Vecrino - doppler noise

Hello everyone,

I want to measure rms velocities and  turbulences with a Vectrino in a lobor flume. Therfor I have to estimate the velocity component of the doppler noise

(short term error). I have read that the doppler noise is a function of signal strength and flow contition.

The manual describes a doppler noise about 1% of velocity range at 25 Hz. My configuration is 200 Hz and velocity range 1 m/s.

Is it possible to estimate it for a special Vectrino configuration?

Thanks!

Markus

Sorry for mistake! Vectrino - doppler noise

Dear Markus, there are ways to estimate the noise fluctuations in a Vectrino but is really quite complicated and I am not quite sure if the predictive models are good enough to be used for practical purposes.  Here is what I do:

a) Estimate the frequency spectrum from the velocity time series
b) Find the section of the spectrum that is flat.  Approximate the average noise level in this portion of the spectrum
c) Integrate the noise level over the full frequency range o find the variance of the noise. The square root of this number is your measurement uncertainty.

If you cannot find a portion of the spectrum that is flat, the measurement uncertainty is too low to estimate - other than to find a upper bound using the minimal spectral level.

Best regards, Atle Lohrmann

PS There are several good Velocimeter papers that have been published over the last couple of years.  We will try to upload some of them to our Bibliography pages over the next couple of days.

Hello.

Thank you for information!

The variance of the doppler noise is the integral of the average noise level over the full frequency spectrum of the velocity? Is this right?

Must I estimate the doppler noise in every single breakpoint or is this a average value of the special Vectrino configuration?

How can I estimate the frequency spectrum of the velocity?

Best regards,

Markus Schmitt

Dear Markus

Yes, the variance of the Doppler noise is the integral of the average noise level over the full frequency spectrum of the velocity.

You can calculate the frequency spectrum in many different ways, depending on what kind of filtering scheme you want to use and also what software you have available.  Around here we mostly use the fft function in Matlab but we really are no experts.  I would recommend maybe looking at the online manual for the ExploreV program, it is very well documented. A second source could be some good books on time series analyses.

Best regards, Atle Lohrmann

I´m including as an attchment, a paper published in JHE ASCE regarding Acoustic doppler velocimeters and Doppler noise effects on turbulence parameters

Best regards,

C.Marcelo

I´m including as a new attachment a paper published in Experiments if Fluids. The section 6 of this paper is related to the Methodology used to compute the turbulence on the basis of velocity signals recorded by Acoustic Doppler velocimeters. This section also includes details of methods to remove Doppler noise effects in the computation of each turbulence parmeter

Best regards,

C.Marcelo