[emspinosa]

Hi everyone,

I am posting something related to Experimental Fluid Mechanics, PIV (Particle Image Velocimetry data) but I'm sure you could give me some advice anyway.

I am processing PIV data. I have a series of 2D vector fields in a plane normal to the wall in turbulent boundary layer (in other words the velocity profile plane). Each vector field is like a picture of the flow at a certain instant and all the fields are acquired say at certain sampling frequency, set by the PIV system synchronizer (2 Hz).

I have to compute the mean velocity profile and the turbulence intensity profile across the boundary layer. To do that, I extract the velocity components u and v along a wall-normal line with Tecplot and then I average over all the fields. For the turbulence intensity I calculate the squared difference betwen the instantaneous component value in each field and the averaged value and apply the usual formula for the rms.

In your opinion, does the sampling frequency of the acquisition (as I said, 2 Hz in my previous experiments) with PIV matter for a correct estimation of the rms or is it just the number of samples that matters the most??
I am just wondering whether with a low sampling frequency I lose the contribution of the fastest fluctuations which may give a significant contribution to the rms value.

What do you think? I think it's more or less the same as whan you're computing the rms from a set of many numerical vector fields.

Please let me know if you have understood my point and what you think about it.

Thanks a lot.

Emanuele