[niamhiefitz]

Basically I'm doing a 2D simulation of a vertical axis turbine, its in simple terms its just 3 aerofoils rotating around an axis. I have successfully carried out sliding mesh simulations with my current mesh, no convergence issues or weird results, for constant rotational speed. Now I'm trying to carry out dynamic mesh simulations to see the start-up of the turbine with a given inlet velocity. However, no matter how small I make the time step (even to 1e-5) my solution starts diverging a lot. Like values of Cm and Cd in 100,000s.

I've set the aerofoils as rigid bodies as well as the rotating domain (which has been set to passive). Would these be the right settings?

I'm using SIMPLE and k-omega SST, all 2nd order. I'm wondering if it is the mesh, but it worked fine for sliding mesh. Although my skewness was quite high, around 0.91.

Any suggestions on what to do? I've tried changing the relaxation factors and that didn't work.

Edit: I've also been monitoring the courant number but I don't think it's that. No matter how small it is divergence will still happen and make the courant number increase with it.