Transient-FSI, Density-based, compressible -very wierd behaviour, help
I'm having lots of difficulties with setting up a FSI case with system coupling and the density-based solver. Using the pressure based solver I get converged solutions quite fast, but my intent is to simulate pressure shock waves hitting and propagating from a structure.
The fluid is hydraulic oil, and it is modeled as compressible using a bulk modulus compressibility and speed of sound UDF. I'm also using a velocity type inlet governed by an UDF.
When I solve the case with the pressure based solver, i get converged solutions, and my structure indeed responds to the fluid in a expected manner, but I'm not getting the desired pressure-wave effects.
When I run the case with the density-based solver (be it implicit or explicit formulation), I do get the propagation of the pressure waves, but the data transfer seems to be broken somehow. My structure deformation is almost nonexistant (even thou the pressure difference is very similar to the PB solver), and what is even more shocking to me...it deforms in the opposing direction as it should...what the?
Above all that, when I tried to use the explicit DB solver with explicit time formulation, I received a warning that I can't use global time steping for incompressible flow, but my flow IS compressible...I've read through all of the documentation twice regarding System coupling, the solvers etc...and I just can't find any reason what this is happening.
1. Why is the system coupling data transfers not functioning properly with the density based solver? I've tried changing the participants order, but to no effect.
2. How can one overcome the "You can't use global time stepping for incompressible flow" error, why does it even show up?. My flow, as stated before, is compressible.
3. If you'd have any experience in simulating pressure waves, could you lend me some tips? Am I on the right track? Am I missing something obvious?
I can answer Q1. The force sign passed from Fluent to System Coupling is incorrectly reversed with the density based solver. To correct this, turn on beta features in Workbench then in System Coupling add the Expert Parameter DataTransfer_ScaleFactor_Force with a value of -1.
thank you so much for your help, though it seem so awkward for such an obvious bug to pass through software testing :)
|All times are GMT -4. The time now is 15:10.|