Hi all:

I study the dual time stepping recently and add it to

LES code. But I still confuse in subiteration. Before

each physical-time-step, I have to make the subiteration

convergence in pesudo-time and then use the result to do

tempral iteration. Am I right?

Hi,

Yes, you seem right.

The point is more that if you use LES, you may need small time steps to achieve a temporal resolution high enough to capture "high" frequencies. The DTS method is more used to compute flows with mainly low frequencies, its main advantage being to allow the use of large time steps ... That is to say that if you use DTS with very small timesteps to capture more physics, you should verify that the cost of the DTS subiterations is not higher than the one for a classical explicit approach.

Regards

Buch