SeiDPierre
commented
1 year ago In this example mmg is called at the first step of the simulation, then every 20 steps. At the first step, it works then at step 22, the number of elements increases in a wrong manner, leanding to segmentation fault.
It looks very fine at step 1:
Then when step 22 is reached:
It looks like I forgot to plot the metric, but it is there, it is just constant everywhere. After mmg calling:
We go from 474 nodes to more than 22000...
Regarding the fluid simulation, after reaching step 2 until next mmg call, the simulation seems to explode in terms of convergence:
MMG Remeshing Process: Remesh finished FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 0.635771; exp.ratio = 0.001 abs = 5.65185 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 0.254644; exp.ratio = 0.001 abs = 2.40528 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 0.136389; exp.ratio = 0.001 abs = 1.21797 FractionalStepStrategy: Fractional velocity iterations did not converge. FractionalStepStrategy: Calculating Pressure. FractionalStepStrategy: Updating Velocity. [WARNING] NavierStokesSolverFractionalStep: Fluid solver did not converge for step 1 corresponding to time 0.01
Fluid Dynamics Analysis: STEP: 2 Fluid Dynamics Analysis: TIME: 0.02 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 1.00033; exp.ratio = 0.001 abs = 1.07149e+13 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 12715.4; exp.ratio = 0.001 abs = 1.07114e+13 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 58.9819; exp.ratio = 0.001 abs = 8.43665e+08 FractionalStepStrategy: Fractional velocity iterations did not converge. FractionalStepStrategy: Calculating Pressure. FractionalStepStrategy: Updating Velocity. [WARNING] NavierStokesSolverFractionalStep: Fluid solver did not converge for step 2 corresponding to time 0.02 [...] Fluid Dynamics Analysis: STEP: 21 Fluid Dynamics Analysis: TIME: 0.21000000000000005 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 1.0078; exp.ratio = 0.001 abs = 4.67798e+19 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 0.705469; exp.ratio = 0.001 abs = 3.87245e+19 FractionalStepStrategy: CONVERGENCE CHECK: FractionalStepStrategy: FRAC VEL.: ratio = 1.17514; exp.ratio = 0.001 abs = 5.59391e+19 FractionalStepStrategy: Fractional velocity iterations did not converge. FractionalStepStrategy: Calculating Pressure. FractionalStepStrategy: Updating Velocity. [WARNING] NavierStokesSolverFractionalStep: Fluid solver did not converge for step 21 corresponding to time 0.21000000000000005
loumalouomega
Description When running the example beam2D or ChannelSphere2D, Kratos returns segmentation fault. beam2d_internal_interpolation, beam_spr, cavity2D, channelsphere3D and contact hessian are outdated. Remaining examples are working. General issue with the use of mmg as a process.
Scope In Examples-master/mmg_remeshing_examples/use_cases/channel_sphere2D/source/
To Reproduce Simply run the MainKratos.py in Examples-master/mmg_remeshing_examples/use_cases/channel_sphere2D/source/
Expected behavior Reproduce the result from the ChannelSphere2D example here:
Environment