[alsemio]

Hi everyone
I was trying to run a car model with 1million cell aproximately using k epsilon turbulence model and I have the following problem :

Turbulent viscocity limited to viscocity ratio of 1.0000 in 400 cell
and the number of cell goes increasing

after many iteration appear to me the following message
Error:divergence detected in AMG:temperature
Error)

Can please anybody help me to solve this problem?
thank you all

see you in advance

Alsemio

[swiftaircraft]

Have you tried running it with the energy equation off? I assume you are assuming a real gas. When this converges a little you should then be able to activate energy again. However for low speed flows it would not add that much to the accuracy of the simulation.

[alsemio]

Hi David
First of all thank you for anwer me and help me with this problem
let me tell you that I´m trying to solve a simple car simulation at 50 m/s speed using k-e solver
As well I turn off the energy equation because I using incompresible flow
so I cannot understand why do I have this problem

Fluent always show me turbulent viscocity ratio in xx cells and the number of cell rise

Thank you so much and I hope that someone else can help me to solve this I dont know what to do because I changed set up and still same problem

See you in advance

Alsemio

[swiftaircraft]

How have you initialised the problem? Do you use fmg initialisation? What are your underrelaxation factors? Are you using a velocity inlet with an outflow outlet? I assume for defining the inlet turbulence you have used a low turbulent intensity. What is the maximum skewness in the model? Please provide as many details as possible.

[alsemio]

Hi David thank you for help me I´m new using CFD fluent so to answer you question I´m using Gambit to mesh so I used 1.200.000 tetrahedral elements with skew below 0.8 only I have 98 elements between 0.8 and 0.9 and in Fluent I use k- e viscous model Standard I left the turbulent kinetic energy and turbulent dissipation ratio as default 1 (I dont understand to compute this values) I iniatilize with velocity in inlet
I hope to hear from you
Thanks a lot David
Alsemio

[swiftaircraft]

Set the inlet turbulence using Intensity and Viscosity Ratio. Set Intensity to 0.05% and a viscosity ratio of 1. Initialise the flow with the values obtained when you select the name of your inlet in the "Compute from" drop down box. Then in the Fluent window type "sol ini fmg y". Do not use the quotation marks though. Also use the realizable k-epsilon model to start. Initially set the discretisation to first order. See if that helps. Also change default under relaxation factors so that Momentum is 0.3 and Pressure is 0.7.

[alsemio]

Thank you so much David
I will try your advice I have my finger crossed
let me tell you the result if you want

See you in advance

Kind Regards

Alsemio

[swiftaircraft]

Please let me know if it works.

[ComputerGuy]

I would also ensure that your cells in the boundary layers are appropriate. The message typically "goes away" after the solution gets close to steady state. However, if your boundary conditions are not well defined, or your mesh is poor, they will continue to pop up. Finally, you can always set the solver limits on viscosity ratio to be slightly higher. It's cheap and isn't always effective, but you can give it a shot.

ComputerGuy

Originally Posted by swiftaircraft

Please let me know if it works.

[alsemio]

Thanks a lot Computerguy and Swiftaircraft
I think both you are right
but I dont quite undertand how much and where to put viscocity ratio limits
kind regards
alsemio

[alsemio]

Hi David
Sorry for bother just to know for cuiosity what does fluent do writting "sol ini fmg"

Thank you again
alsemio

[ComputerGuy]

Alsemio,

Check the following for how to enter limits:http://my.fit.edu/itresources/manual...g/node1381.htm

Originally Posted by alsemio

Hi David
Sorry for bother just to know for cuiosity what does fluent do writting "sol ini fmg"

Thank you again
alsemio