Defining regions of laminar and turbulent flows

raven · July 30, 2015, 16:34

Hello,

I'm trying to generate airfoil data at low Reynolds numbers (~300,000). Compared to experimental results, the lift coefficients are very good whereas the drag coefficient are over predicted by about 30 to 50% when fully turbulent SST model is used.

At low Reynolds number such as this, I understand that the flow transitions from laminar to turbulent somewhere on the airfoil and shed laminar separation bubbles (LSB). My guess is that the assumption of fully turbulent flow near the leading edge of the airfoil (where the flow is actually laminar) is responsible for the over prediction.

To investigate this I've tried using the gamma model and the gamma theta model with built in default settings. The results did improve slightly but the drag is still grossly over predicted even at moderate angles of attack.

I keep reading on this forum that using the specified intermittency setting, one can define where the flow is turbulent or laminar (say flow becomes turbulent at 25% chord length). But I cannot find anywhere how to implement this. Does anyone have experience with this and can advise me on this matter?

Thanks.

Opaque · July 30, 2015, 16:42

Here is the documentation

Quote:

CFX-Solver Modeling Guide
Chapter 4: Turbulence and Near-Wall Modeling | 4.1. Turbulence Models |

As well as the two-equation Gamma Theta transition model, two reduced models are available:

Specified Intermittency transition model

A zero-equation model, where you can prescribe the intermittency directly as a CEL expression.

The best way to specify the intermittency is with a user defined subroutine that is based on the x, y and z coordinates. This way, conditional statements can be used to define geometric bounds where the intermittency can be specified as zero (laminar flow) or one (turbulent flow). This method can be used to prescribe laminar regions at the leading edges of the wings, for example.

Gamma transition model

A one-equation model that solves only the intermittency equation using a user specified value of the transition onset momentum thickness-based Reynolds number.

raven · July 30, 2015, 16:43

yes, I saw this as well. I'm looking for more detailed explanation or an example...

Thanks

ghorrocks · July 30, 2015, 19:42

It has been a while since I looked at this, but I recall you simply set intermittency=0 where it is laminar and 1 where it is turbulent. Then you can define it directly.

raven · July 30, 2015, 21:16

So should I physically divide the mesh into two regions?

ghorrocks · July 30, 2015, 21:18

No. Define a CEL expression (either using expressions or a 3D interpolation function) which returns 0 in laminar regions and 1 in turbulent regions. Then set the intermittency to the expression.

raven · July 30, 2015, 21:21

Hi ghorrocks,

I'm not familiar with CEL expressions. Could you give me a quick example of what it might look like or direct me to a useful source?

ghorrocks · July 31, 2015, 01:24

Have a look at the CFX tutorials, in the help menu.

July 30, 2015, 16:34	Defining regions of laminar and turbulent flows	#1
raven New Member Dustin Join Date: May 2014 Posts: 11 Rep Power: 12	Hello, I'm trying to generate airfoil data at low Reynolds numbers (~300,000). Compared to experimental results, the lift coefficients are very good whereas the drag coefficient are over predicted by about 30 to 50% when fully turbulent SST model is used. At low Reynolds number such as this, I understand that the flow transitions from laminar to turbulent somewhere on the airfoil and shed laminar separation bubbles (LSB). My guess is that the assumption of fully turbulent flow near the leading edge of the airfoil (where the flow is actually laminar) is responsible for the over prediction. To investigate this I've tried using the gamma model and the gamma theta model with built in default settings. The results did improve slightly but the drag is still grossly over predicted even at moderate angles of attack. I keep reading on this forum that using the specified intermittency setting, one can define where the flow is turbulent or laminar (say flow becomes turbulent at 25% chord length). But I cannot find anywhere how to implement this. Does anyone have experience with this and can advise me on this matter? Thanks.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFX Treatment of Laminar and Turbulent Flows	Jade M	CFX	18	September 15, 2022 08:08
Ratio of eddy viscosity to molecular viscosity : Laminar or turbulent flow?	JuPa	CFX	7	September 9, 2013 08:45
laminar and turbulent flow in two regions	mozafarie	FLUENT	0	March 11, 2013 14:42
Defining wall Roughness in laminar flows	SH_P	FLUENT	1	April 28, 2012 17:34
SimpleFoam: Laminar vs. Turbulent Convergence	JasonG	OpenFOAM	0	June 2, 2011 09:29

July 30, 2015, 16:43		#3
raven New Member Dustin Join Date: May 2014 Posts: 11 Rep Power: 12	yes, I saw this as well. I'm looking for more detailed explanation or an example... Thanks

July 30, 2015, 19:42		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	It has been a while since I looked at this, but I recall you simply set intermittency=0 where it is laminar and 1 where it is turbulent. Then you can define it directly.

July 30, 2015, 21:16		#5
raven New Member Dustin Join Date: May 2014 Posts: 11 Rep Power: 12	So should I physically divide the mesh into two regions?

July 30, 2015, 21:18		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	No. Define a CEL expression (either using expressions or a 3D interpolation function) which returns 0 in laminar regions and 1 in turbulent regions. Then set the intermittency to the expression.

July 30, 2015, 21:21		#7
raven New Member Dustin Join Date: May 2014 Posts: 11 Rep Power: 12	Hi ghorrocks, I'm not familiar with CEL expressions. Could you give me a quick example of what it might look like or direct me to a useful source?

July 31, 2015, 01:24		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,872 Rep Power: 144	Have a look at the CFX tutorials, in the help menu.