Learning Forum

[jimbo]

The system coupling cannot complete convergence. Is there any expert who can provide a better way to solve this error? Thank you very much

[Rahul]

Before coupling your participant problems, verify that each one of them can be solved individually, as decoupled solutions. By running each one independently, you can get an approximation of their expected inputs to a coupled analysis.

Ensure that you good mesh quality and mapping of coupling interfaces.

Also, refer to Building Up to a Coupled Analysis, Improving Coupled Analysis Stability

[sheng]

 

Hi, I am also using system coupling. I can try coupling and it works. The MAPDL part keeps converged, but the fluid flow part cannot converge.

I want to know why. 

thx

 

[Rahul]

When running System Coupling and it stops running, the first thing to do is examine the output files for each solver to look for clues to what happened. For Fluent, look at the transcript .trn file. For Mechanical, look at the .out file. For CFX, look at the .out file. Also, look at the System Coupling log file, scl. You can also try re-running the case and stopping it just before the timestep when it crashes, then examine the mesh in Fluent or CFD-Post to get more clues.