Learning Forum

[heattransferguy]

Hello community,

this question has been asked before but unfortunately it didnt really get resolved (https://innovationspace.ansys.com/forum/forums/topic/why-does-fluent-become-slower-when-automated-and-increasingly-slower-after-every-design-point/). I ran into basically the same problem like the linked post and I am clueless how to solve this.

I am simulating a wind flow over a cavity and im trying to get insights about the heat transfer coefficient on the back wall of the cavity depending on a bunch of geometry parameters. Running a design point manually (open DM to generate new geoemtry, generating the mesh, open Setup and running the case) takes between 5-10mins to converge. But selecting multiple design points in the 'Parameter Set' Table of Design Points and 'Update Selected Design Points' takes extremely long to solve (around 4x the time running the design point manually). 

Did anyone made a similar experience using the Parameter Set and knows why this is happening/how to solve this?

Thank you!

[Federico]

How are you initializing each solution for each design points when Updating Selected Design Points? How does that compare with your initialization when you do it manually?

Update of Design Points can initialize a solution as defined in the solver or use an existing solution from a previous design point. If your changes in the geometry are too important, perhaps the latter can take longer to converge

[esbenn]

I am also having that experience with updating design points from workbench instead of manually doing it in Fluent. I'm running a simple flow simulation on the DES realizable k-epsilon model across a bunch of different flow speeds. Running it manually takes 10 minutes max but updating design points instead takes like 5-10x as long per design point. I've only got 1 million elements in my mesh too, so not it's not a crazy  demanding use case either.