Mechanisms of turbulence production

wilson · December 8, 2009, 18:06

I'm doing tutorial 9 in cfx and I am trying to answer a question about the main mechanisms of turbulence production in the different principal regions of the flow, can anyone tell me what this means please? The mechanisms are described from plots contours of eddy viscosity.

Thanks

andy_ · December 9, 2009, 11:05

Quote:

Originally Posted by wilson

I'm doing tutorial 9 in cfx and I am trying to answer a question about the main mechanisms of turbulence production in the different principal regions of the flow, can anyone tell me what this means please? The mechanisms are described from plots contours of eddy viscosity.

Thanks

Can you link to the tutorial?

wilson · December 9, 2009, 12:15

I've attached the eddy viscosity plot and I am wondering what the mechanisms of turbulence production are?

Thanks

andy_ · December 9, 2009, 12:53

It was not just the picture but the context. Is this online or only on your computer as part of your cfx installation?

The answer you can always give is flow instability regardless of the pictures but I doubt that is what is being asked for hence the request for context.

wilson · December 9, 2009, 18:23

The tutorial looks at the flow through a pipe containing a butterfly valve and I'm being asked what the mechanisms of turbulence production are. I've spoke to one person who says its due to the deep velocity gradients but I'm looking for a more detailed explanation than that

sfalsharif · December 9, 2009, 18:56

Turbulence is produced by the interaction of mean velocity gradients (shearing and staining of the flow field) with the fluctuating velocity field. Look for regions of high shear or normal strain, these will be the regions were turbulence is produced.

wilson · December 9, 2009, 19:19

Thanks.

How will I be able to determine the regions of high shear or normal strain?

sfalsharif · December 9, 2009, 19:42

There should be a way in the software to look at contours of the velocity gradient tensor components (individually). I know you can do it in Fluent, surely it's possible with CFX as well.

andy_ · December 10, 2009, 05:15

The production of turbulence is the rate of work done by the mean flow against the Reynolds stresses. If you consider the face of a control volume then the rate of work done is the product of the stress component and the velocity component in the direction of the stress component (i.e. force times distance moved in direction of force from school physics lessons). If you add up the various contributions on all the faces you will determine the net contribution to turbulence generation. However, note that opposite faces will cancel if you do not have a gradient in the velocity and/or stress components which is where the gradients come in.

In high regions of turbulence generation you typically have a large stress component and a large gradient in the velocity component in the direction of that stress component. In your example this is the free shear layer and boundary layers. The generation of turbulence tends to increase the Reynolds stress but decrease the velocity gradient so the level of generation grows and then reduces as the shear layer, for example, mixes out in your example.

This type of view of fluid mechanics combined with the details produced by CFD simulations of the stress and rates of strain components in the flow allow engineering decisions to be made in a systematic and reliable manner and I suspect your tutorial question is after something like the above. But it does not address the question/mechanism of why the flow is unstable and the turbulent stresses arise which you opening question appeared to be about. It would have been interesting if it was.

wilson · December 10, 2009, 17:37

Thanks very much Andy for taking the time to explain. I take it you looked at the eddy viscosity plot I attached in an earlier post?

December 8, 2009, 18:06	Mechanisms of turbulence production	#1
wilson New Member Mike Join Date: Dec 2009 Posts: 14 Rep Power: 16	I'm doing tutorial 9 in cfx and I am trying to answer a question about the main mechanisms of turbulence production in the different principal regions of the flow, can anyone tell me what this means please? The mechanisms are described from plots contours of eddy viscosity. Thanks

December 9, 2009, 18:56	Turbulence production	#6
sfalsharif New Member Join Date: Mar 2009 Posts: 15 Rep Power: 17	Turbulence is produced by the interaction of mean velocity gradients (shearing and staining of the flow field) with the fluctuating velocity field. Look for regions of high shear or normal strain, these will be the regions were turbulence is produced.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Discussion: Reason of Turbulence!!	Wen Long	Main CFD Forum	3	May 15, 2009 10:52
PRoduction terms of a turbulence modeL	nada	FLUENT	1	December 30, 2008 23:40
Turbulence Production Based on the Curl of the Velocity Field	ngj	OpenFOAM Running, Solving & CFD	0	April 4, 2008 06:22
non-physical production of k, k-w turbulence model	David Garrido	Main CFD Forum	9	September 17, 1999 16:47

December 9, 2009, 12:53		#4
andy_ Senior Member andy Join Date: May 2009 Posts: 301 Rep Power: 18	It was not just the picture but the context. Is this online or only on your computer as part of your cfx installation? The answer you can always give is flow instability regardless of the pictures but I doubt that is what is being asked for hence the request for context.

December 9, 2009, 18:23		#5
wilson New Member Mike Join Date: Dec 2009 Posts: 14 Rep Power: 16	The tutorial looks at the flow through a pipe containing a butterfly valve and I'm being asked what the mechanisms of turbulence production are. I've spoke to one person who says its due to the deep velocity gradients but I'm looking for a more detailed explanation than that

December 9, 2009, 19:19		#7
wilson New Member Mike Join Date: Dec 2009 Posts: 14 Rep Power: 16	Thanks. How will I be able to determine the regions of high shear or normal strain?

December 9, 2009, 19:42		#8
sfalsharif New Member Join Date: Mar 2009 Posts: 15 Rep Power: 17	There should be a way in the software to look at contours of the velocity gradient tensor components (individually). I know you can do it in Fluent, surely it's possible with CFX as well.

December 10, 2009, 05:15		#9
andy_ Senior Member andy Join Date: May 2009 Posts: 301 Rep Power: 18	The production of turbulence is the rate of work done by the mean flow against the Reynolds stresses. If you consider the face of a control volume then the rate of work done is the product of the stress component and the velocity component in the direction of the stress component (i.e. force times distance moved in direction of force from school physics lessons). If you add up the various contributions on all the faces you will determine the net contribution to turbulence generation. However, note that opposite faces will cancel if you do not have a gradient in the velocity and/or stress components which is where the gradients come in. In high regions of turbulence generation you typically have a large stress component and a large gradient in the velocity component in the direction of that stress component. In your example this is the free shear layer and boundary layers. The generation of turbulence tends to increase the Reynolds stress but decrease the velocity gradient so the level of generation grows and then reduces as the shear layer, for example, mixes out in your example. This type of view of fluid mechanics combined with the details produced by CFD simulations of the stress and rates of strain components in the flow allow engineering decisions to be made in a systematic and reliable manner and I suspect your tutorial question is after something like the above. But it does not address the question/mechanism of why the flow is unstable and the turbulent stresses arise which you opening question appeared to be about. It would have been interesting if it was.

December 10, 2009, 17:37		#10
wilson New Member Mike Join Date: Dec 2009 Posts: 14 Rep Power: 16	Thanks very much Andy for taking the time to explain. I take it you looked at the eddy viscosity plot I attached in an earlier post?