Hi Ramos, I am not sure if due to your strong current velocities, the vibrations can be neglected. Note that the ADV are quite sensitive and therefore if vibrations occur in a flow which is already very turbulent, there will be a huge variation in the measured values of the ADV.

Take a look on SNR values make sure they are above 11, and check your std deviations just to make sure the measurements are not varying a lot.

I believe that the best solution will be to strengthen the support structure of the ADV. If you are dealing with such a strong current, it leads me to think that the cross section of your flume is quite wide. If that is the case, including a more intrusive structure may not have such a bad effect and may allow to correct the vibrations.

In the meanwhile maybe you would like to check the "Users guide for physical modelling and Experimentation", by Froestick L. (Hydralab Network). For scour tests it is also common to assume that there is no huge effect of the flow contraction, if the occupied section by an ADV a Pile or a Probe is lower that 1/6 of the useful cross section (Whitehouse R. - 1998 - scour at marine structures - this is valid for currents as well).

Bottom of the line - go for a stronger support for the ADV,

Other idea - if you find that the vibration induces to a systematic error in the velocities, you might be able to correct the measurements once they are in txt file. But I am not really sure if you can find a systematic error. Note that the turbulence of your flow might have its proper randomness. 

All the best,

Tiago Ferradosa

Hello Ferradosa,

firstly, thanks for your valuable considerations.

The SNR values are higher than 25 dB. I focus my attention on this parameter since the water I was dealing was so clean (because of a previous filtering process) that it couldn't reflect the signal. So we injected small particles of clay in the water and in some points I have a  really good SNR (30 dB). So, this parameter is not a problem.

As you know by your current research, introducing an object in a flow, such a cylindrical pillar, can considerably change the flow (please, check attached picture). Therefore, I am not sure if I can improve substantially the structure. But I can do a change in the pillar, not in the size, but in the shape. For instance, to adopt a teardrop shape for my pillar would reduce the vibrations. What do you think?

About the measurements, I am dealing with all the raw data and I don't find a clear pattern between the different parameters values and the vibration/shaking of the device. Although the correlation factor is clearly lower in the core of the jet (where the Vectrino vibrates).

Thank you for your insights,

Pedro Ramos 

Well I believe that a teardrop shape may generate lower vibrations. Are you able to change the place where the ADV is measuring the velocities? Sometimes researchers just want to measure velocities right in front of the pile and that might be quite hard to do. Sometimes I put my ADVs far way from the structure to minimize vibrations induced by the flow changes near the pile. Maybe that is not an option for you, because you might be more interested in realistic values near the jet or the pile (I am not sure how your setup is done - please forgive me if I am providing wrong guesses).

As I suspected you are not being able to detect a systematic error produced by systematic vibrations. Let me just make a note to your SNR values, you have great SNR values, good job! I often find researchers that don't pay attention to them, and therefore perform tests with non-realistic values of velocity and turbulence.

How are you placing your ADV? i.e. how is your support of the ADV done? In my last tests I was using a metal bar and some "clamps" (não eram bem clamps eram mais braçadeiras de plástico) which attached the ADV to the metal bar. I believe you will have to improve the stiffness of your ADV support. However, I am not really sure if that might have an effect on your scour tests or not. I will try to take some time to think about that problem and send you some sketches about possible ideas for ADV mounting device.

Good luck,

Tiago

I am measuring the velocities of a jet and not near a pile. I put an illustration of the flow around a pier in my later answer to illustrate the problem in using a cylindrical bar to hold the ADV device. I strongly believe that the shape of this bar can affect the quality of my results, among other factors.

The Strouhal number and the frequency of vibration are related with the diameter of the cylinder. Increase that diameter can lead to a stronger downflow (see again my latest picture) and a vortex shedding downstream. Although the vortex to downstream doesn't affect my results, the downflow can do. So, instead of changing the diameter I will study a more hydrodynamic shape, such as the teardrop.

I also used the same "clamps", but I am thinking in putting the cable inside the bar. The problem is how to make the end of the probe stiff (the 4 beams). I am thinking in a solution.

I strongly suggest the use of some clay (bentonite, for instance) to increase the SNR. This parameter is important. If you have the chance to do it, I think it is worth your effort.

In your case, to study the scour, using an intrusive method is quite dedicated. Good luck.

Than your for your valuable input,

Pedro Ramos.

Guilherme,

thanks for your insights.

I had never thought about the spiral approach. It seems very logic to do it and I will try it. I will keep you informed about the results. However, it doesn't prevent the shaking of the 4 beams in the jet core, but I don't think there is a solution for that. Maybe I will have to take that in account in the post-processing or using other instrument :/

Also, I think in some cases the velocities go higher than 3.0 m/s, which can be a problem, since the range of the Vectrino for horizontal velovity is +/- 3.0 m/s.

Measuring such a strong jet with the Vectrino profiler seems to be a demanding task.