transient timestep
 

Hi, all,

I am doing a transient simulation of a flow coming out of a small orifice into a mixing tube. I have a question about the timestep value.

Due to the 0.5 mm diameter of the orifice, I have very fine mesh at the orifice region. The velocity out of orifice is about 40 m/s. So my timestep has to be very small to keep the solver converging within 10 iterations. But this small timestep(1e-4) is too small that it takes forever to simulate total 5 seconds. Everytime I incrase timestep, it will not converge.

Does anyone has similar experience? How can I speedup the simulation? The mesh has about 300,000 elements. Most of them are tet mesh.

Thanks. Tao

Re: transient timestep
 

Hi,

Welcome to CFD. Long simulation times are common, indeed inevitable for any serious analysis.

How to speedup? Go parallel and faster CPUs are the common approaches.

Glenn Horrocks