Fluent 6.2 + LES

Sowjanya

Hi

I am working on a simple straight pipe flow problem. I am using LES - Dynamic model to solve this problem. I have solved the problem initially using the standard k-epsilon model. I have let the solution converge using this model and then started the LES simulation as suggested in the fluent user guide by using the "initialize-instantaneous-vel" command at the TUI. The time step I am using is 0.002. The flow Reynolds number is 100,000. After letting the solution run for some 20,000~30,0000 iterations I see that the fluctuations RMS X,Y,Z components look quite right. Even the mean streamwise velocity(flow direction) looks right and is comparable to the RANS model. The only thing that does not look right is the wall shear stress. I am looking at the area weighted average value of the wall shear stress value and it is very very less when compared to the value obtained from the RANS simulation. I am not able to figure out whats happening.

Could anyone guide me to where and how this can happen? Also could anyone let me know approximately as to how many iterations it would be before we can get a decent LES solution.