Hi,

I would like to have some references of LES in one dimensional flows. Could anybody give me some references.

Q

It sounds strange talk about LES in one dimension since turbulence is an intrinsically tridimensional phenomena.

everything is stange if one talks in one dimension. So shall we not do any one dimensional simulaions and directly jump to 3-dimensinal cases.

Ofcourse 1-Dm simulations are needed to understand the simulations models and results.

Do u agree or not.

Q

No. As LES pointed out turbulence is an intrinsically three-dimensional process. This means that *all* simulations of turbulence need to be performed in 3D. If you don't agree then you need to spend some more time in a library.

I know man all that do u think all the people who are doing cfd in one and 2d dimension are completing missing the turbuence.

So according to you 2d and 1D simulations are wrong since turbulence is 3D.

1D and 2D simulations are done because it is simple to analyse and get results and it helps to understand the turbulence phenomena.

Right or wrong?

Wrong.

One and Two-dimensional calculations are never performed when the turbulence is simulated.

One and Two-dimensional calculations may be performed when the turbulence is modelled.

Turbulence is a three-dimensional phenomena. The basic physical mechanisms do not exist in one- or two-dimensions.

There is a very significant difference between simulating (LES, DNS etc) and modelling (RANS etc) turbulence.

Wrong turbulence can be 2D. Take care do not say something wrong. For exemple Oceanographic flow, atmospheric ...

And LES is a Spectral concept. Read Sagault or Lesieur if you not sure. And second Wrong : LES simulating only one part of the TURBULENCE and the second one is MODELLING.

1D calculations is a good way for understand a model, but you can't make science with it and no interpretation is possible. With 2D simulations interpretation can be possible, but dangerous.

I belevieve that, 1D and 2D simulations is a good way to unterstand numerical method etc. But 3D is necessary to make the real 3D phenomena.

Best Regards,

From Turbulence, Coherent Structures, Dynamical Systems and Symmetry by Holmes, Lumley and Berkooz (the only book I have to hand). Page 3.

'In fact, the non-linearity, rotationality and the dimensionality interact dynamically to feed the turbulence - hence, to suppose a realisation of the flow is two-dimensional also destroys the problem'.

Explain to me all the paper about mixing layer, or boundary layer, with 2D simulations. Is it stupid ?

I never say that 2D simulations is good to perform a great LES. But i say it's, can be, necessary for understand LES !

And second time, I say what you write a mistake when you say that turbulence is a only '3D phénomena'. It's wrong, and I give some examples.

Finally, I know your book and I totaly agree with the authors. , and I never say something different with us.

Best regards,

Phillipe,

I'm no expert but trying to understand your point. Are you saying that mixing layer, or boundary layer concepts are 2D effect. Or is it that they are 3 dimensional but the effect of the 3rd dimension is negligable and so can be effectively ignored ?

If that is the case does that then mean we can do a 2D LES simulation ? or should we do a 3D simulation and expect to see the lack of 3D effects ??

Stu

90% or more of DNS and LES configurations in the literature are not 3-D.

Reason --> resources

Question --> Had scientists better to do nothing and let this field of research wihtout any progress?

We can still study two-dimensional configurations - but we need to do three-dimensional simulations (as I think you will find most of your 90% did).

Question over your question:

Would those scientists have made more progress if they had used 3D approaches? (assuming they had resources to solve 3D problems)

Let me go a bit further:

Who have found out that turbulence is a three dimensional phenomena? Well, I guess, it was discovered by the same scientists which in the past were solving 2D problems - (maybe, they have noted there was something lacking in their experiments and simulations ;-)

Except some very special case (stratify mixing layer in the ocean ... ), mixing layer and boundary layer are a 3D turbulence phenomena. If you can ignore a dimension, it's not a 3D phenomena but a 2D ... Laminar regim (not turbulent by definition) are stricly 2D for example.

And finaly if you do a 3D simulation with LES model, necessary you have 3D phenomena. 2D LES Simulation is only to learn and to understand the effect of LES model on a simulation.

I have not a great memory, but I realy interresting if someone can give a reference of paper with 2D-LES. Really.

Best regards,