Objective
The mission of this project was to make a low-cost, portable profiler that can be deployed on a submersible. The additional constraints are:
Unobtrusive, i.e. not affect the dynamics of the flow to be measured
Capable of measuring velocities from 0.1-10m/s
Measure velocities in a window on the order of 50mmx50mm
Plan
The solution I came up with to solve the problem was based on established research for profiling flow from a submersible, specifically An Autonomous Open-Ocean Stereoscopic PIV Profiler by Steinbuck et al., which used a 3 W, 532 nm diode-pumped solid-state laser. These lasers are typically large and expensive, so I instead opted for a 4 W 532 nm diode laser, which are typically used in hobby CNC laser cutters and are cheap (on the order of $100) and much more compact than other types.
I also forwent the beam-shaping optics and used a simple mirror and the laser’s natural dispersion to create a desirably-sized investigation window.
Instead of expensive high-speed cameras, I opted for more recently-developed, cheaper global shutter “Arducam” camera modules. To do the real-time image processing, I chose to use an NVIDIA Jetson which supports Arducam’s cameras to run OpenCV. The hope is that with recent advancements in technology, it might be possible to replace all of the expensive components of Steinbuck et al.’s approach to make a truly compact and cheap flow-profiler.
In accordance with UCSB laser safety guidelines, I had to build an enclosure with the proper safety features such as a key, emergency stop, and light seals. I also had to get the apparatus inspected by the safety officer. The enclosure would also have the benefit of making the experiment more robust and portable, and be a good frame for mounting electronics such as the cameras and computer.
Execution
Dry Ice Sanity Check
Results
Raw Stereo Output
Experimenting with Velocity Color Maps in OpenCV
Frame-Averaged Streak Velocimetry
Adding Sediment
Conclusion
The setup worked well. The highlights are:
Successfully visualized particles in the water—No tracer particles were added to the tank; all flow visualization was happening with just the ambient dust particles that collected in the tank over the experiment’s deployment. This implies the laser and camera setup will certainly be able to pick up ambient ocean water particles, even on clear days.
Desired velocity measurement range of 0-1m/s seems feasible based on video, though the velocities haven’t properly been analyzed yet, so it’s just an estimation. Higher velocities are still unclear.
Using the default Arducam lens, the image was good at a range of approximately 6 inches from the beam, the field of view around 100mm.
With further analysis, the limits of this apparatus will be made clear. The ability to visualize ambient dust particles more than surpassed my expectation. Overall, the project is very promising.