No periodic flow structure while one is expected

ziggo · September 4, 2013, 07:06

I am running a transient 3D simulation on the geometry shown in the attachment.
I use the k-epsilon standard model.

I have done successful grid refinement study and the aerodynamic coefficients are close to the wind tunnel reference data.
However, no matter what time step I use (tried 0.01, 0.004, 0.0005s) I don't get a periodic wake structure with shed vortices.
The flow is as shown in the attachment for the full running time (except for the beginning where the flow is not fully established yet).

I've varied my Reynolds number (tried 500000 and 50000) but this did not make any difference. I kept getting a steady state solution with no flow oscillations.
I have the feeling that some artificial damping is preventing the periodic shedding.

What can be the problem?
If you need more information I am happy to provide it.

duri · September 4, 2013, 08:52

Quote:

Originally Posted by ziggo

However, no matter what time step I use (tried 0.01, 0.004, 0.0005s) I don't get a periodic wake structure with shed vortices.
The flow is as shown in the attachment for the full running time (except for the beginning where the flow is not fully established yet).

I've varied my Reynolds number (tried 500000 and 50000) but this did not make any difference. I kept getting a steady state solution with no flow oscillations.

Usually wake from turbulent flows are not periodic when shedding from stationary structures. How do you know wake is periodic. If so, at what reynolds number and what is the periodicity of wake.

ziggo · September 4, 2013, 09:33

I was not aware of the fact that turbulent flows are a-periodic. Have you got references about this?

I found the graph shown in the attachment in the book of Blevins called Flow Induced Vibration. This graphs shows that a 2D wedge has a Strouhal number of around 0.17 for Re=1e4.

I figured that Re=5e4 is in the same order of magnitude and expected some vortex shedding.

duri · September 4, 2013, 11:54

From my understanding, with increase in reynolds number in turbulence regime vortex shedding will lose its periodicity. Periodic vortex occurs only in some range of reynolds number that ends in low turbulence flow.
Probably you need to figure out the reynolds number at which periodic vortex occurs for your configuration. Comparing 2d flow and 3d flow doesn't make much sense.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Periodic Pipe Flow LES; parabolic profile	jrrelx	FLUENT	34	August 13, 2015 18:58
Periodic boundary conditions for 2D pipe (water flow).	thess	FLUENT	3	May 31, 2012 14:38
Periodic flow	librasu	FLUENT	1	August 18, 2008 13:24
Instability and Vortex/roll structure	Z. Zeng	Main CFD Forum	3	May 10, 2000 06:18
Possible?: Periodic conditions with non-constant mass flow rate.	Ray	FLUENT	0	April 10, 2000 07:10

September 4, 2013, 09:33		#3
ziggo Member Bart A. Join Date: Feb 2012 Posts: 45 Rep Power: 0	I was not aware of the fact that turbulent flows are a-periodic. Have you got references about this? I found the graph shown in the attachment in the book of Blevins called Flow Induced Vibration. This graphs shows that a 2D wedge has a Strouhal number of around 0.17 for Re=1e4. I figured that Re=5e4 is in the same order of magnitude and expected some vortex shedding.

September 4, 2013, 11:54		#4
duri Senior Member duri Join Date: May 2010 Posts: 245 Rep Power: 17	From my understanding, with increase in reynolds number in turbulence regime vortex shedding will lose its periodicity. Periodic vortex occurs only in some range of reynolds number that ends in low turbulence flow. Probably you need to figure out the reynolds number at which periodic vortex occurs for your configuration. Comparing 2d flow and 3d flow doesn't make much sense.