turbulent viscosity limited to viscosity ratio of

 Register Blogs Members List Search Today's Posts Mark Forums Read

 April 15, 2003, 12:19 turbulent viscosity limited to viscosity ratio of #1 Hua Guest   Posts: n/a hi, everyone I met this warnning when I run fluent, I am using RSM, and Reynolds number 10E6. What is the problem for this and how to solve it. Thanks a lot

 April 15, 2003, 16:13 Re: turbulent viscosity limited to viscosity ratio #2 Alex Munoz Guest   Posts: n/a Hi There is a lot of discussions about this topic, therefore I suggest that you search other cases on this web. The problem should be solved by identifying the cells where you have the problem and then refine the mesh on those areas. Best regards Alex Munoz

 April 15, 2003, 17:24 Re: turbulent viscosity limited to viscosity ratio #3 Jongdae Kim Guest   Posts: n/a Why don't you try as follows (If you still have the same warning, please go to the next step. Usually, the initail flow condition used for the RSM run is obtained from the RNG k-e model result); First step: Solve - Controls - Solution -Default => iteration Second Step: Decrase "Under-relaxation factors" => iteration Third Step: Adaptation of cells : I usaually use y+ and velocity gradient conditions => iteration Fourth Step: Regenerate mesh, goto step 1 If your solution stats to converge, you can increase under-relaxation factors. If you have converged solutions, you can increase the order of the discretization parameters (for ex. 1st -> 2nd -> QUICK etc.) KIM

 April 15, 2003, 20:59 Re: turbulent viscosity limited to viscosity ratio #4 Alex Munoz Guest   Posts: n/a Hello Hua Kim as many others play with the relaxation factors to avoid during the first iteration the warning! Personally, I do not like to reduce this relaxation factor, even more, usually I increase those parameter to obtain solucions in a shoter period of time. Now, Kim suggests to adapt your mesh according to y+ and soon.... Hua, every case is different anybody can tell you where to adapt your grid without getting an idea of which type of geometry are you working with. Therefore, I strongly suggest that you check first where are you getting high turbulent viscosity ratio Another point, Kim suggest that you run first the k-e model or one of its relatives, this suggestion is very important, I forgot to tell you because I think you have done this option before. Now, come the question have you getting the same warning with k-e model?? Another point, Kim send you to use Quick discretization and soon... Careful, again this type of discretization is very sensitive. I strongly suggest that you try harder with the second order discretiztion before applying Kim's idea Perhaps, You should know that overall what it is more important is that you provide a good estimation for the boundary condition (velocity, turbulent kinetic energy and dissipation rate) after all waht this codes made is just proyect your boundary condition into the domain. My advice, Careful with the suggestion that you recieve, each person in this forum thinks that all the cases behave the same and that is not true You are the only one who can evaluate if you keep going with this Kim' scheme, or you develop a systematic analysis of your case!!

 April 21, 2003, 18:37 Re: turbulent viscosity limited to viscosity ratio #5 Hua Guest   Posts: n/a Thanks you two, for your reply. I tried both of your ideas. I have changed to k-epsilon model first and adapt the grid base on Y+. The result seems converging, but there are still warnings that viscous ratio is too high in 80 cells, do I need to adapt the grid more, or can I trust this results? I check the FLUENT manual, not find the meaning of the turbulent viscosity ratio, can you please give me some explaination what it is and how it affect the results? Tnanks a lot.

 April 21, 2003, 19:04 Re: turbulent viscosity limited to viscosity ratio #6 Alex Munoz Guest   Posts: n/a Hi Once again, you are too broad I don't know your case, your domain, your objectives, or even the project goal. Without this information it is impossible to provide an advise. Hua, I know that the project is confidential and soon. However, You cannot be so general. For instante, you write it seems to converge. a case converge to a number in your case I don't know 1e-04??? The turbulence models selection depends of the complexity of the flow field and the need for accurate prediction, you can get in most of the cases convergence with the k-e model. However, you don't know if this model provide the most accurate flow field description until you compare with experiments or other source of data. Anayway, I cannot help you till you provide realiable information. Best regards Alex Munoz

 April 24, 2003, 16:45 Re: turbulent viscosity limited to viscosity ratio #7 Hua Guest   Posts: n/a thanks, Alex. Actually I know the numerical result, if not compared to the experiment, can not be viewed as accurate or reliable. For some reasons, I can not talk about the detail of the problem, even inconvinient to do that since there is no figure-posting function here. For that run, what I can tell is that convergence is to 1E-5, but there is still warnnings about turbulent viscousity. So I am asking according to your experience, if this also happens in some of your work. And also if not troubling you too much, give us an introduction about the conception of this headache term--turbulence viscosity ratio. I believe I am not asking for specified case, just general explaination is fine. Thanks again, Alex.

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post ZQWY FLUENT 15 June 19, 2016 03:00 omar.2002bh FLUENT 2 September 5, 2012 11:04 eespi002 FLUENT 3 June 30, 2009 13:24 Cyril FLUENT 8 February 7, 2006 21:34 David Yang FLUENT 3 June 3, 2002 06:13

All times are GMT -4. The time now is 22:01.