Star-CD and K-Epsilon turbulence model

hamoudi · June 1, 2009, 11:19

Hello Everybody,

I have a simulation on a numerical jet of air For this I use the Star-CD with Turbulence Model K-epsilon
as you know at the beginning of the simulation we need to initialize the values of
K and epsilon ( Energy and Dissipation rate of turbulence )
to calculate these two parameters
Energy
K = 3 / 2 * (U * I)
Velocity of Jet . U = 286 m / s
I= Intensity = 5 %
K= 360,73 m2/s2

Epsilon

Epsilon = (cmu 3 / 4) * K (3 / 2) / L

cmu = 0,09,

L= Mixing Length = 0.07

Epsilon = 1,25 e +04

My simulation diverge always

Please I am waiting for your Answers

****** Is this possible ********

gmg · June 2, 2009, 13:00

Hi,
can I get your one which you are doing.. because I am working on k-e model.

thanks

gmg

gmg · June 2, 2009, 13:02

my e-mail address is mustafa_ruet@yahoo.com.

thanks.

SKK · June 5, 2009, 06:02

Try using smaller URF for k and e to start the case and then later on increase as the solution has run for a few iterations and stabilised. You could also initiate the solution with smaller value of k. General k-e models under-predict the 'spread' of k values from meanline in jet modeling.

June 1, 2009, 11:19	Star-CD and K-Epsilon turbulence model	#1
hamoudi Member Hamoudi djamel Join Date: Apr 2009 Posts: 50 Rep Power: 17	Hello Everybody, I have a simulation on a numerical jet of air For this I use the Star-CD with Turbulence Model K-epsilon as you know at the beginning of the simulation we need to initialize the values of K and epsilon ( Energy and Dissipation rate of turbulence ) to calculate these two parameters Energy K = 3 / 2 * (U * I) Velocity of Jet . U = 286 m / s I= Intensity = 5 % K= 360,73 m2/s2 Epsilon Epsilon = (cmu 3 / 4) * K (3 / 2) / L cmu = 0,09, L= Mixing Length = 0.07 Epsilon = 1,25 e +04 My simulation diverge always Please I am waiting for your Answers **** Is this possible ******

June 2, 2009, 13:00	Hi	#2
gmg New Member Join Date: May 2009 Posts: 3 Rep Power: 16	Hi, can I get your one which you are doing.. because I am working on k-e model. thanks gmg

June 2, 2009, 13:02	Hi	#3
gmg New Member Join Date: May 2009 Posts: 3 Rep Power: 16	my e-mail address is mustafa_ruet@yahoo.com. thanks.

June 5, 2009, 06:02		#4
SKK Member Join Date: Mar 2009 Posts: 55 Rep Power: 17	Try using smaller URF for k and e to start the case and then later on increase as the solution has run for a few iterations and stabilised. You could also initiate the solution with smaller value of k. General k-e models under-predict the 'spread' of k values from meanline in jet modeling.