[reespanau]

Hello everyone,

I'm working on a detached eddy simulation model to analyze river turbulence and I want to investigate:


1) How to quantify the loss of information from a benchmark LES model simulation as a result of implementing coarser grids and how to validate the accuracy of the representation of turbulence structures?
2) How do main hydraulic features, primary, and secondary flows, shear layers, main and secondary eddies manifest under different domain resolutions (from coarse grids to finner)?

I have tried to perform temporal analysis, taking different points in the domain and recording each point's velocity values over time. Doing this, I plotted the energy spectra for each of these points "individually" (i.e. FFT analysis of the velocity signal for each point). This approach allows me to compute a two-point correlation of the velocity fluctuations, which could be an indicator for characteristic length scales of turbulent structures (i.e. eddy sizes that your coarse grid could resolve) and compare them with highly-resolved LES benchmark. I want to know if there is another way to investigate these questions, maybe other approaches or ideas?.