code development - non dimensional n-s code vs dimensional k-epsilon model

Garp · May 12, 2015, 01:16

Hi all, I have developed a cfd solver based on NON-DIMENSIONAL equations and the code can produce results the way it should be. But the problem arises when I want to implement the k-epsilon turbulence model in the code which is in DIMENSIONAL form. Can I just add the additional term in the k-epsilon equations (Reynold stresses term) into my N-S solver without non-dimensionalising the k- epsilon equations?

Please help me

som87 · May 12, 2015, 18:48

I'll be precise: No you cant, k & epsilon both are dimensional parameters and they are needed to be in same 'form' as rest of your equations.
Furthermore, k-epsilon is a dynamic model unlike a static one like Spalart Almaras, so you must take extra care while you are linking the equations.

Try to look for Morrison J.H.'s work for detail (I think you'll be able to find all other related questions as well)

Garp · May 13, 2015, 06:50

Thanks for the reply. So how do I implement the k epsilon model? Do I need to develop new solver based on Dimensional form? That is huge.. Or do I need to non dimensionalise the k epsilon equations in which I have never seen anyone doing that

Do you have any suggestion? So, how does commercial cfd codes such as fluent deal with it? usually new solver will be based on non dimensional form. I can bet on that.. Thanks a lot for your time

May 12, 2015, 01:16	code development - non dimensional n-s code vs dimensional k-epsilon model	#1
Garp New Member Monkey D Garp Join Date: May 2015 Posts: 2 Rep Power: 0	Hi all, I have developed a cfd solver based on NON-DIMENSIONAL equations and the code can produce results the way it should be. But the problem arises when I want to implement the k-epsilon turbulence model in the code which is in DIMENSIONAL form. Can I just add the additional term in the k-epsilon equations (Reynold stresses term) into my N-S solver without non-dimensionalising the k- epsilon equations? Please help me

May 12, 2015, 18:48		#2
som87 Member Join Date: May 2012 Posts: 68 Rep Power: 14	I'll be precise: No you cant, k & epsilon both are dimensional parameters and they are needed to be in same 'form' as rest of your equations. Furthermore, k-epsilon is a dynamic model unlike a static one like Spalart Almaras, so you must take extra care while you are linking the equations. Try to look for Morrison J.H.'s work for detail (I think you'll be able to find all other related questions as well) __________________ Best wishes, Somdeb Bandopadhyay Last edited by som87; May 13, 2015 at 07:52.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Calculation of Epsilon in LES model	Yuri Almeida	OpenFOAM	2	January 18, 2013 08:43
Open source CFD code development, possible?	Dr. Yazid Bindar	Main CFD Forum	27	July 18, 2000 01:18
Activities of national CFD code development?	Dong-Kyu, Chung	Main CFD Forum	1	July 6, 2000 10:12
Use of an hydrodynamic model integrated in the vertical by the PHOENICS code	Hafsia Zouhaier	Main CFD Forum	2	March 15, 1999 09:23
public CFD Code development	Heinz Wilkening	Main CFD Forum	38	March 5, 1999 12:44

May 13, 2015, 06:50		#3
Garp New Member Monkey D Garp Join Date: May 2015 Posts: 2 Rep Power: 0	Thanks for the reply. So how do I implement the k epsilon model? Do I need to develop new solver based on Dimensional form? That is huge.. Or do I need to non dimensionalise the k epsilon equations in which I have never seen anyone doing that Do you have any suggestion? So, how does commercial cfd codes such as fluent deal with it? usually new solver will be based on non dimensional form. I can bet on that.. Thanks a lot for your time