Heat Transfer

Heat Transfer

How do I get rid of convergence problems with energy equation in Fluent (especially since R15)?

    • SolutionSolution
      Participant

      Besides the tips in solution#2000941 there are some further recommendations: Since Fluent 15, the default cycle-type in the AMG-settings for energy is F-cycle. This can improve the convergence, as shown in the attached slide 1. This is true especially for cases where diffusion is the predominant effect and for cases with high jump in thermal conductivity. However, for other cases this new default could lead to instabilities/crash. In such cases it is recommended to use the old default, which was Flexible cycle. Use Least-Square Cell-Based gradient-reconstruction method, when solving energy equation. See slides 2-4. Maybe the instability comes from ideal gas. Try to start w/o ideal gas and also w/o energy equation. Run a few iterations and then enable ideal gas. Monitor max and min values of temperature in the volume. If limiters are hit, inspect these cells by means of adapt/iso-value. If a bad mesh causes the issues, mark these cells (mark instead of adapt) and in TUI: solve/set/poor-mesh-numerics/user-defined-on-register yes, followed by the name or ID of the register. This will apply poor-mesh-numerics on these cells, even if the quality is ok.

      Attachments:
      1. Improve-Convergence-of-Energy-F-vs-Flexible.pdf