The Toroidal Propeller

Drones could transform wildlife research, except their own propeller noise confounds the observations and stresses the animals being studied — one cited study measured a 300% jump in black bear heart rate after a flyover. The noise is dominated by the propellers, and it lands squarely in the 1–10 kHz band where mammals hear best.

Looping each blade back on itself into a closed toroid reshapes the tip vortices that cause most of the noise — that much was MIT Lincoln Laboratory's contribution. What nobody had published was which of the shape's parameters actually matter. My own addition, the leading blade segment height offset, raises where the loop's leading segment begins so the two segments no longer lie in one plane: it reaches the same 2.5 N thrust at lower RPM, and lower RPM is quieter.

Eight geometric parameters, isolated one at a time. Geometry parameterized in SolidWorks and screened in SolidWorks CFD (RANS, k-ω) to eliminate hopeless designs cheaply, then the survivors 3D-printed and measured on a thrust, torque and acoustic stand I designed and built, inside an anechoic chamber at the Alfred Gessow Rotorcraft Center — with two-tailed t-tests against a commercial control, and a mesh-convergence study that put the simulation within 1.1% of published thrust data.

4.4 dBA quieter than a conventional propeller, with a 40.7% reduction in mean sound pressure across 1–10 kHz — 1.9 dBA of that from the LBSHO alone. Figure of merit came out roughly 10% higher, though the paper flags that the efficiency gain still needs validating in injection-moulded nylon; the acoustic result carries no such caveat, because printed and moulded controls measured statistically identical sound. Patent pending (PCT) — the paper is written, and we're holding its submission until the patent clears.