Learning Forum

[Vu Le]

Context: I'm attempting to simulate fluid flow in a heat pipe. It's a two-phase VOF model with k-omega. I'm running the Lee model for evaporation-condensation with alternative energy treatment enabled for transient calculations using PTL table. The relaxation parameters in the Lee model are 0.1 evaporation (default) and 10,000 condensation. I have two meshes of the heat pipe (27k elements and 45k elements). Both meshes have inflation layers near walls and the mesh quality is good (skewness and orthogonality are good).

My residuals look fine until everything suddenly spikes and the sim fails. I reduced my under-relaxation parameters but everything still abruptly diverges. There's no indication of imminent failure in the residuals or physical variables.

I'm at my wits end here. I could try to increase the mesh density but other papers I've read have used similarly sized meshes to great results. I'm not sure what else I can do if there's nothing in the logs for me to troubleshoot.

[Rob]

Can you restart from a saved data set and run beyond the failure point?  If not, review the flowfield to see what's going on. How does the fluid temperature & velocity look, are you at risk of drying a section of the wet domain out? Ie is the solution heading towards a result where something could change very quickly?