Rotational flow turbulence model

ntrask · February 2, 2009, 13:16

Hello all,

Can anyone please tell me whether one of the RANS turbulence models takes into account the effects of rotation? I'm looking to apply this to a high Reynolds number, swirling jet.

Alternatively, if anybody could point me toward a reference/thread describing what regimes the included turbulence models are built for I would be very appreciative. I've had limited success searching the forums and reading the code.

-Nat

gwierink · February 3, 2009, 05:30

Hi Nat,

The RANS k-epsilon model is not the best choice for flows with strong rotation, shear, and/or confined flows. This is due to the way of averaging and assumptions on the isotropy of turbulence. It is not a shortcoming of OpenFOAM, but it is because of the turbulence modelling. In fact, in OpenFOAM you have the possibility to change the solver so that it will fit your case

. Anyway, if you have the possibility I would try one of the LES solvers for your problem. You can find details on k-epsilon on many websites, e.g. this one: http://www.google.fi/url?sa=t&source=web&ct=res&cd=2&url=http%3A%2F%2Fc fd.me.umi st.ac.uk%2Ftmcfd%2Fpdfs%2Ffile_pdf-1023456176.pdf&ei=kguISamTLZiq-gaA0v3VBw&usg= AFQjCNHUcHvKTzgxxxyMj89mEtm4vMIREw&sig2=Kw7XRAIrXj REhEqWAmGZlw. The OpenFOAM user guide lists some general guidelines for the solvers on p. U-83 and onwards and there is a solver tutorial in §3.2.4 on p. U-74.

Regards, Gijsbert

billy · February 3, 2009, 05:55

Maybe a Reynolds stress turbulence model or DNS (just kidding).

ntrask · February 3, 2009, 11:59

Thanks for the quick replies. Unfortunately, my turbulence model is also coupled to a subgrid atomization model based off the Reynold's decomposition assumptions, so an LES won't work (I need the turbulent kinetic energy, not just the sub-grid kinetic energy). I know that the stand kEpsilon models included in Foam won't work for anisotropic flows, but when I look into the other models (especially the ones included in rasInterfoam) they all seem to be tailored to low Re number flows.

I'd even be interested in implementing a turbulence model and submitting it to the next release, if someone can suggest a model not currently in the code.

February 2, 2009, 13:16	Hello all, Can anyone pleas	#1
ntrask New Member Nat Trask Join Date: Mar 2009 Location: Providence, RI Posts: 17 Rep Power: 17	Hello all, Can anyone please tell me whether one of the RANS turbulence models takes into account the effects of rotation? I'm looking to apply this to a high Reynolds number, swirling jet. Alternatively, if anybody could point me toward a reference/thread describing what regimes the included turbulence models are built for I would be very appreciative. I've had limited success searching the forums and reading the code. -Nat

February 3, 2009, 05:30	Hi Nat, The RANS k-epsilon	#2
gwierink Senior Member Gijsbert Wierink Join Date: Mar 2009 Posts: 383 Rep Power: 18	Hi Nat, The RANS k-epsilon model is not the best choice for flows with strong rotation, shear, and/or confined flows. This is due to the way of averaging and assumptions on the isotropy of turbulence. It is not a shortcoming of OpenFOAM, but it is because of the turbulence modelling. In fact, in OpenFOAM you have the possibility to change the solver so that it will fit your case . Anyway, if you have the possibility I would try one of the LES solvers for your problem. You can find details on k-epsilon on many websites, e.g. this one: http://www.google.fi/url?sa=t&source=web&ct=res&cd=2&url=http%3A%2F%2Fc fd.me.umi st.ac.uk%2Ftmcfd%2Fpdfs%2Ffile_pdf-1023456176.pdf&ei=kguISamTLZiq-gaA0v3VBw&usg= AFQjCNHUcHvKTzgxxxyMj89mEtm4vMIREw&sig2=Kw7XRAIrXj REhEqWAmGZlw. The OpenFOAM user guide lists some general guidelines for the solvers on p. U-83 and onwards and there is a solver tutorial in §3.2.4 on p. U-74. Regards, Gijsbert __________________ Regards, Gijs

February 3, 2009, 05:55	Maybe a Reynolds stress turbul	#3
billy Senior Member Billy Join Date: Mar 2009 Posts: 167 Rep Power: 17	Maybe a Reynolds stress turbulence model or DNS (just kidding).

February 3, 2009, 11:59	Thanks for the quick replies.	#4
ntrask New Member Nat Trask Join Date: Mar 2009 Location: Providence, RI Posts: 17 Rep Power: 17	Thanks for the quick replies. Unfortunately, my turbulence model is also coupled to a subgrid atomization model based off the Reynold's decomposition assumptions, so an LES won't work (I need the turbulent kinetic energy, not just the sub-grid kinetic energy). I know that the stand kEpsilon models included in Foam won't work for anisotropic flows, but when I look into the other models (especially the ones included in rasInterfoam) they all seem to be tailored to low Re number flows. I'd even be interested in implementing a turbulence model and submitting it to the next release, if someone can suggest a model not currently in the code.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Eul-Eul flow, k-e-kp-ep-Theta Turbulence model	us	FLUENT	5	April 5, 2011 03:29
Turbulence model for Laminar flow	Terry	Main CFD Forum	8	August 14, 2010 06:45
validating turbulence model on flow around a car	Pedro	CFX	1	February 20, 2008 17:32
Compute Torque on Rotational Model	rachid	FLUENT	4	December 1, 2005 11:51
Algebraic Turbulence Model for 3D flow	Dmitriy Makarov	Main CFD Forum	0	June 28, 1999 02:40