I have observed that when I stop and resume the solver run in SimCFD, after resuming, the residual plots take an unusual fluctuation before stremlining again. From what I have heard, this phenomenon was not present in the CFDesign solver.
Can this anamoly be a reason for divergence down the line?
Vee Bee Filtrations UK
Solved! Go to Solution.
This is actually an expected result when running certain simulations. It is because the momentum terms of the solver need to be solved for again. The residuals should return to the previous trend after a few iterations.
This makes sense since the recalculation of momentum equations will affect all other coupled equations because of seggregated solver's operation-flow (I think). But interestingly, this was not the case with CFDesign 2010/2011! Perhaps some other pressure-velocity coupling scheme was used?
Anyway, if this doen't surely affect the convergence, I'm happy.