# Want to implement a turbuence model! How?

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

 January 14, 2014, 20:09 Want to implement a turbuence model! How? #1 New Member   David Join Date: Feb 2013 Posts: 16 Rep Power: 6 My advisor want me to implement a 2-equation turbulence model, like K-W or K-E, I am wondering those two equations, should I simultaneously solve them together with N-S equations? Or, can I solve N-S equation firstly, and then treat the flow field as stable field, and then only calculate k-E equations for the fluctuating variables. Normally, how do you guys do it? Thanks.

 January 15, 2014, 17:25 #2 Member   Join Date: Jun 2011 Posts: 51 Rep Power: 8 Hi David, I would be curious to know how you will handle the clousure problem. AFAIK u need to related the unkonw Reynolds stresses to the known mean flow variabes through the turbulent transport equations. The latest can be 1eq. models, 2eq. model or second clousures. U must end up with: Full NS eq. - Averaged NS eq. = eq. for fluctuations. Regards,

 January 16, 2014, 12:46 #3 Senior Member   Martin Hegedus Join Date: Feb 2011 Posts: 486 Rep Power: 12 In general, the turbulence model is decoupled from the N.S. equations during the solution process.

 January 17, 2014, 00:51 #4 Member   Join Date: Oct 2012 Location: IIT-Hyderabad, India Posts: 42 Rep Power: 7 Hi!! This is how I had implemented k-e model 1. Initialize some velocity field. 2. Calculate k. 3. Calculate epsilon 4. Calculate turb viscosity. 5. solve NS to get new velocity field. 6. if velocity < stopping criteria, goto 2, else stop. If you have any doubts, please reply back!!

 January 17, 2014, 01:38 #5 Senior Member   Martin Hegedus Join Date: Feb 2011 Posts: 486 Rep Power: 12 A note, from what I understand, the two equation models are finicky about how the convection terms are discritized. The SA model is forgiving, so first order upwind is OK. Also, the two equation implicit methods are a little more tricky, assuming of course that an implicit method is used.

January 17, 2014, 12:45
#6
New Member

David
Join Date: Feb 2013
Posts: 16
Rep Power: 6
Quote:
 Originally Posted by samurai_01 Hi!! This is how I had implemented k-e model 1. Initialize some velocity field. 2. Calculate k. 3. Calculate epsilon 4. Calculate turb viscosity. 5. solve NS to get new velocity field. 6. if velocity < stopping criteria, goto 2, else stop. If you have any doubts, please reply back!!

THanks for all.

When calculating k and epsilon, what algorithm can I use? for example, if I used SIMPLE for velocity field, do I need do any special treatment for the K and E equations? Thanks.

 January 19, 2014, 01:03 #7 Senior Member   TWB Join Date: Mar 2009 Posts: 148 Rep Power: 10 Hi, I'm also trying to update my laminar code to turbulent. I'm reading Wilcox's turbulence modeling for CFD. Is there any books or online material which details the steps for a novice to follow?

January 19, 2014, 02:20
#8
Member

Join Date: Oct 2012
Posts: 42
Rep Power: 7
Quote:
 When calculating k and epsilon, what algorithm can I use? for example, if I used SIMPLE for velocity field, do I need do any special treatment for the K and E equations?
No there shall be no special treatment for that. just do the numerical discretization and apply Tri Diagonal Matrix algo if solving using FDM or go for regular explicit/implicit time stepping for the same. just remember that for collocated FVM grid, you will be required to substitute variables such as effective viscosity at face center rather than going for cell centers.

Give me some time, and i'll upload a document regarding the same, just remind me once after wednesday

Quote:
 I'm also trying to update my laminar code to turbulent. I'm reading Wilcox's turbulence modeling for CFD. Is there any books or online material which details the steps for a novice to follow?
I think Wilcox is the best for numerical techniques, but also have a look at Turbulent flow by Eswaran and Biswas, and also refer Peric's book on cfd.

January 19, 2014, 13:40
#9
New Member

David
Join Date: Feb 2013
Posts: 16
Rep Power: 6
Quote:
 Originally Posted by samurai_01 No there shall be no special treatment for that. just do the numerical discretization and apply Tri Diagonal Matrix algo if solving using FDM or go for regular explicit/implicit time stepping for the same. just remember that for collocated FVM grid, you will be required to substitute variables such as effective viscosity at face center rather than going for cell centers. Give me some time, and i'll upload a document regarding the same, just remind me once after wednesday I think Wilcox is the best for numerical techniques, but also have a look at Turbulent flow by Eswaran and Biswas, and also refer Peric's book on cfd.
Thanks, I'll read those for details, and hope you can upload those soon. Thank you very much!

 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 kick OpenFOAM 2 November 19, 2013 18:28 roukaia FLUENT 0 December 24, 2011 10:10 sherifkadry CFX 2 September 7, 2009 20:51 rystokes CFX 3 August 9, 2009 19:13 sven OpenFOAM 0 July 19, 2009 15:47

All times are GMT -4. The time now is 04:54.