OF2.1: turbulentInlet for LES, description of how the fluctuations are generated
 

Hi

Does anybody know how the turbulent fluctuations are generated when using the turbulentInlet as an inlet condition in a LES simulation in Openfoam v2.1 ? If anybody could point to a scientific publication where the method employed is described it would be very helpful for me.

thanks

From what I've been able to interpret from the source code in turbulentInletFvPatchField.H and turbulentInletFvPatchField.C the turbulent inlet generates random fluctuations, and is not based on any method that tries to create synthesized turbulence through fourier modes.

Has anybody implemented a synthesized turbulence generating inlet himself? If, are you willing to share the code? :)

If not, would it be hard to implement it? If anybody who is knowledgeable could give me some general start up guide lines how to implement for exemple the method proposed in Smirnov et al, "Random Flow Generation
Technique for Large Eddy Simulations and Particle-Dynamics Modeling", J. Fluids Engineering, 2001, Vol. 123

It would be helpful. I might give it a try...

Hi Jan,

There are lots of threads already on the subject - for example concerning re-mapping look for "de Villiers".
If you really want a turbulence synthesizer, there is one available in the code of LTT Rostock (http://www.kroegeronline.net/exchang...Extensions.tgz - call inflowGenerator). It's based on the transport of vortex to generate the turbulent fluctuations following a given scale and Reynolds tensor.
!! The inlet should be oriented such as the x-axis is pointing towards the inside of the domain - that's hard coded :s

Hope that it will help

Hi Frederic

Thank you for the reply. I will look into the code from Rostock. I also found the threads about the inlet mapping, this might be an option as well for me.

/Jan

hi Ian. I know a nice method from statisctical approach: you can read
Efficient Simulation of Multidimensionaol Random Fields
(Jin, Lutes and Sarkani, journal of Engineering Mechanics, oct. 1997, 1082-1089).
This method is able to generate correlated random fields with controlled spectral density.
Nevertheless, I am not sure this algorithm gives a divergence-free U field (I have to try that since several years...). I know "home-made" implementations of this algorithm, but I am not (yet ;)) a Foam programmer... Let me know if you find something in openFoam about this!

HI,
you can use random number generator to implement the turbulent inlet boundary condition. The random generator interval should be between (-10 % and 10 % of inlet velocity).

Hi,thank you for your sharing, but i cant open the link,http://www.kroegeronline.net/exchang...Extensions.tgz could you send it to me, thank you very much!!

my email box: 981588592@qq.com

hello,Mohsen.i cant download the code,could you send it to me. 981588592@qq.com
thank you very much!!!

regards,
wen

The link doesn't work. Could you please mail me the code at saeedur.rahman.imece@gmail.com ?

Uni Rostock's OpenFOAM extension is hosted in GitHub
Code: https://github.com/LEMOS-Rostock/LEMOS-2.3.x
Features: http://www.lemos.uni-rostock.de/en/d...nfoam-content/

Does anyone knows what is the paper that turbulentInlet boundary condition is based on? It sure looks like to be implemented from a paper.

Les
 

hello why the random generator interval should be between (-10% and 10%of inlet velociity

please why it should be between (-10% and 10% )

Hi, do you know if turbulentInlet uses a random seed, for example clock time? I seem to get one and the same simulation results when running it multiple times, so it seems to me the seed is not arbitrary, but hard coded. Could anyone confirm this?