Hi All,

I have been looking at some papers about Backward facing step problems. In all of these papers they refer to the step height as h and the free stream (inlet) velocity as Uo. They then give a Reynolds number of, say, 25,000 based on these quantities. However, it is rare that someone says "we used a step height of x m" . Why is this? I assume it is simply because a choice of h defines a choice of Uo through the Reynolds number and this is non-dimensional?

They often gien h in terms of the boundary layer thickness at the end of the step, why is this?

Also, in the spanwise direction people tend to use 32 nodes. Why is this? The axial and vertical directions are always well refined. I asusme it is because the flow is quasi-2D?

Also, does anyone have any guide lines for defining a mesh spacing size (and expansion ratio) along the shear layer (and perpendicular to it) at the end of the step in an LES?

Thanks for any help,

J.

Surely someone knows the answer to some of these?! Please!

Hi there

How did you generate 2D structured grids for this problem (Backward Facing step) I am looking for some grid generators. I request your kind help in this regard.

Thanks

CFDtoy

I did it myself. I simply created the grid points in an array. I am however, looking for a poisson grid smoother if anyone knows where I can get one.

Thanks.

I'm not an expert, but maybe i can help you a little bit why don't you copy the spacing from le, kim & moin (stanford) or some other work? they are all using les bfs is not 2d from Re=1000 to 7000, but it's always simetric (barkley,gomes,henderson) the boundary layer thickness tells you if the flux is turbulent before the step, it influences the entrainment of fluid in the shear layer and if you use adimensionalization with the boundary layer, you can generalize the results with other configurations ciao

Hello Fluids

Can you send me a structured grid for backward facing step ? I wanna check my code on a structured grid. my email is CFDtoy@hotmail.com

Thanks

CFDtoy

[mahfuzsarwar]

hi!
I am recently working on simulation of backward-facing step. For validation, I will need the DNS data of Le & Moin - 'Direct numerical simulation of turbulent flow over a backward-facing step'. Please let me know if anyone can help me. mahfuzsarwar@yahoo.com

[Engr.RZA]

Why u not using the mesh independence study?

Start with the initial mesh size mentioned in the literature , make it course and fine. Check the results for the mesh independence and use it for your detailed study.