Advice for hydraulic master thesis.


I’m a civil engineering student and I’ve always had a special interest for electronics, sensors, and arduinos.

I’m finishing my degree for which I have to develop a thesis. After some research and based on some articles, I proposed to my hydraulics Professor to monitor the solids transportation (at the bottom of a river) using a hydrophone.

This idea consists on measuring the intensity/quantity of sediments (like rocks and muds) passing through a certain spot. These measurements are made using a hydrophone put underneath a seel plate – when the sediments pass beneath this plate they create impulses or frequencies, which will then be analyzed.
What I would like to know is the best way to capture these frequencies through the hydrophone – mobile recorder, daw, arduino?
I intend to connect the hydrophone to the computer, either through the arduino or through a sound plaque, but I’m afraid these solutions are more prone to the appearance of noise.

The frequency range to analyze is between 10Hz and 30kHz.

Will it be a good solution to connect the hydrophone to the arduino?

Another issue in my mind is the handling of the data gathered – should I use excel, matlab or python?

Carlos Esteves

sort by: active | newest | oldest
Forget Arduino, its not the solution for real time recording at this kind of data rate. Just record onto a PC using "Audacity". Since you can record in Stereo, you might be able to profile the arrival of materials at the phones, by measuring signal phase.

There'a HUGE dataset to crush here. I suggest Matlab or Octave.

Sparticles4 years ago
I am no expert, but the A/D converter on an arduino only has 8 bit resolution right? Can't you directly attach your hydrophone to the best recorder you can find? (e.g. a laptop whith at least 16 bit resolution and high sampling rate.)
I mean, after all, you will be measuring small rolling dust particles, right?
I don't know for a real time analyse, but if you save your sound samples as .wav files, you can get the raw data out of it pretty easily with C or C++.
And you will probably use a FFT algorithm to get the spectrum. I have only used the fftw library in c++ once and it was pretty easy and straight forward. (But must be possible in python and matlab too.)

And in my limited personal experience Python > Matlab > ... > minesweeper > excel.
amagro (author) 4 years ago

Thanks for the answer,

I'm curious about the smoke detector, why do you refer this sensor? will work under water?

The hydrophone will work on a flume in a laboratory, so will only read instant data, maybe less than 30seg each time.

Carlos Esteves
I was only thinking to use a beam of light and a photoreceptor to measure the opacity of the liquid like a smoke detector sees but if you are looking for amount of solids suspended in water, don't you have to centrifuge or dry it out and measure? Conductivity tests might tell something.
caitlinsdad4 years ago
Just thinking out loud there an accurate sound profile or difference with muddy silty water passing by than an algae bloom or oil spill? Rocks floating through or debris would just be noise so I don't know how you can really extrapolate quantity or intensity. Maybe it would be better to sample by optical means like a smoke detector? Is this measuring device to be dropped in the water to sample over time or an instant read? Crunching the data would be your preference for what cranks out the better charts. Good luck.