3D Backward Facing Step (Reattachment Length)

nskelly · April 23, 2018, 09:58

Hi all,

I am currently simulating flow over a backward facing step using LES. I am in the process of validating Armaly (1983) paper with air, but will later do some simulations with shear-thinning fluids.

At the moment I've been having difficulties in getting the correct reattachment length with respect to Reynolds number. In the laminar regime (up to Re=800) my results agree with the experimental data of Armaly, however beyond this my results are far from the experimental. For instance, at Re=3400 the RL is about 15 where it should be about 6.

I have a very fine mesh at the step and upper and lower walls. And I am using a mapped velocity BC at the inlet. I am using pisoFoam solver with the Smagorinsky SGS model for LES.

I am using the exact same geometry as Armaly stated in the experimental procedure.

I am not too sure why I cannot seem to get the same results. Any help or suggestions would be greatly appreciated.

piu58 · April 23, 2018, 10:28

> I have a very fine mesh at the step and upper and lower walls.

If you have higher Re numbers, than the fluid transports a lot of turbulence downstream. To cover this it is necessary to have a fine mesh in this region too. It may be that your mesh gets coarse to fast.

nskelly · April 23, 2018, 11:33

The mesh refinement is very similar downstream of the step.

I had to create multiple blocks in the blockMeshDict in order to get similar refinement without it getting too coarse.

I have a max aspect ratio of 48.

nskelly · April 24, 2018, 04:31

Also the Reynolds Number is not substantially high. Max will be Re=7000

April 23, 2018, 09:58	3D Backward Facing Step (Reattachment Length)	#1
nskelly Member Nat K Join Date: Oct 2017 Posts: 68 Rep Power: 9	Hi all, I am currently simulating flow over a backward facing step using LES. I am in the process of validating Armaly (1983) paper with air, but will later do some simulations with shear-thinning fluids. At the moment I've been having difficulties in getting the correct reattachment length with respect to Reynolds number. In the laminar regime (up to Re=800) my results agree with the experimental data of Armaly, however beyond this my results are far from the experimental. For instance, at Re=3400 the RL is about 15 where it should be about 6. I have a very fine mesh at the step and upper and lower walls. And I am using a mapped velocity BC at the inlet. I am using pisoFoam solver with the Smagorinsky SGS model for LES. I am using the exact same geometry as Armaly stated in the experimental procedure. I am not too sure why I cannot seem to get the same results. Any help or suggestions would be greatly appreciated.

April 23, 2018, 10:28		#2
piu58 Senior Member Uwe Pilz Join Date: Feb 2017 Location: Leipzig, Germany Posts: 744 Rep Power: 15	> I have a very fine mesh at the step and upper and lower walls. If you have higher Re numbers, than the fluid transports a lot of turbulence downstream. To cover this it is necessary to have a fine mesh in this region too. It may be that your mesh gets coarse to fast. __________________ Uwe Pilz -- Die der Hauptbewegung überlagerte Schwankungsbewegung ist in ihren Einzelheiten so hoffnungslos kompliziert, daß ihre theoretische Berechnung aussichtslos erscheint. (Hermann Schlichting, 1950)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Long simulation time (Backward Facing Step)	nskelly	OpenFOAM Running, Solving & CFD	3	March 6, 2018 11:59
backward facing step problem	ian8430	Main CFD Forum	1	January 18, 2011 09:58
Backward FAcing Step - Frequency	CFDtoy	Main CFD Forum	8	October 29, 2006 21:51
backward facing step	ryoga	Main CFD Forum	1	March 9, 2003 03:52
Backward Facing Step	Pete	Main CFD Forum	5	June 28, 2000 22:40

April 23, 2018, 11:33		#3
nskelly Member Nat K Join Date: Oct 2017 Posts: 68 Rep Power: 9	The mesh refinement is very similar downstream of the step. I had to create multiple blocks in the blockMeshDict in order to get similar refinement without it getting too coarse. I have a max aspect ratio of 48.

April 24, 2018, 04:31		#4
nskelly Member Nat K Join Date: Oct 2017 Posts: 68 Rep Power: 9	Also the Reynolds Number is not substantially high. Max will be Re=7000