Fluid mechanics problem

[AliMadayen]

I'm trying to simulate a 2D channel with a flap like obstacle in middle. I generate the grid in another application and export to the FEBio using Gmesh format.
I assign pressure difference to the left and right and appropriate zero velocities to other boundaries. The problem is when the waves travelling from the inlet to the outlet hit the obstacle, something similar to a source/sink appears on the top of obstacle and ruins the solution. Interestingly, the solution does not diverge and keeps on going. I have tried every type of mesh (tetra, hexa), with or without near wall resolution. I have attached an example.
Can somebody please check my case and help? I very much appreciate it.

Compressed xplt: https://drive.google.com/file/d/16T1...ew?usp=sharing

[ateshian]

Hi,

First, please keep in mind that CFD analyses are dynamic, so prescribing a fluid pressure as a step-wise function of time will cause oscillatory behavior because of the sudden acceleration. So the first thing I suggest is to attach a sigmoidal loadcurve to your inlet pressure that rises from 0 to 1 (scale factor), say over a time interval of 1 (since your analysis runs to time 10).

Second, as a general rule I recommend adding a boundary layer mesh on no-slip boundaries in CFD analyses. (Mesh->Boundary Layer)

With these two changes I was able to get your model to work, see attached ( Model1-gaa.fs2 ). Keep in mind that, under a prescribed pressure, the fluid will accelerate for a long time in this problem, until viscous dissipation is sufficiently elevated to provide the frictional force necessary to reach terminal velocity. But the viscosity of the fluid is so low that this limit will not be reached before your analysis simply breaks down due to turbulent flow.

Best,

Gerard

[AliMadayen]

Thanks for your prompt reply Gerard

I ran your model and I faced the same problem (attached picture). I just saved and clicked run.
Is there any problem with my installation?

[ateshian]

You have a peak velocity at the tip of the obstable which is skewing the color map. Just use user constraints on the Color Map, such as Min range type (user): 0 and Max range type (user): 0.05, and display the X-nodal fluid velocity. You will see a more distributed profile. Also, you can select a face on the inlet boundary and another on the outlet boundary, then New Graph and plot (e.g.) the X-nodal fluid velocity, you will see that the inlet and outlet values nearly match.

Best,

Gerard
Model1.png

[AliMadayen]

Thank you very much for your help.
I managed to solve the problem with my models.
But there is one problem that I can't solve, I have attached the model and picture. I prescribe a sine wave velocity input at the inlet but the inlet seems that is unable to affect the domain.
I have changed value of the time step and the results is same.
I'd be grateful if you could help me with this last problem.
Edit: I have also tried both negative and positive values for normal velocity. both lead to the same problem.

[ateshian]

Hi,

I was able to get your model to run by making the following modifications: I deleted the zero-dilatation BCs on the four inlet boundary curves. Instead, I checked the box "prescribe_rim_pressure" in "fluid normal velocity". With these changes the model ran to completion.

Best,

Gerard