SimScale CAE Forum

The meshing algorithm failed with a floating point exception

Hello. I am doing an incompressible simulation on one of my diffuser models. But during meshing, it shows this floating-point error.

I have checked the post regarding the floating-point error which says that my turbulence values are zero. I am and have been using default values for a long time but I haven’t encountered any such problems before. Can anyone guide me through this problem?

(Also you free to check out my diffuser model and give any suggestions on it :wink: )

Link to my project:
https://www.simscale.com/workbench/?pid=2446195873483880936&mi=spec%3A104446e4-1349-407c-be5d-e72884423ed6%2Cservice%3AMESHING%2Cstrategy%3A191&sh=190

Checkout simulation named 30-10 DD for the problem!

Thanks in advance

1 Like

Hey Mihir,
I will have a look and inform you as soon as possible :slight_smile:
Cheers!

1 Like

Hi again,

this issue can be created for many reasons, and most of them are listed here: Error: Floating point exception
However, for cases with material point-based meshes, it is common that this error is created because of the coordinates of the material point. So, what you can do next time, is to try to relocate the point somewhere close.
I used the coordinates below, and the mesh was successfully created:

Cheers!

1 Like

Hello,

I did the changes in this. But is there any specific reason this is happening?

Also, I have placed material point this far in earlier iterations too but no error like this happened then while running sims.

Thanks in advance

Hi again, I did not mention that I changed the bounding box resolution too:
bounding box resolution
I read a while ago in an OpenFoam forum that this happens because the material point is placed on the boundaries of a cell, and it is suggested that the coordinates are set with a lot of decimals, so this doesn’t happen.
If I find this post again, I will let you know :slight_smile:

1 Like

Thanks for the support. I still am not getting the reason for the placement of material point. If I am understanding your point correctly, I shouldn’t place the material point on the edge of any meshed cell. It would be really difficult to figure out where the potential edge of a cell would come before the start of the mesh.

Also, I tried meshing after bringing the material point closer to the body, still, it failed.

Along with you, I would also like to tag @dschroeder and @yosukegb4 to help out as they helped me with my previous question on the forum

Thanks in advance!!

@MihirRayganga,

I unfortunately cannot help with this situation as i dont have much experience with these types of problems. If another power user doesnt know what the problem could be then someone from the simscale staff has to get involved.

Dan

Thanks, Dan for trying. Perhaps I need help from @jousefm.

Thanks in advance for the help

https://www.simscale.com/projects/rgaonkar27/bwb_wing__trial_3/
This project is an aerodynamic analysis on blended wing body and I was trying to determine the flow separation at different angles of attack.
I was playing around with Inflate boundary layer to try to get y+ = 1 ( check mesh named Mesh 1 without relative sizing) but it is showing the error mentioned above.
I tried mesh with default values for Inflate boundary layer and it ran successfully (check mesh named Mesh 1 with relative sizing) but it doesn’t have an appropriate y+ value.
The document page form learn more link has suggestions only if the problem of floating-point exception occurs during simulation run and no suggestions if this occurs during meshing.
Can you suggest ways to deal with this issue

1 Like

Hi there, thank you for using the forum!

Have you considered using the Standard mesher instead? It is a reliable meshing algorithm an it is highly recommended. I can help you set the mesh if you haven’t used it in the past : )

Let me know what you think and prefer.

Best regards,
Fillia

I haven’t used standard mesher but I will try it and update whether I am able to get the required mesh or not

1 Like

Thank you

1 Like

Great, looking forward to the update, let’s see if it works out :grin: Cheers!

1 Like

I tried using standard mesher but was not able to figure out how to turn off relative sizing for the layer addition process. So my only option was to go with relative sizing turned on and I was able to get a mesh with y+<5 in the area of interest on the wing using the previous Hex dominant parametric mesh(Mesh 1 with relative sizing), so my issue is resolved for now.

1 Like

Thank you