CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Turbulent model & NS (http://www.cfd-online.com/Forums/main/5407-turbulent-model-ns.html)

new one November 13, 2002 04:40

Turbulent model & NS
 
Principally NS equations describe both turbulent and laminar flows. Then why in codes additional turbulence model information is required?

Thanks,

new one

sheetal November 13, 2002 05:17

Re: Turbulent model & NS
 
The DNS of the turbulent flow is not possible with the current computing capabilities, since the realistic turbulent flow consist of large range of turbulent length and time scales. So it is necessary to reduce the complexity of the model. So RANS are introduced. RANS represents transportation equations for mean flow quantities instead of resolving turbulent fluctuations.

regards, Sheetal

mukhopadhyay November 13, 2002 07:05

Re: Turbulent model & NS
 
eqn of motion (popularly called ns eqn) describes flow in terms of the instantaneous velocity in a flow situation, the instantaneous vel is the combination of the time averaged component (mean flow) and the fluctuating component. thru reynold's decompostion (rans) one addresses the first one while the fluctuating component is addressed thru various turbulence propositions. a good understanding may be developed if you look at the book by anderson, tannehill and pletcher. dns is also possible and available now for many situations - formulation quite difficult and computation quite expensive although.

Holidays November 13, 2002 08:59

Re: Turbulent model & NS
 
DNS is only possible at small scale and low Re number, hence the need for LES or RANS version of the N-S eqns.


All times are GMT -4. The time now is 18:23.