid is present at this point of time, the general default settings with CFL=1.5 (CFL=1 for CGI) and deactivated residual smoothing are applied. If for challenge 6 the mesh is present, Omnis detects that a structured Autogrid is used for domain 1 and uses the recommended default setting with CFL=3 (CFL=3 for CGI) and residual smoothing activated.

What to do now?

• Either please change the settings to CFL=3 (also CFL=3 for CGI in advanced mode) and activate the residual smoothing. Check that you mass flow at inlet and outlet converges to 20.7357kg/s.
• Create a new simulation (no copy & paste), select the Open solver and add the domain. As the grid is present the default settings with CFL=3 should be set.

Please excuse this issue. We did not see this coming, but will note it down for the next round of the CFD Online Course.

Best regards
Manuel

Hi Mick,

good you mentioned the point that you are aware of the meaning of the warning message. Good also you tried the setting with the residual smoothing yourself! I double checked the settings and the domain setting in the numerical tab should be:

I was wondering whether we missed that setting when we created the hand-out.

In the hand-out we only specify the settings, which need to be changed compared to the default settings (with only a few exceptions). After investigating the case a bit further, I noticed that the default settings potentially can vary based on the mesh input. In this case you would have these settings, which I assume you got:

Explanation for the different default settings

When you create a new simulation setup, you choose a solver and add a domain. At this point of time the recommended default settings are set based on:

• Chosen solver
• Provided grid if present

If no grid is present at this point of time, the general default settings with CFL=1.5 (CFL=1 for CGI) and deactivated residual smoothing are applied. If for challenge 6 the mesh is present, Omnis detects that a structured Autogrid is used for domain 1 and uses the recommended default setting with CFL=3 (CFL=3 for CGI) and residual smoothing activated.

What to do now?

• Either please change the settings to CFL=3 (also CFL=3 for CGI in advanced mode) and activate the residual smoothing. Check that you mass flow at inlet and outlet converges to 20.7357kg/s.
• Create a new simulation (no copy & paste), select the Open solver and add the domain. As the grid is present the default settings with CFL=3 should be set.

Please excuse this issue. We did not see this coming, but will note it down for the next round of the CFD Online Course.

Best regards
Manuel

Hi Manuel,

thank you very much for this detailed answer.

This must be the problem, because as you say the default settings are different in my setup and look exactly like in your second screenshot.

I will apply the correct settings in the numerical tab which should certainly fix my issue.

Best regards

Mick