Learning Forum

[AgriCFD]

Hi,

currently I do a transient calculation (k-epsilon) on a fast rotating (3000rpm) mower unit. I only simulate one rotor.

 

The Problem:

Simulation is running for a few time steps (approx. 5-30, depending on the initial flow field). The residuals are high but very evenly. Then Epsilon residual starts vibrating. Then, suddenly I get a floating Point exeption, with the residuals always being below 1e-01.

Console Output:
Divergence detected in AMG solver: epsilon
Error at host: floating point exception
Error at Node 0: floating point exception

I found an area of extremly high turbulence variables (turbulent intensity: >1500%). I think this is the problematic area, but I dont know what to do. Can someone help me out with this Problem.

I tried a lot of Things:

- created a better mesh ()

- enlarged the fluid domain around the mower (with smaller Domain I had backflow and high turbulent viscosity ratio)

- reduced the relaxation factors

 

Possibly interesting: I did the same calculation with a rotational Speed of 300rpm and the Simulation is running without Problems. But even in this case the turbulence in the problematic areas is extreme (turb. intensity > 200%). Is there a possible reason for this?

 

[DrAmine]

Better mesh with low aspect ration, lower skewness and low cell volume change. I won't touch anything else before ameliorating the grid in the areas where you think that is related to that. Just have a look into the areas where you high TVR, locate them and check the values of the mesh metrics.