I would like to study decay of vortices structures behind a vortex generator at Re_bulk=5000 in a plain channel, isn't LES a bit of an overkill there ?

hi, i also work on LES. But what exactly, your problem is? Can you tell?

I would like to quantify the viscous dissipation of the structures behind the winglet and link it with the heat transfer. I would like to compare stationnary RANS and non stationnary calculations.

which type of winglet u are using...delta wing or winglet pair....Which paper you would like to compare , related to RANS?

If I understant correctly, your Re_bulk is not very high. Can you not attempt a DNS? If DNS resolution requirements are too high, I think attempting LES is quite reasonable. If you could explain why you think LES for this is an overkill, I might be able to explain better.

I want mainly to compare with "Heat transfer in a rectangular channel with a vee-shaped vortex generator" by Sohankar. However, I have a single delta winglet while he have two. He uses dns and LES, and found the results comparables.

I thought LES was interesting for high Reynolds numbers (more than 10000), but for less, RANS and URANS might be sufficient.

Yes that is correct. So, with the present computational power, computing DNS for your case should not be much effort, don't you think?

hi, For the winglet problem....Do, anyone need to satisfy Y+=1 around the winglet body also? as near the wall it must be a criterion for les? Can anyone tell what is the reason?