I am trying to simulate a wing in water. The meshing seems to be of decent quality now, but i cannot get the simulation to run, it just errors without any log or explanation.
any sugesstions to where it errors out??
Hi @akhm74!
The mesh is too coarse at the moment. I would just follow this tutorial which gives you a good impression on how to create an adequate model: Step-by-Step Tutorial Session 1 - Front Wing Design Study (Part 1). If you have any questions you can ask me anytime.
Cheers,
Jousef
The tutorial is the basis for my project, and it is rather comprehensive.
The simulation did run on a less fine mesh, (LC1 Medium2), but as the boundary layer has been added to the mesh (LC1 Medium 3), it just fails…
I know the bounding box mesh is coarse, but i have refined rather heavily through 3 regions and with surface and feature refining i am down to a “pretty” ok mesh on the geometry. I want to refine to the smaller details on the geometry, keeping core hours to a low.
Hi there @akhm74!
Will have a closer look at the issue tomorrow morning if you have not figured it out yet. I also saw that you could make it work with adjusted settings but I would still argue that the mesh is too coarse and you would have to perform a mesh convergence study anyway so let’s keep the sim now and definitely compare it to one that has double the fineness/half the grid width.
Cheers!
Jousef
Hi Jousef,
I did by mistake delete the sim in question. I am currently recreating the sim. (LC1 Medium 3 - Run 1)
I will also try a smaller bounding box, but with finer mesh initial mesh. I am hitting memory issues with 0.5x0.5m initial mesh, trying a ~0.625x0.625. Less level of refinement should result in a decent boundary layer.
Btw: I have gotten the mesh without boundary layer inflation (LC1 Medium2 - Run 4) to finish without any issues. Seems like convergence is reached at approx 300s (it did require a very long max run time…)
Thanks for now.
It is starting convergence plot, so there is hope…
1 Like
Hi @akhm74,
Just checked your latest run. Looks good so you can proceed to post-process from there and fine tune the results if they are out of margin.
Cheers.
Regards,
Barry
1 Like
The results does seem to be in line with the tank trials and the previously achieved results (ansys - screenplots)
Cheers to all of you
2 Likes