Regarding IBM Implementation on 2D Cylinder Case

[StepTang17]

Hi there,

I was trying out for the 2D Cylinder simulation presented in "A high-order immersed boundary method to approximate flow problems in domains with curved boundaries" using the provided test case of IBM Cylinder located in /Solver/test/NavierStokes/IBMCylinder for the most part I was successful in getting similar Pressure and Velocity Contours for the Volume Penalization IBM method at polynomial order of 3. However I face some significant difference in my pressure coefficient results. By looking at IBMClass.f90 it seems by default the IBM uses the inverse distant weighting interpolation, but it seems I have some significant deviation from the VP IBM results. I was wondering if there are specific control file settings needed to achieve similar results to what is shown in paper ? Many thanks

[loganoz]

Hi @StepTang17! Thanks for showing interests in our results.
@stecolumbus was the developer of this methodology. @stecolumbus could you please solve @StepTang17 doubt? Thanks!

[stecolumbus]

Hi @StepTang17! Thank you for trying the methodology implemented in Horsees3D!
As you correctly said, VP BIM uses inverse distance weigthing for computing the forces. In order to get the quantities on the surface there are 2 possible ways:

1. Add to the .control file, in the IBM section, the string

clip axis = 3

for this specific case it means that the STL is clipped alonz the z axis so that it is completely inside the computational domain.

2. Directly in paraview with the command:

Resample with Dataset.

(Keep in mind that the result may depend a bit on the STL file and the number of triangles).
Let me know if you are still getting problems!

[StepTang17]

Hi @stecolumbus, many apologies for the late reply. I've tried option 1 recently and it actually improved the results. This test was for my research consideration whether to implement IBM in a specific research my lab is interested in.