SimScale CAE Forum

Still can't fix the float point expectation error

I’m posting this again because i REALLY REALLY need this to be fixed. I have to get this done by next week! Also PLEASE don’t reply to me to "check 1. Error: Floating Point Exception | SimScale Knowledge Base
2. Most Popular Errors of Simulation - #3 by jprobst " Because if I’m posting this AGAIN it means I already have tried all the solution meantion in these articles.

My runs always fail with the error Floating point exception error, I’m running an incompressible so the pressure can be 0. The kinetic energy is non 0. I have no idea how to fix this error. Can anyone please check the project for me?
Here is my project: https://www.simscale.com/workbench/?pid=3741705651761276277&mi=spec%3A731f8338-3ad5-4012-94d9-588659950b35%2Cservice%3ASIMULATION%2Cstrategy%3A39&ps=analysis%2FIncompressible–oneOf%2FinitialConditions%2FomegaDissipationRate

Hi,

A few comments, before talking about the floating point exception:

  • I noticed a few issues in the setup. For instance, the velocities are currently pointing in the wrong direction:

  • 0.9 for delta t is likely going to cause issues for your simulation, since we cannot reach 1000 iterations with steps of 0.9.

  • Now, about the floating point exception, the issue here is likely related to the unclean CAD model, which results in a mesh with very poor quality (see the recommended metrics here).

This is likely breaking the calculation at some point, resulting in a floating point exception. Even if the simulation didn’t fail due to floating point exception, it would definitely diverge with the current metrics.

For instance, I tweaked some of the mesh settings, and the simulation runs normally, but it fails because the mesh quality is very poor:

This is likely caused by all of these tiny artifacts, that are all over the model. To give you some perspective, the smallest face is currently 1e-9 m².

For cleaner models, the metrics would be better.

In short, CAD clean up is required.

Good luck!