Human-Powered Submarine

The previous generation steering system used a rack and pinion machined from Delrin via wire-EDM. While functional in dry conditions, the system suffered from precision limitations inherent to the manufacturing process and was vulnerable to water ingress at the mesh interface — a critical failure point for a fully submersible, depressurized hull operating at depth.

For the current generation, we redesigned the steering system around a slider-crank mechanism. The pilot interfaces with a throttle-style input, converting rotational motion into linear displacement of the slider. This system runs parallel to the yaw and pitch fin rod assemblies, where the fin rods replace the throttle input in an equivalent kinematic arrangement.

Initial prototypes were 3D printed to validate the concept, but the material (PLA) proved insufficiently robust for the operating environment. We sourced components from McMaster-Carr to evaluate what could be purchased off-the-shelf versus fabricated in-house — concluding that the linear rail could be bought directly while the slider itself could be replicated. Anodized aluminum was selected as the primary material based on budget constraints and its proven corrosion resistance in marine environments. The primary risk is coating damage — any scratch exposes bare aluminum to saltwater — so pre-deployment inspection and protective spray coating are planned as standard maintenance procedures. 

To integrate the system into the hull, we coordinated with the hull team to request a flat mounting surface on the 3D-printed hull. Non-load-bearing mounting blocks were 3D printed for the rail ends to maintain level alignment throughout the assembly.

The objective of pilot testing was to evaluate whether pilots could generate sufficient steering force to operate the mechanical system under real conditions — 20 feet underwater, in low visibility, under potential stress.

We measured pilot steering strength using a weighted wrist curl protocol. Both pilots returned factors of safety of 8 and 11, respectively. While these values appear acceptable in isolation, they are insufficient given the operating environment — a panicked pilot underwater may exert forces far beyond their measured baseline, risking system failure at a critical moment.

During testing, we also identified a procedural error: pilots performed the exercise with the back of the hand parallel to the bench rather than perpendicular. These two positions engage fundamentally different muscle groups, meaning our measurements did not accurately reflect the actual steering motion. This finding invalidated the current test setup and informed a revised testing protocol for the next design iteration.

Both outcomes — the marginal safety factors and the procedural correction — are being carried forward into the next phase of steering system development.

Training new members is one of the most important — and most overlooked — parts of leading an engineering team. I knew early on that I wouldn't always be available, and that the team needed to be competent enough to find answers independently, not just follow instructions.

To make that possible, I developed a structured training program spanning CAD, airfoil theory, MATLAB, composite layup, tooling, and engineering design — delivered through hands-on sessions and a library of 10 instructional videos members could revisit on their own time. The goal wasn't to hand them everything. It was to give them enough of a foundation to go figure the rest out.

The result: our sub-team members have gained the fundamental skills and technical know-how to work independently. Furthermore, we have achieved the highest member retention rate on the team.

Featured Training Videos: