[annu]

hi friends,
i am doing my project based on species transport without reaction. Its basically release of gas through a tunnel of 1200 m. i am taking pressure based solver,unsteady state,k-epsilon model(2 eq) ,pressure velocity coupling---PISO,first order upwind. step size=0.01 max iteration per time steps is 20...when i run these it says "turbulent viscosity limited to viscosity ratio of 1.0000e+005 in 6328 cell". i had refined the grid but problem remains the same.
gas is released through a 180mm hole. flow rate is very much high. i am taking fluid as incompressible.

please help me friends, its urgent

thankyou in advance..

[misterdd]

try to change air viscosity to sutherland instead of constant value. hope this will help your problem.

[RodriguezFatz]

Sometimes this is due to bad initialization and the warning will vanish after a few time-steps. Do you "need" the first time-steps or could you just run the simulation and start evaluating later?

[ghost82]

Following RodriguezFatz advice, what about to calculate a steady state solution with only air and then change the injection of air with your tracer?

Daniele

[annu]

I had firstly run the simulation without tracer in steady state for 240 sec Then I inject the tracer.

I had changed thermal conductivity and viscosity to ideal gas mixing law.warning is same. Turbulent viscosity is not disappearing after 200000 iterations.

Thankyou all for your responses..

Please help.

[Zaktatir]

-Look after the inlet / outlet specification of your turbulence unerliying properties for the calculation of the kinetic energy and turb. dissipation. Use middle intensities and viscosity ratios. Start first of all with first order turbulence handling and then monitor the ratio to see whether your simulation iw q kind of fullscale simulation where you may expect very high ratio of turb. to molcular viscosity

[annu]

Zaktatir , i am using turbulent intensity and hydraulic diameter as specification method for inlet and outlet.

from 240 sec i mean i had run for a week then injected tracer.



misterdd- i had changed it to sutherland viscosity, but viscosity in number of cells is increasing. no impact.

[RodriguezFatz]

Look at the results: Where does the viscosity limiting occur? At the inlet / outlet / at some corners?

[annu]

RodriguezFatz...
sir, u mean plotting the countours of turbulent viscosity ratio?

I had plotted, it shows red colour contours means maximum at velocity outlet that is from hole from gas is leaking..

I had again meshed the grid by taking more interval counts for hole..results are same..

[RodriguezFatz]

Can you post the pictures?

[annu]

sir it exceeds the limit of 97.7

[annu]

sir had attached ..

[RodriguezFatz]

1) What's your y+?
2) I can not see the problematic regions in the pictures, since you limited the scaling to 1. Where is it 1e5 ?

[annu]

y is height of tunnel. i.e 7m

[RodriguezFatz]

y+ is the solution of the numerical grid at the wall. You need to plot it on the wall.

[annu]

sir,I dint get U. I am new to fluent. Please explain.

[RodriguezFatz]

This is too substantial to be explained in the forum. You should get some basic CFD book and try to understand it before starting with simulations. "y+" will be explained in every good book.

[annu]

mine problem is turbulent viscosity ratio is not disappearing after so many iterations.

[annu]

vertical plane which i made is middle of tunnel and also it is centre of sqaure shaped hole mm dimension

y is height
and z is length
and x is width

[Zaktatir]

1/familiarize your self with basic turbulence modeling and understanding
THAT IS THE MOST IMPORTANT POINT IF YOU WANT TO UNDERSTAND WHAT ARE YOU SIMULATING

2/For full scale geometry is possible to get very high ratio: I simulated a Train and the ratio was above 10 power 8. The large scale of the eddies are dependent on the geometry
3/Monitor the viscosity ratio after increasing the limit of it (let say to bigger value than the limit)
4/Look whether the value is converging to a constant range: if not you have a problem
5/Do you have any installations modeled in your tunnel