are the turbulence models that used for newtoninan fluid diffrent from non-newtonian fluid in fluent? how about in general?
thanks in advance
I asked ANSYS tech support, whose response is below.
The turbulence models for use with non-Newtonian fluids are the same as those for newtonian fluids. Newtonian fluids are implemented by changing the viscosity dependency for the fluid material; the other selections stay the same.
I do not quite understand so I'm asking for clarification. However, I thought I'd go ahead and respond with the information thus far.
I am not sure about it.
what about bossinesque approxiamtion?
how is turbulence viscosity in non-newtonian fluids defined?
turbulence characteristics of newtonian fluids seems to be different than newtonains.
I am not quite satisfied with the response either. When I asked for elaboration, tech support responded
I meant that all the turbulence models we have support non-newtonian fluids. The implementation of non-newtonian viscosity is similar to making property change to the fluid. The Non-Newtonian model just allows viscosity to depend on Shear stress and other variables.
I also submitted a new service request to see the response from another person, who stated
With non-Newtonian models we generally alter only the dynamic viscosity term, other terms are calculated in the same fashion.
I still do not completely understand but this is all the time that I have to devote to this issue at the moment.
From the CFX documentation, it looks like built-in non-Newtonian dynamic viscosity models are new in Release 12, which is probably a great improvement. However, I do not see anything in reference to turbulence modelling. I spent a great deal of time understanding turbulence modelling earlier this year and did not see anything in reference to non-Newtonian fluids in the CFX documentation. I am not sure as to a good source to refer you to. If you have access to ANSYS tech support, I would suggest directing questions to them.
I agree with Hamid. The Boussineq approximation does not hold for non-newtonian fluids and the definition of turbulent viscosity has to be modified. There was a study on the addition of polymer particles in blood a while ago (2002 maybe) and they found that the flow became turbulent at Re ~ 100.
You can trying using the Reynolds stress models (provided the numerical computations remain stable). Another option might be to check the documentation of OpenFOAM online. The implementation of a non newtonian turbulence model might be similar in Fluent and OpenFOAM.
|All times are GMT -4. The time now is 12:10.|