Page 1 of 1

LES - wall discretization in street canyons

Posted: Tue Sep 06, 2022 10:40 am
by VDavid
Hello all,

I am working on optimizing the PSWIFT parametrization of complex street canyon geometries. For that, I am currently setting up an LES simulation for a periodic street canyon (see picture) and I was wondering about the cell sizes near the walls. My university tutor told me to aim for y+ = 2 normal to the walls and y+ = 50-100 along the flow, this being valid for attached BL only. This would result in a huge mesh for me. Since my model represents a cavity and thus a detached flow, what cell sizes should I be aiming for close to walls inside my street in order to represent the flow patterns correctly? The output information I am looking for is the macrospopic flow structure, so I guess I don't need to know that much about the BL flow. My only concern is that the main flow patterns are changed significantly due to insufficient wall resolution.

Merci d'avance!

Victor
LES_mesh.png

Re: LES - wall discretization in street canyons

Posted: Thu Sep 08, 2022 6:06 pm
by Yvan Fournier
Hello,

I'll remind turbulence specialists to answer this. As far as i! know, although we prefer wall-resolved LES when possible, there have been LES studies run with coarser meshes near the wall, due to mesh size/compute power/time constraints.
But I am not an expert in this domain, and am not sure what Y+ value is reasonable or not here.

Best regards,

Yvan