[Richy]

I am working on modelling a high speed impeller. I have done the Tutorial of the fan (both the reference frame and rigid body motion). I am having trouble running the unsteady simulation. I follow the steps of the tutorial (obviously using parameters appropriate to my model) but I keep getting a floating point error?

Any suggestions? I think it might be mesh related? Do i need a special mesh for the motion to work correctly?

Thanks

[malfrienz]

Hi has your problem been resolved? I have been getting the same error and i tried using very low residuals as mentioned in multiphase (my simulation has multiphase models as well). And also perhaps you could try first with first order and then once settled, change it to 2nd order.

Let me know how you solved this problem.

thanks.

[frownless]

When I did the second part of that tutorial where you move from steady to unsteady I found that it had frozen the solver. I needed to untick the box in the solver settings to get it to run properly.

Hope this helps!

[Richy]

Thanks for the help. I changed the inner iteration number and that seems to have solved the issue. The only concern I have now is the residual plot. During each time step, my residuals decrease sufficiently but as a new time step commences, they shoot back up. To me, this suggests that within each time step I am reaching some level of convergence. I know I need to look at other parameters to judge convergence, but I was just wondering about this?

[olauer]

That's the way, RBM works. Everytime you move your mesh you increase the residuals in the interface regions.
For convergence checks you could have a look on pressure gain or torque or ...

[Richy]

Thanks. That makes sense actually