About convergence
 

I tried to run the same case file in single and double precision solver. The convergence limit is 1E-5. The double precision converges while the single one does not. Why this is so?

Re: About convergence
 

Numeric diffussion.

Re: About convergence
 

Then how can I avoid this? Thanks.

Re: About convergence
 

Using the double precision solver obviously works. If you are using second order upwind try switching to first order, which is less susceptible to numeric diffusion.

Re: About convergence
 

I'm using QUICK scheme now. I will try 1st order upwind. Thank you!