Exploring a childhood fascination with clocks, an ingenious mechanical invention that uses the mass dependent period property of pendulums to regulate the accelerating drop of a weight, I designed and manufactured a fully functioning 24 hour mechanical clock. I conceived the idea prior to a product design course, but the course provided the opportunity to fully realize and test my designs. In addition to a tight deadline, the class project imposed certain restrictions like the necessity to mount the mechanism on a 15x15 board of duron and manufacture a certain amount of parts out of wood and plastic.

The project started with studying resources on different clock mechanisms, with a primary focus on the escapement mechanism, the heart of a watch movement. The escapement mechanism is reliant on the period equation of a pendulum: period = 2π√(length * g), (note length is a reference to the distance from the axis of rotation to the center of mass of the pendulum). Notably absent from this equation is the actual distance that the pendulum is pulled back, meaning that the only variable that changes the time of one pendulum swing is the center of mass, which can be controlled by an adjustable set of weights set on the pendulum. From here, I derived my equations of motion for the optimal length to center mass of a pendulum, and started building a model in CAD. After experimentation with different geometries and materials of varifying coeffecients of friction, I settled on this 3d printed geometry using ABS plastic that I then finely sanded:

From here, I had to design gear reductions, which although simple, took multiple prototypes to get right, considering the high friction between wooden gears, and the tendency to seize up if they weren't meshing at the optimal distance. I then had to cut the frame from duron, and find the exact location for all 6 shafts to be mounted. After I had found all my final dimensions, I arrived at my final CAD:

With a some patience for final prints, laser cut, a finicky assembly, and some tuning of the weights on my escapement mechanism, I finally arrived at a functioning clock, which tetsed to an accuracy of ±30 seconds per hour. It could run for an hour with the weight set from a height of about 8 feet, though the running time is constrained by which gear the weight is mounted on. Heavier weights mounted on higher torque gears (those closer to the hour hands) can run much longer, as each spin of the weight spool will result in that many rotations of the particular gear. However, my duron assembly and wooden gears limited me to putting the weight on my seconds hand, meaning each full rotation, or minute, would unravel my weight spool by one rotation. If I remake this assembly out of full plastic or metal parts, I could increase the time between windings and the accuracy. All in all, despite the numerous hours of reworking tolerances and prototyping, this project immensely improved my skills with design of functioning assemblies, and was an immense pleasure to finally make work.