CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   FSI time (

YY April 25, 2008 12:38

FSI time
Hi guys,

I am very confused with the time in an FSI simulation. I'm basically doing an FSI which is steady for both the structural part and fluid part.

The first time which I could set is the Coupling Time. As I know this does not have any meaning in a static analysis, so I just input the total time as 1s and the substep time as 1s as well. Is this correct?

Second, its the time under solver control, I think that is for the fluid part. People have said that its a pseudo time which I dont really know what it means. Anyway i treat this like a substep, so I use a physical time scale and i use a 0.01s time. Does this mean that when CFX solves in 1 stagger iteration, the whole load step which is 1s will be divided into 100 substeps? and if so, if I set the outer loop max iteration to be 100, would it mean 100 for each substep or 100max for the entire 100 substeps. Please advice.

Another thing is for the under-relaxation factor, if I have convergence difficulty say negative volume, do I lower the under relaxation factor or do I lower the pseudo time for the CFD? I am modelling a flow in a pipe where the pipe can be severely collapsed. I would think that its the displacement of the pipe which is too much that is causing the fluid mesh to crash. I tried lowering this under-relaxation factor but the solution crashes earlier. Why is that so? Thanks guys.

All times are GMT -4. The time now is 16:23.