I am interested to model and simulate the flow of fluid in T-shaped geometry.

Is there a free software for this purpose? Or among the commercial ones which is the best for this purpose?

thank you

I have recently read a paper regarding the simulation of T-type mixers. CFD-ACE code, which is a commercial code, was used for the simulation. You might be interested to read this article:

Gobby, D., Angeli P., and Gavriilidis, A., "Mixing characteristics of T-type microfluidic mixers", J. Micromech. Microeng, Vol. 11, pp. 126-132, 2001.

Anyway, I am going to simulate non-Newtonian fluids too!

Dear Farished

Thanks for your response and information. Are you considering two dimensional(2D) or three dimensional (3D) prblem?

What is exactly the difference between 2D and 3D CFD in terms of computational approach.

Thank you solomon

Hi all,

I think it would be interesting to make clear what you mean by "non-Newtonian fluids".

Are you talking about viscoelastic, viscoplastic, shear-thinning, ... fluids ?

The kind of fluid (and the corresponding constitutive equation that models the behaviour of the material) will definitively influences the choice of the numerical method you would adopt or the commercial code you would use to simulate your problem.

Depending on those answers, you would have to face some very peculiar (but anyway very interesting and challenging) problems often called "high Weissenberg numbers" for viscoleastic fluids, "high Bingham number" for viscoplastic (yield stress) fluids. If hopefully for you, your fluid has only shear thinning (or shear thickening) properties, then you will only have to solve a classical velocity-pressure viscous problem, which is quite straightforward.

Rheologically complex material often highlight strange and funny properties, that could be noticed experimentally but difficult to simulate (convergence of the solving algorithm fails).

I would be glad to answer your questions if you don't mind to be a little bit more specific.

Anyway, it is nice to see that some people involved in CFD projects or studies are interesting in something else than turbulence !!

Best regards, Anthony

Dear Anthony

Well I am new to the area of CFD. I want to get the way to start my research. I want to model the flow of non-newtonian fluid in 3D T-shaped geometry.

By non-Newtonian I mean the viscosity is not constant and dependent on the rate of shear strain.

The fluid I am modeling is blood. Whatever appropriate assumtion need to incorporated I want to include and taste it. At this point I have no idea what to specify in this regard.

Hi Solomon,

Thanks for your answer, this is exaxctly what I expected : "viscosity of your fluid is shear-rate dependent".

Thus, at this point, if you consider only a shear-rate dependent viscosity model, a standard CFD code as FLUENT will clearly be enough to solve your problem.

I am currently using FLUENT to simulate flows of shear-rate dependent viscosity fluids and viscoplastic (Bingham and Herschel-Bulkley) fluids as well. FLUENT is quite well suited for this type of calculations but cannot handle viscoleastic fluid flows calculations.

Another CFD code completely dedicated to non-newtonian fluid flows is POLYFLOW that can take into account shear-rate dependent viscosity, viscoelastic and viscoplastic models. I have never tested this code myself, thus I cannot give you my opinion about its features and performances but I heard that it is not too bad ...

You are working on blood flows : that's very interesting and challenging. And blood and biofluid fluids is a really new Fluid Mechanics and CFD area.

Good luck

Anthony

Dear Wachs

That is a very wonderful guideline you gave me about the problem. But still I have some other related question.

How about the capability of FLUENT in handeling flow in T-shaped three dimensional geometry.

Thank You Solomon

Solomon,

I don't think Fluent will have any problem with your geometry. It sounds simple enough, t-shaped tubing with (I assume) circular cross section is meshed and preprocessed quite quickly with Fluent (or most other codes for that matter). Provided your bc's are straighforward of course.

Martin

Solomon,

Regarding the difference between 2-D and 3-D simulations, because you would like to use a commercial code and you will not write your own code, the only difference is in computational costs (CPU, memory, etc.).

Anyhow, concerning b.c.'s which are appropriate for the simulation of non-Newtonian fluids (such as blood), I should aware you that at high shear rates, the no-slip condition breaks down due to the emergance of slip which increases rapidly with augmentation of shear rate. Particularly, this may happen at the corners. So, definig b.c.'s might not be as straighforward as usual.

Good luck with your simulation.

I to am using non-newtonian fluids. In my case i am simulating the flow round a rotating auger (drill) and will probably be using bingham or other visco-plastic models as well as possibly visco elastic models. At the moment i am using star-cd for my initial simulations as i have not yet got into programming the non-newtonian behavoir. How suitable is star for such visco elastic/plastic flows and how easy is it to program?