Learning Forum

[mimozas1]

Hallo everybody!

I am conducting a Detached Eddy Simulation of a 3D wing.

My main focus for now is capturing cd-cl fluctuations once the wake is being set to periodical motion.

For that reason i have encountered some thought that troubled me. I know since near the wall-boundary the simulation is like RANS, i need y+<=1 for k-omega model. But fot the LES part of the grid i am having trouble defining exactly the size of the grid. I have made a lit.review but found no solid and universal criteria for grid size. The idea i am having is defing that through experimenting-changing-re experimenting.

If anyone has any knowledge on the field of the grid size for the LES part of the DES grid i would be very glad and thankfull to share it.

Again thanks for your time people!

[Rob]

Check the courses (top of the page) as at least one covers turbulence and LES. From memory, when the LES model first came out we used to run k-e (would be k-w in your case) to get a flow field and an idea of how turbulent the system was. There may be additional guidance in the Fluent documentation and references quoted there.

[Murari Iyengar]

Hi mimozas1 LES requires a finer mesh than those typically used for RANS. In the DES model, based on the turbulent length scaleand grid spacing, RANS/LES model is chosen. For more information, please see the links below.
4.16.1. Overview (ansys.com)
10.1. Estimation of the LES Resolution Quality (ansys.com)
13.2.2. Scale-Resolving Simulation (SRS) Models (ansys.com)