CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Main CFD Forum

Doubts about Larsson non-linear model

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

Reply
 
LinkBack Thread Tools Display Modes
Old   January 11, 2007, 05:11
Default Doubts about Larsson non-linear model
  #1
Julian Jaramillo
Guest
 
Posts: n/a
I implemented and I'm using non-linear eddy viscosity model presented by J. Larsson in Proceedings 13th ISABE conference, pp. 1214-1222, 1997. However, when I've applied this model to the impinging flows studied by Abdon and Sunden in Num. Heat Transfer: part A, #40, 2001, convergence and instability problems have been observed. Also, checking formulation used by Abdon and Sunden I've found some differences with the original formulation.

Could anyone confirm if this new formulation is correct? or if the original model has been changed for some reason?. Because using the formulation by Abdon results are different and convergence is improved.

Thank you very much.
  Reply With Quote

Old   January 11, 2007, 10:16
Default Re: Doubts about Larsson non-linear model
  #2
Jonas Larsson
Guest
 
Posts: n/a
The original version of the model has been used by several people and has been implemented in several codes. It usually gives fairly realistic results and does not lead to any significant convergence problems compared to a normal linear two-equation model. Many other non-linear models can be quite tricky but this model is fairly nice I think.

Sometimes the model produces less turbulent viscosity in regions with large normal strain (stagnation regions, shock regions and regions with very strong acceleration). This can sometime give stability problems because the incorrectly prediced high turbulence levels caused by most classical models tend to stabilise the flow and reduce unsteady behaviour and robustness. Having a robust model which is robust just because it predicts too much turbulence isn' very good though.

I'm not sure what Abdon did change in his implementation of my model. I remember emailing with him about this many years ago though, so he probably did do something. I can't help you with explaining excatly what he did and what causes it has though. I've never used his version. Sorry.
  Reply With Quote

Old   January 12, 2007, 05:54
Default Re: Doubts about Larsson non-linear model
  #3
Julian Jaramillo
Guest
 
Posts: n/a
Thank you for your soon answer. I've used the original formulation of the model in channels and backward gacing step configurations with success. But problems have appeared in the impinging jet case, specially as you commented near the stagnation region. In this zone I've observed that the coefficient of the non-linear term, which involves square root, can take undefined values (square root of negative values). I've restricted the term within the square root to be strictly positive. But, instabilities are still present.

Could you please recommend me any other strategy in the implementation of this model or tell me about some paper where this model has been used? (apart of those by Abdon).

Thank you very much for your kindness.
  Reply With Quote

Old   January 12, 2007, 06:51
Default Re: Doubts about Larsson non-linear model
  #4
Jonas Larsson
Guest
 
Posts: n/a
The only one I know of who has used the model on impinging flows is Andreas Abdon I'm afraid. It has been used on many turbine blade simulations though, where you at least have an impingement region at the leading edge of the blade.

Sometimes during convergence you might run into stability problems as you describe. One strategy which I used when instabilities etc. occured was to set a minimum level of k at the start of the similation in order to not obtain too low turbulence levels during convergence. Once the flow has stabilized and started to converge it should hopefully be possible to reduce and eventually remove this lower limit of k. Sometimes the flow is inherently unstable though and then it will be difficult to get it to conerge well in a steady simulation.

I've also sometimes used an upper length-scale limiter in order to not obtain completely unrealistic turbulent length-scales during convergence. There are many tips and tricks in the turbulence modeling world I'm afraid. Excatly which limiters and constraints you are using in your code is sometimes just as important as which turbulence model you are using. This is true for most two-equation models.

I haven't used this model myself for many years (I'm working in industry now). Have you checked out if Shih in his original formulations of this realizable non-linear model formulation has had any similar problems with the C2 constant?
  Reply With Quote

Old   January 17, 2007, 07:38
Default Re: Doubts about Larsson non-linear model
  #5
Julian Jaramillo
Guest
 
Posts: n/a
Thank for your answer. I agree with you, tips and tricks are very important in turbulence modeling at any level. But the problem is that, generally, they are not documented. Until now I haven't used Shih model, but I'am using Craft, Launder and Suga model and Abid, Morrison, Gatski and Speziale model. All of them are more difficult to converge than linear models in impinging situations. However, with the model I've found more problems is with the Larsson model, specially when the grid is refined.

Could you tell me more details about these constrictions you used?, or could you recommend me any book or article where I may find information about these tips?

Thank you again
  Reply With Quote

Old   January 20, 2007, 12:49
Default Re: Doubts about Larsson non-linear model
  #6
raju
Guest
 
Posts: n/a
Can somebody send me this paper by J. Larsson in Proceedings 13th ISABE conference, pp. 1214-1222, 1997. I would like to got through it.

My email id is mandhpati@yahoo.com
  Reply With Quote

Old   January 24, 2007, 06:51
Default Re: Doubts about Larsson non-linear model
  #7
Julian
Guest
 
Posts: n/a
You can find it in this web:

http://www.cfd-online.com/Users/jola/
  Reply With Quote

Reply

Thread Tools
Display Modes

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 Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Use of k-epsilon and k-omega Models Jade M Main CFD Forum 23 February 8, 2017 15:27
Atomization and Breakup model xck1986 OpenFOAM 4 May 31, 2014 04:43
Water subcooled boiling Attesz CFX 7 January 5, 2013 04:32
IdeasUnvToFoam Bug amp Fix benru OpenFOAM Bugs 42 November 13, 2009 08:59
Multiphase flow. Dispersed and free surface model Luis CFX 8 May 29, 2007 18:13


All times are GMT -4. The time now is 01:00.