Skip to content

CHT Convergence Troubleshooting

Video topic: how to actually tell a steady CHT run has converged, and the concrete levers to pull when it won't.

Talking points:

  • Don't trust residuals alone — probe velocity/pressure/temperature at surfaces/points of interest and watch for values to stop fluctuating; that's a more reliable convergence signal than a residual plot. (SimScale blog)
  • Residual targets and interface heat-flux balance — typical targets of 1e-4 to 1e-6 (energy to 1e-6), plus explicitly checking heat-flux balance across the solid-fluid interface rather than relying on residuals in isolation. (Neural Concept)
  • Stage the validation — sanity-check materials, boundary conditions, and heater power on a simplified run (e.g. "dry air first") before committing to the full physics; catches setup errors cheaply. (Ansys FENSAP-ICE)
  • Where NOT to put a heat source — placing a heat source directly at a CHT interface patch causes convergence problems; keep sources away from interfaces. (Ansys FENSAP-ICE)
  • Relaxation factor tuning — solid-region enthalpy often needs different (higher) relaxation than fluid momentum/pressure — a concrete example: h_solid=1 vs. U=0.4, p=0.3. (SimFlow)
  • Temperature bounding — bounding T to a physical range (e.g. 290–600K) is a practical stabilization trick during solve. (SimFlow)
  • Troubleshooting order — check mesh quality first, interface type second, numerics last, rather than reaching for solver tweaks immediately. (Ansys forum)

Sources:

  • Simscale — blog entry (convergence probes)
  • Neural Concept — convergence/residual entry
  • Ansys — FENSAP-ICE (staged validation, source placement) and forum (troubleshooting checklist) entries
  • SimFlow — relaxation factors, temperature bounding entry