CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Specfying Turbulent boundary-backward-facing step (

Ben Sellars December 4, 2006 15:56

Specfying Turbulent boundary-backward-facing step
Can anyone send me the fluent tutorial for Turbulent flow over a backward facing step in the Fluent website( i dont have have any access to their website. Or does anyone kwown how to specfy a turblent boundary in fluent, with referance to this problem:

Problem Description: Backward facing step flows involve separation, free shear layers, recirculation region, and flow reattachment. It has important industrial applications, such as in oil/gas pipes and flow around buildings. This case provides a good test to the capability of CFD software regarding the mathematical models (turbulence models), the numerical algorithms, and the mesh sensitivity. After verification of your solution approach, validation can be carried out by comparing your results with the experimental data available by Driver and Seegmiller, which are provided at the module's webCT website.

Figures 1 shows a schematic diagrams of the backward facing step of height H = 1.27cm. The experiment was carried out in a wind tunnel with the step on the floor and the top boundary is the tunnel wall. The tunnel span is 12H.

(figure won't copy and paste) inlet height = 8 H outlet height = 9 H depth/ width/ spanwise = 12H where H= Height of the step

Figure 1.

Inlet Flow Condition: Uin = 44 2 m s-1, Min = 0.128, at atmospheric total pressure and temperature.

The turbulent boundary layer thickness at 4H upstream of the step:  = 1.9 cm

Note that the boundary layer thickness is significant in comparison with the step height. It is therefore important to match this condition as closely as possible. This can be done by (1) extending the upstream channel length; or (2) specifying a turbulent boundary layer distribution at x = -4H.

Only the undeflected upper wall case is considered in this assignment.

P.S: I have look at the paper "Compasrison of near-wall treatment methods for high Reynolds number backward-facing step flow" J.Y kim, A.J.Ghajar, C.Tang, G.L.Foutch

And hence hearding in the right general direction just need to knwon how to: "specifying a turbulent boundary layer distribution at x = -4H." Thanks to all, any help would be azaming

RoM December 13, 2006 08:39

Re: Specfying Turbulent boundary-backward-facing s
The velocity profile as a function of the y coordinate can be written as

u(y)=ufree*(y/delta)^n with delta=inlet_height/2

The boundary layer thickness is the distance from the wall where u(y) is 99% of the free stream velocity.

With u(y)/ufree=0.99 , delta=5.08cm and y=1.9 cm you can calculate n to get the velocity profile function. To use this function as a bc in Fluent you will need an udf or write an profile file (see manual chapter 7.26).

To get the k and epsilon profiles construct a new channel in gambit. Use the inlet dimension of your backward facing step problem and a channel length of at least 100H. Use the velocity profile from above as input value and let Fluent calculate the k and e profiles at the outlet. Extract the profiles for k,e at the outlet (see 7.26) and use it as bc for your backward step.

Hope it helps, RoM

All times are GMT -4. The time now is 16:38.