one or two receivers see low data quality metrics and odd histograms

We recently ran a lab experiment where we had students use a Vectrino+ to measure the velocity profile in an turbulent round jet. A few of the groups have had some oddities show up during the data collection that we are having a hard time understanding. In the attached velocity histogram you can see in the second and fourth velocity components (V and W2) secondary peaks show up (black line is the raw data). These have good SNR and correlation and a normal screening for low DQ indicators doesn't remove them, nor does an adaptive gaussian filter (the red and blue lines respectively). These measurements were made on the edge of the jet at x/D = 70 (i.e. well away from the orifice) with the nominal velocity range set at 2.5 m/s. What could cause the secondary peaks? It doesn't seem like phase wrapping would cause this and the measurements are of good quality, so we are at somewhat of a loss. Weak spots would also not seem to be an issue, since they are 3 and 10 cm from the wall for this velocity range, and the measurements were made 35-40 cm from the wall. Aside from the histograms, another oddity some of the data is that each receiver saw much different return signal strength and SNR. In some instances we saw Receivers 1-3 with very good DQ indicators, and Receiver 4 have very poor indicators (correlations around 30-40 %, SNR below 10). In the data set above, Receiver 4 actually have the highest SNR, approximately 10 dB above Receiver 1, but its data looks far worse. The instrument was set up with one pair of receivers oriented streamwise with the jet and the others oriented cross stream. Receiver 1 (assuming it's the red banded one) was streamwise, pointing towards the jet orifice with the probe head looking down. What is the convention for numbering the receivers? I haven't been able to locate this in the manual. Thanks!

Dear PJ, again I have to apologize -- March has been really hectic around here with conferences, seminars  and the holy Norwegian Easter vacation where everybody goes to taste the last snow in the mountains.  Anyway, we will try to do better.

Jets are troublesome and the least understood application for velocimeters. Our general sense is that the large spatial velocity gradients - and possibly also large variations in scattering - give rise to problems with the Doppler estimation technique we use.  My suggestions is to use the max. velocity range available to try too find out what is real and what are artifacts of the measurement process. Once you know, you can then decrease the velocity range again to try to improve the precision.  Sometimes it is also worth it to change the orientation of probe for the same reason.

Numbering conventions for down-looking probes are 1,3 in the x-direction in 2,4 in the y-direction.  In other words, 1,2,3,4 counterclockwise for a down-looking probe when seen from above.

Best regards, Atle Lohrmann