Equality in Quality ParametersUp to Velocimeters - single point
Equality in Quality ParametersPosted by davidciochetto at January 08. 2003
At one point, I read here or in a paper that the correlation should be equal or roughly equal in all three beams. However, I have forgotten where I saw that info to re-read it.
It seems geometrically reasonable that the correlation should be similar if not equal since two particles that are moving in a turbulent flow would have 100% correlation if they were both moving the same relative to a beam. If their velocity in the beam direction were equal but the other directions were out of phase, then only one beam would report good correlation and the others would report bad correlation. The data point would be rejected since the turbulence was of a smaller scale than the measurement volume or time. Is this correct? What other situations arise?
My real question is why should the correlation and amplitude in all three beams be equal for normal operation? If one is consistently higher, what is that an indication of?
Equality in Quality ParametersPosted by davidciochetto at January 09. 2003
There are several reasons why the quality parameters (SNR and correlation) should be the same for all three beams.
1) It is an indication of production quality. For example, if one beam has weaker amplitude it may mean that the efficiency of this receive element is poorer, which again means that the system is not as good as it can get.
2) When you calculate u'w' and you see how this parameter relates to the original along-beam velocities Vi, you will see that there are terms like c*(V1'^2 - c*(V2'^2 (differences between variances). Since the noise variance is a strong function of the correlation, these terms will not disappear unless the system is well matched. The practical consequence is that the ability to measure small shear values is reduced.
The correlation is also a function of the shape of the curve that you find in the probe check program, especially if you operate with a high velocity ranges. Basically, the more 'peaky' the shape, the higher the correlation. This is something that will vary somewhat from system to system but not something that we can easily influence.
The correlation will also vary with the velocity direction such that the beam pointing in the main velocity direction usually has the lowest correlation.
The main reason for differences in the correlation is otherwise the amplitude. As a consequence, the Vector generally has higher correlation at higher SNR.
- Atle Lohrmann
Re: Equality in Quality ParametersPosted by Ayal Anis at March 29. 2012
Fine Nortek Folks,
I am looking at a dataset collected in a small lake from a moored, upward looking vector (see attached log and hdr files).
During the whole time pitch and roll were relatively small, with values less than 3 degrees for each.
Attached is a plot showing the burst-averaged values of several parameters (1024 points per pseudo-burst,
since data was sampled continuously).
My question is: is this normal for the correlations to differ from each other by as much as seen in the figure?
Beam3 seems to be "better" than beams 2 and 1, with the latter being the worst regarding correlation.
Re: Equality in Quality ParametersPosted by P.J. Rusello at March 30. 2012
It looks like Beams 1 and 2 we're having issues from either a wake only affecting them (or debris caught on the frame or instrument). Were there any boundaries close by? A small tilt could mean those two beams were having issues with pulse interference that Beam 3 wasn't subject to because of a different acoustic path.
Just to make sure it's not an instrument issue, try doing a probe check in a bucket in the lab and also collecting velocity data. Ideally all three beams will look the same in all parameters, but it's not unknown for one or two beams to have some problems.