I want to combine my LES (3D time resolved field) with experimental data (2D planar distributions) in order to improve my results. The flow field is highly unsteady. Data assimilation can be applied but probably would not be perfect or even worse. Are there any similar studies?
The experimental data are averaged in time and along the third direction or you have just the two components in the plane?
Dear Professor,
I really appreciate your response. The experimental data is pressure distributions extracted by PSP (attached paper). Probably, I do need to perform Stereo PIV to have flow field data as well as pressure distributions on the surface. In this way, data assimilation might have a good result.
Article Structure analysis of adiabatic film cooling effectiveness i...
yes, I think that you should consider the results obtained by PIV, then you have to do the correct statistical analysis
Is there any other similar mathematical technique to merge numerical data with experimental?
The key is to make congruent the two data set... if you have a LES solution that means you have 3D and unsteady fields and the solution provides the functions in the sense of locally filtered data. Thus, your experimental data and LES solution must match. From LES solution you can perform ensemble averaging and extract the statistics you need. The meaning of the experimental data depends on the type of experimental devices.
Data assimilation is not a specific mathematical technique. It combines many kinds of algorithms (variational, statistical) which merge measurement data with observations. The question is how to do it in in your case. Do you solve compressible NS equations or Boussinesq? If compressible, I would try to use your measured planar pressure distribution as BC in equations or to use data assimilation for modification of boundary conditions at this plane with using measurements.
Professor Filippo Maria Denaro,
I really appreciate your response. The problem is that LES was unable to agree with my experimental data in the near field (x/D4).The same discrepancy was captured by other similar LES/DES studies. I want to reconstruct my LED domain. Fortunately, spatial and temporal quantities on the surface were resolved using fast response PSP measurement technique. Maybe reconstruction would lead me to integrate both numerical and experimental such as attached link.
http://www.aero.jaxa.jp/eng/research/basic/windtunnel/efd-cfd/
I cannot tell you if the problem is in the LES solution or in the experimental devices. Often, the results do not match simply because you are not reproducing the same conditions of the experiments. As you wrote that the results match after x/D>4 I can suppose you provides an inflow condition in the LES that does not match exactly the experimental condition so that you need some diameters to let the solution better correlate. But I cannot say more without more details about the problems, the LES formulation and the experimental data.
- Badges
- Science topic