[ari003]

Hello everyone, I m Ari currently doing my thesis related to various type of Turbulent inlet method suitable for wind energy application and I'm in the search of various methods like Digital Filtering, Synthetic Eddy Method(SEM) etc . I'm using OpenFoam and Python for my project. I have the datas of 5 met mast values which is arranged linearly at different height (pic attached) and I need to generate the velocities at all the point or cell along that plane which is considered as my inlet plane.

I need some assistance regarding how can I arrange the velocity data based on lets say SEM on the inlet plane at different time intervals keeping these input datas as a constraint. As of now I was able to do it with KSEC method (Kaimal Spectrum with Exponential Coherence) only and need to implement with other Synthetic Turbulence method. Any kind of suggestion will be of great progress towards my thesis work. I'll remain hopeful to get a response from you.

[sbaffini]

I can't answer for OF related issues (note, however, that there is a lot of material on the OF implementation, even a dedicated python package if I remember correctly), so I will only give you a general guidance on how these methods work.

Typically, they are built on top of your average flow inlet, adding spatio/temporal fluctuations to it, according to a certain spectrum and given reynolds stress.

So, in your case, you first need to somehow extend your profile (which I assume is related to the average inlet, otherwise you have to average it yourself at each point) to the whole inlet plane. That is, you have inlet along a single line but you need it for a plane. As this is a wind energy application, I expect that your profile is constant in the transversal direction, so you can just replicate it along that direction (according to the specific implementation details, you might even skip this, but I don't know for OF).

But, this is typically not sufficient to prescribe an LES inlet. All the methods I know of also require a specification for, at least, two turbulence quantities (say, k and epsilon, for the vortex method) or, more commonly, the whole average Reynolds stress tensor. You ideally need it in the same location where you know the inlet profile (in your case, again, I expect that knowing it in the points of your profile is enough, as every averaged quantity should be constant in the transverse direction).

So, in conclusion, I suggest to first investigate how OF handles this stuff and the availability of anything that could help (like the package I mentioned, whose name I can't recall now). Then adapt your input to it, taking in mind that you might need more than you currently have now.

[sbaffini]

Here it is the package I was referring to:

https://www.sciencedirect.com/scienc...52711018300487

https://eddylicious.readthedocs.io/en/latest/

https://github.com/timofeymukha/eddylicious

But I have no idea how/if it works

[ari003]

Originally Posted by sbaffini

Here it is the package I was referring to:

https://www.sciencedirect.com/scienc...52711018300487

https://eddylicious.readthedocs.io/en/latest/

https://github.com/timofeymukha/eddylicious

But I have no idea how/if it works

Thanks for your response. I'm aware of this Eddylicious unfortunately it only creates precursor inlet methods whereas I m looking for synthetic one.