Blending funtction F1 for SST turbublence model

wooing · September 1, 2009, 13:35

I am confused about the blending function for SST model:
in Menter's paper, he said, F1 switch to one inside of boundary layer? But in my case i found in the free surface, I got F1=one also.

Is that ok? or something wrong?

Thanks in advance!

SKK · September 2, 2009, 05:48

F1 is formulated so that it is 1 in boundary layer and in regions of adverse pressure gradient. This switches the k-w part of the model in these situations. In the freestream it should be one. Check your formulation and the variables of F1. I would suggest have a look at the value of vorticity in the problematic area to see if it is behaving the way it should.

wooing · September 2, 2009, 13:50

Thanks for your reply!
I checked the formula, it was right and also the wall distance calculated rightly. For my cases I got a good convergence with regular k-w, but if i use sst k-w it did not convergence very well and F1 is one in some other region besides boundary layer.
Last day i tired to set F1=one and F2=zero everywhere and switch the SST boundary to K-w boundary conditions, It was supposed to get the same results as regular k-omega, but I did not.
And I found somethings confused at wall boundary for omega equation:
1, for regular K-w, w=7.2*Nu/Y**2 at y=y1
2, for SST, w=10*6*Nu/beta_1/Y**2=800*Nu/Y**2 at y=0 (beta_1=0.075)
My question are:
1, what does y=y1 and y=0 mean? what i understand is: y=y1 mean we specify this value to the first cell above the wall boundary, and y=0 mean we need to specify this value at wall face. Is that right?
2, cause w in SST model is 1/0.09 times w in the regular K-w model, so those two values are not identical.If they are same, the SST model w should equal to 6*Nu/beta_1/Y**2. SO why they are different?

Any suggestions will be appreciated

Thanks!

SKK · September 4, 2009, 05:06

I don't remember much detail about the Wilcox model, but according to the paper by Menter http://ntrs.nasa.gov/archive/nasa/ca...1993013620.pdf , in Wilcox's model, y is the distance from cell to the wall. It is the same in SST model, but in SST at the near wall cell, the value for Wilcox's w is multiplied by 10 as it gives much more accurate value for w. In general, because w is used to dampen the turbulence near wall, w values are expected to work better when they are high. Menter modified the w values to increase accuracy and to make it easier to implement in traditional CFD environment.

September 1, 2009, 13:35	Blending funtction F1 for SST turbublence model	#1
wooing New Member Join Date: Apr 2009 Posts: 5 Rep Power: 17	I am confused about the blending function for SST model: in Menter's paper, he said, F1 switch to one inside of boundary layer? But in my case i found in the free surface, I got F1=one also. Is that ok? or something wrong? Thanks in advance!

September 4, 2009, 05:06		#4
SKK Member Join Date: Mar 2009 Posts: 55 Rep Power: 17	I don't remember much detail about the Wilcox model, but according to the paper by Menter http://ntrs.nasa.gov/archive/nasa/ca...1993013620.pdf , in Wilcox's model, y is the distance from cell to the wall. It is the same in SST model, but in SST at the near wall cell, the value for Wilcox's w is multiplied by 10 as it gives much more accurate value for w. In general, because w is used to dampen the turbulence near wall, w values are expected to work better when they are high. Menter modified the w values to increase accuracy and to make it easier to implement in traditional CFD environment. Last edited by SKK; September 4, 2009 at 05:40.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
model	han	Main CFD Forum	0	September 27, 2005 09:35
references about the fan/radiator model	Mihai ARGHIR	Main CFD Forum	1	January 8, 2001 16:49
references about the fan/radiator model	Mihai ARGHIR	Main CFD Forum	0	December 21, 2000 04:06
references about the fan/radiator model	Mihai ARGHIR	Main CFD Forum	1	December 17, 2000 08:01
Why Turbulence models are not universal.	Senthil	Main CFD Forum	4	July 5, 2000 05:34

September 2, 2009, 05:48		#2
SKK Member Join Date: Mar 2009 Posts: 55 Rep Power: 17	F1 is formulated so that it is 1 in boundary layer and in regions of adverse pressure gradient. This switches the k-w part of the model in these situations. In the freestream it should be one. Check your formulation and the variables of F1. I would suggest have a look at the value of vorticity in the problematic area to see if it is behaving the way it should.

September 2, 2009, 13:50		#3
wooing New Member Join Date: Apr 2009 Posts: 5 Rep Power: 17	Thanks for your reply! I checked the formula, it was right and also the wall distance calculated rightly. For my cases I got a good convergence with regular k-w, but if i use sst k-w it did not convergence very well and F1 is one in some other region besides boundary layer. Last day i tired to set F1=one and F2=zero everywhere and switch the SST boundary to K-w boundary conditions, It was supposed to get the same results as regular k-omega, but I did not. And I found somethings confused at wall boundary for omega equation: 1, for regular K-w, w=7.2Nu/Y2 at y=y1 2, for SST, w=106Nu/beta_1/Y2=800Nu/Y*2 at y=0 (beta_1=0.075) My question are: 1, what does y=y1 and y=0 mean? what i understand is: y=y1 mean we specify this value to the first cell above the wall boundary, and y=0 mean we need to specify this value at wall face. Is that right? 2, cause w in SST model is 1/0.09 times w in the regular K-w model, so those two values are not identical.If they are same, the SST model w should equal to 6Nu/beta_1/Y**2. SO why they are different? Any suggestions will be appreciated Thanks!