What is eddy viscosity value in advanced turbulence parameters?

CFD-fellow · August 27, 2014, 06:15

Hi
Im modeling a centrifugal pump with sst turbulence model. In high flow rates the pump produces much more head than experiment one and in low flow rates produces less head. In other words the H-Q curve of experiment is steeper than the numerical one. But when I set the eddy viscosity value to 1e-5 in advanced turbulence parameters, I get really well data.
What happens if I set a constant value to this parameter?
What is the reasonable value for it?
Is it logical to change this value to get proper results?

ghorrocks · August 27, 2014, 07:28

Which parameter are you setting to 1e-5? Please post your CCL or output file.

CFD-fellow · August 27, 2014, 07:45

Hi
The parameter is attached in the picture.

ghorrocks · August 27, 2014, 08:31

I am not familiar with that parameter. It is a bit puzzling, it implies that it results in a constant eddy viscosity - so then why have a turbulence model?

You are going to have to look this up in the documentation, or possibly contact CFX support. Or maybe somebody on the forum is cleverer with CFX turbulence models than me.

Lance · August 27, 2014, 08:59

I got curious and looked it up in the documentation:
"Parameter/Description
Eddy Viscosity / Enables you to specify a value that will be picked up by the specified turbulence model; in this case the internal algorithm will not be used."

But I dont understand why anyone would specify a value for the eddy viscosity?

CFD-fellow · August 27, 2014, 09:29

I thought that this parameter fix the average value of eddy viscosity in the domain, but I got contours of eddy viscosity and everywhere has the same value 1e-5. So Glenn is right and using turbulence model is meaningless.
Also this sentence of documentation gives good ideas "You should use this option when startup with the desired turbulence model is not possible." and also you can define an expression other than what cfx uses for eddy viscosity and this is really good.

But still, I dont know why my H-Q curve is like attached picture?

ghorrocks · August 27, 2014, 10:10

Please put labels on future graphs so we do not have to guess what you are showing.

It looks like at low flow/high head you are producing less head than anticipated. Have you checked whether the model is predicting excessive separation at low flow rates? This could be linked to your comment about the turbulence values.... if the constant 1e-5 is more accurate than the SST model then it suggests something with wrong with your turbulence modelling - maybe your inlet turbulence level or surface roughness is wrong.

August 27, 2014, 06:15	What is eddy viscosity value in advanced turbulence parameters?	#1
CFD-fellow Senior Member Behrooz Jamshidi Join Date: Apr 2013 Posts: 110 Rep Power: 13	Hi Im modeling a centrifugal pump with sst turbulence model. In high flow rates the pump produces much more head than experiment one and in low flow rates produces less head. In other words the H-Q curve of experiment is steeper than the numerical one. But when I set the eddy viscosity value to 1e-5 in advanced turbulence parameters, I get really well data. What happens if I set a constant value to this parameter? What is the reasonable value for it? Is it logical to change this value to get proper results?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem with divergence	TDK	FLUENT	13	December 14, 2018 07:00
Help with calculating turbulence parameters!	Dan Chambers	Main CFD Forum	0	January 16, 2011 16:02
Link between eddy viscosity and subgrid-scale turbulent viscosity in LES	Marvin	Main CFD Forum	6	December 4, 2009 12:08
Turbulence inlet boundary conditions - eddy viscosity approximation	zboud	CFX	3	June 1, 2009 19:42
Turbulence viscosity ratio	Haitham	FLUENT	5	February 5, 2008 05:49

August 27, 2014, 07:28		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Which parameter are you setting to 1e-5? Please post your CCL or output file.

August 27, 2014, 08:31		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I am not familiar with that parameter. It is a bit puzzling, it implies that it results in a constant eddy viscosity - so then why have a turbulence model? You are going to have to look this up in the documentation, or possibly contact CFX support. Or maybe somebody on the forum is cleverer with CFX turbulence models than me.

August 27, 2014, 08:59		#5
Lance Senior Member Lance Join Date: Mar 2009 Posts: 669 Rep Power: 22	I got curious and looked it up in the documentation: "Parameter/Description Eddy Viscosity / Enables you to specify a value that will be picked up by the specified turbulence model; in this case the internal algorithm will not be used." But I dont understand why anyone would specify a value for the eddy viscosity?

August 27, 2014, 10:10		#7
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Please put labels on future graphs so we do not have to guess what you are showing. It looks like at low flow/high head you are producing less head than anticipated. Have you checked whether the model is predicting excessive separation at low flow rates? This could be linked to your comment about the turbulence values.... if the constant 1e-5 is more accurate than the SST model then it suggests something with wrong with your turbulence modelling - maybe your inlet turbulence level or surface roughness is wrong.