[A.A.]

Hi,

I am simulating a hybrid rotor with pimpleDyMFoam and my calculations are to slow.

I am using a normal PC with four cores and Ubuntu to simulate my case with a mesh of 300k cells. I want to run the simulation for 0.2 sek., but its extremly slow. It would take me more than 300 days.

I run the same case with 80k cells in 10 days, but this seems also to be to slow.

What could be the reason for that? I heared up to 1M cells shouldn't be a problem for a single computer.

I briefly explain the case, because the problem could be related to it specificly. The case is a rotating cylinder in an airflow with five symmetric wing profiles which travel around the cylinder and change angle of attack depending on their actual position. I use AMI to do the movement. Could the complex movement be the reason for the slow calculation? (When I use the command moveDynamicMesh only the mesh movment is conducted by OF, but this goes fast within minutes).

I am using k-omega SST as turbulence model.

I would higly appriciate any hints or suggestions to solve this problem.

Many thanks,
Alex

[michielm]

are you specifying a maxCo or running with fixed timestep? And what is the value of these?

[A.A.]

Hi Michiel,

I am running with maxCo = 0.5.

As far as I know from the OF documentation, the max recommended Co number shouldn't be bigger than 1. I put it to 0.5 for safety reasons.

Best,
Alex