[oe15d033]

Hi all,
I am studying the flow past 3D fixed cylinder in range of Reynolds number from 1000-100000. The physics is segregated flow-implicit unsteady - RANS - k omega SST. I have chosen the time step as 0.001s. The results are good and comparable, but simulation time is high. So i want to know, Is there any rule to calculate the time step for this problem like courant number, CFL?


1.How can i chose the time step for this problem?

2.If courant number is the criteria, What is the courant number for the flow past cylinder problem?
3. If I change the turbulence model, cournat number will change?





Thanks in advance.

[ashokac7]

Originally Posted by oe15d033

Hi all,
I am studying the flow past 3D fixed cylinder in range of Reynolds number from 1000-100000. The physics is segregated flow-implicit unsteady - RANS - k omega SST. I have chosen the time step as 0.001s. The results are good and comparable, but simulation time is high. So i want to know, Is there any rule to calculate the time step for this problem like courant number, CFL?


1.How can i chose the time step for this problem?

2.If courant number is the criteria, What is the courant number for the flow past cylinder problem?
3. If I change the turbulence model, cournat number will change?





Thanks in advance.

Courant number is "u∆t/∆x".

Back calculate your time step "∆t" taking Courant number as 1. 1 is ideal. u is free-steam velocity and ∆x is first layer height. Your Courant number may give recommended time step value less than 1ms as you have used. Then the only option is to do trial and error to find out optimum time step for your case.

Courant number doesn't change with turbulence model. Just refer the formula given above.
Hope this helps.

[oe15d033]

Thanks Ashok..