gradient adaption
Hello
when i use gradient Adaption .the second time step has big Courant number and then strat to less.why?how can i solve it?
i simulate drop impact and its transition.what is the best gradient adaption ferquency?
when i use 10 or smaller my solve is not converge?
Best Answer
-
Rob
UKForum Coordinator
When you adapt the mesh you more-or-less half the cell length (look at the adaption section in the User's and Theory Guide), this obviously has implications for the courant number. At the same time the refined mesh may pick up some new flow features further reducing the time step required for stable solving. Have a look at the flow field and reduce the time step to suit. Adaptive time stepping can work in some cases, but with dynamic adaption you may find it changes too quickly for the solution: fixed tends to be a little slower but much more stable.
Answers
When you adapt the mesh you more-or-less half the cell length (look at the adaption section in the User's and Theory Guide), this obviously has implications for the courant number. At the same time the refined mesh may pick up some new flow features further reducing the time step required for stable solving. Have a look at the flow field and reduce the time step to suit. Adaptive time stepping can work in some cases, but with dynamic adaption you may find it changes too quickly for the solution: fixed tends to be a little slower but much more stable.
thanks
@Rob
how can i choice best frequency ?
Does it have a formula?
It does not have a formula. Best to do it frequently if you need to have a crisp interface.
I'd suggest somewhere between 10-20. Too often and you slow the solver down with unnecessary remeshing, too infrequently and as @DrAmine says you'll lose the interface.