Changing a material property causes the V.O.F. simulation to diverge.
|
A VOF simulation that was converging, now has convergence problems.
Check for the following error at the end of your FDSTAT file: ***** E X E C U T I O N E R R O R ( 14104) TIME STEP DECREASING FOR ( 5 ) CONSECUTIVE TIME STEPS EXECUTION ABORTED If you are using a variable time-step, then adjusting the values of UCHAR, VCHAR, and WCHAR in the TIMEINTEGRATION command to reflect the characteristic velocities of the problem may yield convergence. If these values are unknown, try a linear extrapolation to a new value of UCHAR using the base value, e.g.: UCHAR_new=(Re_new/Re_old)*UCHAR_old Also, U(V,W)CHAR values less than zero forces the code to calculate these values at each time-step. Increasing the value of the WINDOW keyword may also help with convergence. Please see the TIMEINTEGRATION command in the FIPREP manual for more details in this regard, as well as the use of characteristic variables for other degrees of freedom. NOTE: Using a fixed time-step DT often helps with the convergence of V.O.F. simulations. The value of DT should be based on the COURANT number, the characterisitc velocity, and the size of the smallest element. |
||
|