Among RANS and LES which one is the best for simulating transient flow in IC Engine.
Jubin,
The main point is that RANS models are less expensive. They simulate all the turbulence spectrum and gives time averaged mean value for velocity field. LES is based on filtering rather than averaging. It calculates only high frequency while computing the smaller ones (high frequency means small scales, which are supposed I somehow to be independent to the configuration of the flow).
Regards
Jubin,
If you look at the derivation of the RANS equations, you will notice that no care is taken when "smoothing out" the equations. This means that the entire spectrum of fluctuations is averaged out, with no preference to any modes or scales.
However, as mentioned by Mounir, the LES equations take care of the spectrum of fluctuations by adding a filter to the vector fields, thus removing certain fluctuating modes (small spatial and time scales) and keeping others in the equations. Hence its name: "Large Eddy Simulations"
Of course these explanations are not enough. Look at the derivations of these models and it'll become clear :)
Good luck.
LES and RANS are two different worlds! The latter is based on statistical averaging, leading to steady equations (eventually also in 2d). A sort of time contribution is in the URANS but it can be debated the real meaning of the temporal mean.
LES is based on a local filtering, the equations are unsteady and 3d. As a consequence, the meaning of the turbulence model as well as of the final solution is totally different from RANS.
I wrote some notes about https://www.researchgate.net/publication/313280320_Lecture_on_LES_-_Part_I_and_II
Dear Naseem Ali,
be careful, there is nothing less the Kolmogorov scale, as it is the characteristic scale where kinetic energy is terminated by the dissipation.
LES introduces a filter at a scale lying in the inertial subrange so the part of the spectrum to be modelled is much more extended than you think.
Hi Jubin,
The basic difference is RANS models all eddies while LES simulates large eddies and models small eddies.
The RANS equations were derived by taking a time average of the NS equations. The effect of turbulence is simulated through modelling the Reynolds stresses. LES is not a time average. The NS equations are solved on large scale to small scale to resolve the eddies in the turbulent flow. Only very small eddies are 'averaged' on sub-grid scale (smaller than 4 elements). An LES simulation is always unsteady.
Dear Filippo Maria Denaro
What is the relevance of the Kolmogorov scale for RANS models?
Abdul Rehman
actually no practical implication of the theoretical Kolmogorov lenght scale in the RANS setting.
But you can consider the wall resolved case, that is the case in which the boundary layer is fully described by the grid close to a wall. In such a case, you have to focus on the y+ distribution. You could consider that non-dimensional number as a ratio between the wall-distance and the Kolmogorov lenght evaluated in the region close to the wall.
@Naseem Ali, what are the eddies that are less than Kolmogorov scales?? I think kolmogorv scales are the smallest which dissipates into heat energy eventually.
There is no eddy less than kolomogov scale. Mr. Fillipo already answered it
- Badges
- Science topic
- Similar topics
- Filing