K and omega residuals

Project Support
1

Hi everyone,

My values for k and omega are converging very very quickly: flattening out within 50 s. What could I possibly be doing wrong for the values to be doing that? I must be solving it incorrectly.

Thanks in advance.

2

Hi there @pippatreacy123!

Sometimes it is not always about the residuals itself but to see also if other relevant quantities like force, velocity and so on - where result control items can become quite useful!

Residuals are however good to have a rough estimate to where your simulation is heading. At first glance there seems to be nothing wrong with your setup. I would personally reduce the domain a bit. CFD Squad, feel free to add your two cents here.

Best,

Jousef

3

Hi,

Thanks for getting back to me so quickly.

I made the domain larger as I thought there may have been interference with the airfoil and the walls. What are the limitations of making the domain too large besides an increase in computational time.

Thanks!

4

While I (sometimes) like this approach the problem really is to avoid big meshes or even boundaries which are too far away having no influence whatsoever on the simulation but you are absolutely right that this is a bit of a trial & error thing to do and we tend to “over engineer” things just to be safe. But try to think more economically and safe some core hours here and there.

Start with a smaller one and for that really take examples from the public projects library and try to reverse engineer them. First look at the results, convergence plots etc. and see if that’s something you need. Some people might have gone through all this pain of tweaking the settings so you just take this information, saving you plenty of time and run your own simulations with that knowledge.

Hope that makes sense :slight_smile:

Jousef