traction calculation in fsiFoam
 

Hello everyone,


i have been reading a lot of threads/post in this forum and it was a LOT of help so far. So thank everyone contributing to this forum.



I have been working with foam-extend-v4.0 and linked compressibleInterFoam (compressible two-phased flow) with fsiFoam (strongly coupled fluid structure interaction). The fluid side seems to work fine when comparing the results to the original compressibleInterFoam solver. However when i actually implement a solid structure which gets deformed it seems that the deformation is way bigger than it should be. I checked the source code on how the traction/stress on the solid is calculated:


In the tractionBoundaryGradient.C file it says on line 120:


traction = 2*mu*(n & symm(gradField)) + lambda*tr(gradField)*n;


The total traction (stress?) is afterwards updated to (line 463):


Traction = (traction - n*pressure);


As the traction has the unit [kg/m*s^2] the pressure must have the same unit. To me this doesnt make sense as fsiFoam is originally coupled with only incompressible fluid solvers where the pressure is defined as p/rho hence the units are [m^2/s^2]. Should the stress not be calculated like
Traction = (traction - n*pressure*rho) for incompressibel cases? Am i missing something? I assumed that i had to divide the pressure for compressibleInterFluid with the density at some point for the stress/deformation calculation.


Second i was wondering if there is any built in feature i am missing to say that for a certain initial pressure (like 1bar) the stress -> deformation is zero?


I would really appreciate some help/insights on this problems.


Best regards


Paul