||March 15, 2013 11:04
Pressure jump and vanishing solver
we are simulating fluid-structure interaction (FSI) using OpenFOAM (a modified interDyMFoam solver) and an external structural solver, coupling them via MPI. After a certain simulation time one of the OF solvers (either for cellDisplacement or U) "looses" one component. Usually (in quasi 2D) we get something like
DICPCG: Solving for cellDisplacementy, Initial residual = 5.561218854736067e-07, Final residual = 5.561218854736067e-07, No Iterations 0
DICPCG: Solving for cellDisplacementz, Initial residual = 1.058528227755323e-06, Final residual = 3.090254563914258e-07, No Iterations 1
DILUPBiCG: Solving for Uy, Initial residual = 0.006568920370747625, Final residual = 2.648992650684406e-10, No Iterations 2
DILUPBiCG: Solving for Uz, Initial residual = 0.002921793460135126, Final residual = 2.591064855961802e-11, No Iterations 2
then suddenly one of those components does not seem to be solved any more (one of the log messages missing). At the same time a pressure jump occurs: The relative pressure distribution still looks OK, but the absolute value is much too high, going back to normal over a couple of time steps.
Has anybody ever experienced something like this? Any hint or rough guess?