CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   some help (https://www.cfd-online.com/Forums/main/4162-some-help.html)

sadbhaw November 29, 2001 18:59

some help
 
Hi this is my first encounter with Fluent... i dont know even know how to work with Fluent.. but want to do a project using Fluent. Can you guide me how to rerun samples(that are provided by Fluent Inc) in Windows NT.

i am first year graduate and will be doing my research in CFD next semester. Thanking in advance sadbhaw

Andrew Parker November 30, 2001 06:57

Re: some help
 
What do you need help with, I would be happy to try to help on a more specific topic, also you may have more response if you post of the fluent forum.

Andy

sadbhaw December 1, 2001 15:38

Re:Re: some help
 
Hi Andy,

I am working on one of the test cases given in Fluent(i.e.Transient Turbulant Flow around Airfoil) for my project. I need to compare results of all the models used by Fluent. In the Fluent documentaion there is guideline to how to use Fluent for Spalart-Allmaras(1 eqn). But I have no idea how to use for other models...specifically I want to use for K-e model and K-e(RNG) model...

Can you give me some guideline on this. Thanking you sadbhaw

Thomas P. Abraham December 1, 2001 17:53

Re:Re: some help
 
Remember, that you are only using FLUENT and not coding. For using other models, you just need to change the option for the turbulence model, that's it. Whether you are using wall functions or not would dictate the type of mesh that you need. For knowing this, you need to understand how the turbulence models work.

Thanks, Thomas

Andrew Parker December 2, 2001 07:57

Re:Re: some help
 
I would agree with Thomas, it is simply a case of changing the turbulence model in the viscous panel, however, from my experience and mistakes I would caution you against two possible problems. These relate to changing the turbulence model once you have a converged flow. If you change to say the standard k-e, then the turbulence parameters at inlet will be set to default, and the converge level will be set to default, if it where me I would always start a new case when you use new model.

A little about the models:

You should firstly read about the law of the wall, it is in the manual. If you can understand this, it will explain y^+ and the near wall treatment. The benefit of using S-A is that you need at y^+ of around 30-course mesh, if using the wall functions. However, it is recommended by fluent that if possible use two-layer zonal and have the y^+ around 4-5. This may not mean that much, so what I would recommend and is what I did is understand when the different wall treatments are appropriate, understand the law of the wall diagram, and the concept of y^+ then you are home and dry.

If you can teach yourself the above then you will understand for your flow situation what is appropriate, then it is just a case of keeping the code happy by specify the correct y^+ and near-wall treatment.

I know this not an answer, in fact I have suggested you go and read, but from my experience, if you put a way the code and spend a day or so understanding the flow features to expect, adverse pressure gradient, separation, sock waves, and so on. Then understand what the code needs to give you the answer, i.e. the near-wall treatment and the y^+ , then you can apply this to the different models keeping in mind the quirks of the different models then you should be ok.

Hope this helps feel free to respond with more questions if it does not, although I am still a student so it will come to a point where I cane be no more help.

Cheers

Andy


All times are GMT -4. The time now is 02:51.