2D turbulent channel flow: pb with TKE results

[gina]

I simulate 2D turbulent channel flow with dimensions (x*y*z) = (100*1*1)m for the helf of my geometry (y is the height, and z is the third dimension). I take, ρ = 1 kg/m3, ν = 0.001695 kg/ms an u=18.4539 m/s to obtain Re_tau=590 and to compared my result to DNS data of Moser et al[1].

I use a periodic Bondary condition.

I can turn on my simulation with saturne and I have got a result BUT, the values of TKE is vey small (the order of 10 **- 2)compared to DNS and Fluent results!!!

I think that I have introduice correctly my data !

I do not know where is the problem ?

Could somebody please have a look at the xml file and let me know where I am making mistake. I just want to plot the turbulent knimatic profil

Thanks and Best Regards,

[1]Moser R.D., Kim J., Mansour N.N.
(1999) “Direct numerical simulation of turbulent channel flow up to Reτ = 590”, Phys. Fluids, Vol. 11, N° 4, 943-500.

[Mickael Hassanaly]

Dear Gina,
could you send me the paper you are referencing to?
Regards
Mickaël

[gina]

thank you Mickaël for your answer.

I review my results .. finally in saturn GUI profile , I made a small mistake , I took turbulent viscosity instead of turbulent energy ... but despite that , the value of TKE wich I get is small compared to the DNS ! Can you please check it I thank you very much

[Debashis Basu]

Hello,
Could you please tell me how you are calculating the TKE ? Is it based on the fluctuating u', v' and w' ?
I can then try to answer your question.
Regards,
Debashis

[Debashis Basu]

Hello,
Are you using Saturne version 2 or Saturne version 1.3.3
Regards,
Debashis

[gina]

thanks,
I use saturne version 2.0.0-rc2
i compared to k+=f(y+)
i calculate k+ for dns like this: k+=1/2*(u'*u'+v'*v'+w'*w')
and from saturne i have the data of k =f(y), then i convert my k to k+ (k+=k/friction velocity**2) and my y to y+ (y+=y*friction velocity/viscosity)

[gina]

Dear all,
Could somebody orient me! Perhaps I do a mistake that I cant see well...
you can found as attached my saturn result compared with DNS
Please i need your help

[Mickael Hassanaly]

Gina,
I'm taking a look to your case. Don't worry ;-)
Mickaël

[gina]

thank you dear Mickaël,
 

[Mickael Hassanaly]

Gina,
i think you attempt to reach a turbulent Reynolds of 590, is that correct? With you set up , this is not the case. More over you don't use the correct values for fluid properties (specially viscosity). I follow the papers you send me and i have same order of value on the turbulent energy (i didn't change the viscosity here). I send you my xml file in attachment.
Mickaël