Speeding up a transient solution

A number of techniques may be used to speed up solution convergence for transient runs.

A number of techniques are available to speed up slow transient runs. The most common cause is a small integration time step. Often this occurs because the default DT = 0 is specified, and RC_MIN is very small. To get around this, specify a physically realistic DT value (e.g., DT = TF/100 or DT = TF/1000) and use an implicit technique.

To determine a suitable DT for your model, re-run your problem with DT/2, and compare the results. If the results are unacceptably different, lower your DT. Lower DT values should yield more accurate results.

Another possible cause of slow transient runs is that the temperature difference convergence criterion of 0.001 is too tight. You can alter it with a Card 9 PARAM TDIFS, but make sure your accuracy is acceptable.