[Edemar]

Hello,

I am simulating the temperature field in a satellite. The simulation is transient, but all the boundary conditions are periodic. I wish to let the simulation running for some cycles until the periodic result is obtained. The criteria to stop is 20 full cycles. However, the simulation stops in the middle of the fifth cycle. I have simulated two cases with different orbital parameters and both have stopped in the same iteration number 5502, in the middle of the fifth cycle.

The error is #004100018 has occurred in subroutine FINMES. Fatal overflow in linear solver.

I checked the residuals and imbalance. Both are very low. I checked all the expressions in the boundary conditions and they are correct. I have tried two refinement of meshes but it didnt help. I have used two timestep,1s and 5s, but didnt work either. In fact, doesnt matter the size of mesh or timestep, the simulation crashes always at the same physical time.

I dont understand the problem because in the first 4 cycles the simulations runs properly. On the other hand, some cases with different spins (a parameter of the satellite) didnt crashed.