Problem with restart in FSI unsteady simulation
 

Hi!


I'm currently working in a FSI software implementation using the python bindings from the SU2 pywrapper. As a baseline for my implementation I used the launch_flatPlate_rigidMotion.py found inside the py_wrapper_folder in TestCases. Prior to execute the FSI unsteady simulation I run a steady case with SU2 in order to use the converged parameters of the steady solution as the initial condition for the unsteady case. To do so, I am using the restart solution option in the PROBLEM DEFINITION part of the configure file (RESTART_SOL= YES) as well as the corresponding options in the UNSTEADY SIMULATION part (UNSTEADY_SIMULATION= DUAL_TIME_STEPPING-2ND_ORDER and UNST_RESTART_ITER= 2). When I run the unsteady simulation with these parameters, SU2 seems to be reading properly both restart files needed for the second order dual time stepping. However, a closer look in the results shows that the first unsteady iteration does not start from the converged parameters ( Residuals, Cl and Cd) coming from the steady case.


Trying to track down the source of the error, I found that the the restart works as expected just running an unsteady case without grid deformation. What means that the problem comes from the way of applying the neccesary grid deformation for a FSI simulation (GRID_MOVEMENT_KIND= FLUID_STRUCTURE). I checked the source code defining the grid movement ( grid_movement_structure.cpp and iteration_structure.cpp) and I found that somehow for the FLUID_STRUCTURE case the stat_mesh boolean variable (static mesh) was set to TRUE. As a first possible solution a changed that to be FALSE, in order for the code to access what I found to be a necessary function to properly execute the restart.


This solution is working so far, but I'm not completely sure if I am overcomplicating things here and there is a simpler solution to this issue. So... have anyone experienced a similar problem??


Thank's in advanced :)



David.

Hi

How did you solve it?