CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Turbulent viscosity limited to viscosity... (http://www.cfd-online.com/Forums/fluent/46106-turbulent-viscosity-limited-viscosity.html)

Cyril September 21, 2007 20:29

Turbulent viscosity limited to viscosity...
 
Hi,

I'm studying 3D external aerodynamics (subsonic M0.8, SA viscosity, coupled implicit) and I have a message saying that "Turbulent viscosity limited to viscosity ratio of 1.00000e+05 in xxx cells (and xx keeps on increasing)

What does this message means and how to fiwx that prolem ? (I set up the Curent number to 15 (but i don't know what it really is)

Thanks!

Cyril September 23, 2007 10:54

Re: Turbulent viscosity limited to viscosity...
 
I tried to adapt the mesh where the turbulent visosity ratio was high.. the number of cells started to decrease, but then started to increase again...

Do you think i'm using the right viscosity/solver scheme?

Thanks

mAx September 24, 2007 03:12

Re: Turbulent viscosity limited to viscosity...
 
Hi Cyril, Did you well described the turbulence quantities in your BC?

Cyril September 24, 2007 04:14

Re: Turbulent viscosity limited to viscosity...
 
I defined the BC as follow:

- fuselage: wall

- Symetry: symetry

- External surfaces (in the fron, behind, below..): pressure far field - gauge pressure 101325 - Mach 0.8 - Temperature 300.

- Operating conditions: 0

- Viscosity: Spalart allmaras

- Air: ideal-gas - sutherland three-coefficient-method 1.716e-05 - 273.11 - 110.56

I've never been working with viscosity before, but I based my BC and on the fluent tutorial (wing in M0.8).. Do you think they are good ?

Cyril September 24, 2007 04:16

Re: Turbulent viscosity limited to viscosity...
 
I also ran a first set of iterations using the first order for both flow and viscosity, and then i tried to put the second order.. (but here again, i tried to find some explanations on the internet, but i still dont understand the difference between second order, power law, quick...)

mAx September 24, 2007 05:35

Re: Turbulent viscosity limited to viscosity...
 
I'll check out tomorrow, I have to go. But for your BC, you should fill the turbulence quantities below Mch- Number etc...

For the first and second order, the difference is the truncated error of the discretization.

In french: c est les developpements limites (1er et 2e ordre)

Cyril September 24, 2007 05:41

Re: Turbulent viscosity limited to viscosity...
 
I set up the Turbulent viscosity ratio to 10 as proposed in the tutorial.

But honestly... I dunno what that means (10.. why not less or more..)

Thanks for help :)

mAx September 24, 2007 07:50

Re: Turbulent viscosity limited to viscosity...
 
switch on Turbulence Intensity and Hydraulic Diameter (Method). *Then you calculate your Hydraulic Diameter by Dh = 4.A/P , where A is the Aera of your Inlet and P the Perimeter *Turbulence Intensity is given by: I = 0.16*Re**(-1/8), with Re is your Reynold's Number


Cyril September 24, 2007 08:15

Re: Turbulent viscosity limited to viscosity...
 
My case is external flows... I dont have a proper inlet.. should I use the the front face of my mesh as an inlet for the area/perimeter?

Cyril September 24, 2007 10:53

turbulent BC.. TVR or Intensity/Hydraulic Diam?
 
Arf... i'm dumb or unlucky..

Tried with Dh=14m and Intensity=1.41% and the residuals diverge and computation stops after 42 iterations :'(

mAx September 25, 2007 01:30

Re: turbulent BC.. TVR or Intensity/Hydraulic Diam
 
ok, but you don't have the warning regarding turbulent viscosity anymore, right? With M=0.8 you are in transsonic flowfield,eg: compressible Did you enable energy equation?

Cyril September 25, 2007 01:58

Re: turbulent BC.. TVR or Intensity/Hydraulic Diam
 
Yeah, the energy equation is on. But as the residuals diverge, the calculation stops by itself..

mAx September 25, 2007 02:42

Re: turbulent BC.. TVR or Intensity/Hydraulic Diam
 
ok, and which variable is diverging first? (do you receive an error message)? I am not familiar with pressure far field BC,so try to switch your front inlet with velocity Inlet. I will send you an document about CFD and extern-aero (best practices)


PratiK Mehta October 2, 2007 10:08

Re: turbulent BC.. TVR or Intensity/Hydraulic Diam
 
i was goin throught the discussion and so i want to ask Cyrll r u still having trouble in solving or u have solved it ?

Cyril October 2, 2007 13:28

Re: turbulent BC.. TVR or Intensity/Hydraulic Diam
 
Still have some problems... mainly Turbulent viscosity ratio limited to ... even if I adapt my mesh on the TVR gradient it doesn't really help.

I tried to run my simulation for Mach 0.2, (nearly no pb) then Mach 0.3 ("TVR limited..." appears as the residuals converge :s)

And when I try to adapt the y+ gradient, the residuals diverge (about all at the same time). (don't have enough ressources (RAM) to refine the mesh with gambit, that's why i do it with fluent.

But my report is to be finished by tomorow..... :( so there's not a lot i can do at this point!

And when I display the turbulent viscosity ratio, it seems good at the rear of the fuselage, but it's strange there are no turbulences behind the cockpit.... http://img251.imageshack.us/img251/2...itrezc1.th.jpg

Oscar October 3, 2007 07:44

Re: Turbulent viscosity limited to viscosity...
 
Hello! I have the same problem, but my case is about a car, at 200 km/h. I've configured all correctly, (i think) using the K-epsilon model for turbulence, without any further complication, and at the second iteration, Fluent say "Turbulent viscosity limited to viscosity ratio of 1e5 in xxx cells" i don't use the energy equations, and my fluid is non-compressible, you know, an easy car analysis. So, where is my problem? in the Mesh? in the configuraiton?

Thanks for your help!

PratiK Mehta October 4, 2007 06:12

Re: Turbulent viscosity limited to viscosity...
 
hi Cyrll, Make sure you have set your turbulence intensity value to about 10e-3, which i suppose u already did ,if not then make it now. Try to switch to unsteady state with small timesteps likei generally use the thumb rule as timestepsize=(L/V), L= reference lenght V=free stream velocity. Value u get should be used as timestep, if u get divergence ,just try to reduce this value.

Keep the pressure overrelaxation to 0.8, just use 1st order scheme for all variable

Use PISO for pressure.velocity coupling

I hope this helps



All times are GMT -4. The time now is 09:35.