Research · Junior year
Moving water without a propeller.
Comparative Performance Analysis of Three Magnetohydrodynamic Propulsion Designs in Seawater for Thrust Efficiency and Acoustic Output
Seawater conducts electricity. Run a current through it while it sits in a strong magnetic field, and the two interact to produce a force — one that pushes the water out the back of a channel. No blades, no shafts, no moving parts. Just physics you can watch.
The Question
Ships move by spinning propellers — loud, mechanical, and full of moving parts. Magnetohydrodynamic propulsion skips all of that. Pass a current through seawater inside a magnetic field and the Lorentz force pushes the water itself. I wanted to know: how well does it actually work, and which design works best?
Designing the Experiment
I designed and 3D-printed three configurations in Fusion 360: a radial channel, a tight rectangular channel, and a semi-compact Halbach-array design. Each used N45 neodymium magnets and stainless steel electrodes, allowing me to compare how geometry and electrode spacing affected flow.
Building and Testing
I tested each design at 5, 10, 15, and 20 V in artificial seawater at approximately 35 PSU, repeating each condition three times: 36 MHD trials. I tracked dye displacement in 240 fps footage to measure exit velocity, and analyzed hydrophone recordings in Audacity. The propeller comparison used three recordings at 5 V. Electrolysis destabilized the current, so I pivoted from direct thrust measurement to exit velocity—not a measurement of thrust-to-power efficiency.
Year One · Honors Science Research
Before the thrusters, biochemistry.
The MHD work was my second year in Brunswick's selective, application-only Honors Science Research Program. In year one, I co-authored a biochemistry study of catechin as an inhibitor of PR-10 food allergens — running SDS-PAGE gel electrophoresis to test how the compound interacts with the proteins behind common food allergies.
Results
What the data showed.
Propulsion designs compared
Maximum exit velocity — Design 2 at 20 V
Lower average acoustic output in the tank comparison
Exit velocity at 20 V (cm/s)
Acoustic output vs. propeller
−33.6 dB on average
Across RMS (overall level), 3000 Hz, and 5000 Hz metrics, the MHD designs averaged 33.6 dB lower acoustic output than the propeller apparatus. These are tank-test comparisons, not a prediction of full-scale vessel noise.
Without rotating components, remaining sound came mainly from electrolytic bubbling and channel turbulence.
Radial · Design 1
15.00 cm/s at 20 V. Wide electrode spacing limited acceleration; electrode passivation increased resistance over repeated use.
Rectangular · Design 2
26.67 cm/s at 20 V—77.8% faster than Design 1. The narrow channel delivered the highest exit velocity, but trapped electrolytic bubbles.
Halbach · Design 3
17.45 cm/s at 20 V. The strongest power-law fit (R² = 0.9842) indicated a consistent velocity response, while wide electrode spacing limited current.
The limitation matters
There was no overall winning design. Electrolysis reduced electrode contact area and destabilized current; at 15 V and above, designs usually stopped working within about 10 seconds. Design 2's faster jet did not establish superior measured thrust or energy efficiency. A larger tank with floating prototypes and a reliable thrust-measurement method would be the next step.
Recognition & reflection
From theory to something observable.
NASA Earth System Science Award
UConn Physics Department Award
Connecticut Science & Engineering Fair — Finalist & 2x Special Award Winner in Physical Sciences
The equations told me it should work. Seeing water move on its own told me why I love doing experiments.