Final Design

This section describes the final design changes the team made to the Jacobs Hall Clock. In a desire to preserve history, most changes are replacements of worn or damaged components. Some improvements have been made to increase the life-span of the clock. 

Step 1) Baseline Testing

To understand how each change effected the clock's performance, a testbed was created. This testbed is comprised of a frame, four sensors, and two MATLAB scripts. 

Frame

The purpose of this frame is to raise the gearbox by at least two inches to allow the pendulum to swing freely, and to secure the gearbox to the table to prevent it from tipping over. The frame is constructed from 80/20 aluminum beams, corner brackets, and 2x4 wood planks (see 1.1.A-B). C-clamps are used to secure the gearbox to the frame, and the frame to the table. 

Accelerometer

This first sensor, MPU6050 3-Axis Accelerometer, is attached to the disk of the pendulum to track its angle as it oscillates. This accelerometer allows for the tracking out-of-plane motion and gradual losses to friction. The accelerometer has been directly soldered to wires that run along the pendulum shaft to an Arduino Uno on the top plate (shown in 1.2.B-C). This configuration minimizes the amount of force that opposes the pendulum's motion, since the wires are attached very closely to the pivot point. 

IR Break Beam

The second sensor of the test bed is an optical sensor to track the ticking of the escapement wheel. This enables us to define a clear start and end time to the clock's active duration, as well as the ticking speed. This sensor uses an emitter to discharge an infrared beam to the receiving end. When this beam is broken, a tick is counted and a timer until the next tick is restarts (see 1.3.A). It was necessary to align the beam in the middle of a tooth's oscillation range. If slightly outside of this range, recoil within the escapement wheel would incorrectly mark double ticks, as shown in 1.3.B. 

Camera

Using a camera to record the pendulum as it swings, a video annotating software, such as Kinovea, can be used to track the speed and angle of oscillation. Also, video recordings can be referenced to understand outliers in data collected from the other sensors. This camera was also used during the YouTube livestreams to check on the clock's performance remotely. 

Data Collection and Processing

Two MATLAB scripts were developed to efficiently collect and process data from our four sensors. The first script, named DataLogger, opens a GUI interface which displays three graphs that update live with data from the temperature, accelerometer, and IR break beam sensors. A camera preview is also displayed at the bottom right, with an option to choose if the video should be recorded along with sensor data. Three buttons within the GUI allow us to start, pause, and reset the data collection. The fourth and final button allows us to download all three of the sensor's data into one combined csv file for processing later on. An inactivity timer is displayed below the save button, which resets after each tick. If no tick is detected for sixty seconds, data collection automatically stops and an email is sent to notify the team. 

The second script, named DataProcessing, uses the csv file created from the previous script to analyze data and create relevant figures. Since multiple tests are run for each change, the input for this script is a folder which contains any amount of csv files. Each file within the folder is processed to find the temperature range, tick accuracy, total runtime, x y z angle ranges, and correlation between temperature and tick speed individually. Next, a summary section is displayed which contains the mean runtime, standard deviation, longest runtime, and shortest runtime. Five figures are created with three subplots each, for the tests with the three longest runtimes. The figures include a temperature graph, tick graph, and a graph for each axes angle. 

Results from the baseline tests are shown below in 1.7.A-F. 

Step 2) Suspension Spring

Beginning with the most accessible component, the team decided that replacing the suspension spring would be the highest priority. This component supports the pendulum weight, while minimizing friction and encouraging single-plane motion. This component was the most clearly damaged part, and easily replaceable (see 2.2.A). With support from Tom Chalfant of the MAE Machine Shop, the team cut out a new piece of steel 0.25 in wide by 3.75 in long and a thickness of x in (see 2.1.A and 2.2.B). Following this change, a major improvement in out-of-plane motion and tick accuracy was observed. 

Step 4) Full Cleaning and Bearings

With this newfound confidence, the team proceeded to disassemble the clock to deep clean and replace harder to reach components. Of these components are the bearings. A total of 13 bearings were replaced across six shafts within the gearbox. It was clear that the old bearings had been over-greased and contaminated with dirt and other debris from improper sealing of the enclosure. This, along with being an aesthetic issue, is a major cause of friction within the system that needed to be addressed. Shafts 1-4, V, and D all required replacement bearings (see Item List). All bearings were replaced with stainless steel shieled ABEC-5 (medium-high precision) with the exception of Shaft 1 which received stainless steel ABEC-7 (high precision) bearings. To properly fit the new bearings, Stephen Porter of the Structural Engineering Machine Shop has remanufactured the spine of the gearbox, to eliminate the need for bushings and improve ease of maintenance (see 4.2.B). The original spine had oversized holes, and bushing were used to compensate for this oversight. Cleaning was achieved using light degreaser, cleaning cloths, soft-bristle brushes, and steel wool to bring back the beauty of the mechanism (4.1.A-D). The brass bob at the end of the pendulum received special treatment of sanding and polishing to truly bring out the original shine. 

Step 5) Escapement Wheel and Pallets

The original brass escapement wheel and steel pallets were also replaced due to significant wear and mechanical degradation. Over years of operation, the pallets had developed deep chipping from repeated contact with the escapement teeth (5.1.C), resulting in increased friction and inconsistent impulse delivery to the pendulum. The brass escapement wheel, while less visibly damaged, exhibited wear patters along the tooth flanks, contributing to poor engagement and timing irregularities (5.1.A). 

To ensure precise reproduction of these critical components, the Stephen Porter employed a digital probing process to capture the existing profile of the components. This data was imported into a CAD software where the profiles were refined to correct for material loss and asymmetry, ensuring that the new components would mesh perfectly. The finalized CAD profiles were used to manufacture the replacement escapement wheel and pallets. The escapement wheel was waterjet cut from a brass sheet, while the pallets were waterjet cut from steel and subsequently heat-treated to increase surface hardness and wear resistance (5.1.B/D).

Step 6) Differential

The final improvement before reassembling the clock was to replace the differential. The differential acts as a disengagement system, allowing the clock to still tick during the raising of the weight. Using the sister differential held as a souvenir by one of the original clock creators, Steve Porter, the old gears on and around the old differential were transferred over. To do this, gears were pressed off of the differential gears and a new hole was drilled on the shaft to secure the worm driven gear.  Steve Porter was given the used differential, which had much more character. 

New Documentation

A key deliverable for this project was to create new documentation for ease of future refurbishment or improvement. To accomplish this, the team created two important documents; an Item List and User Manual. The User Manual comprehensively outlines the gearbox’s operational principles, with clear explanations of how the pendulum, escapement, differential, and winding mechanisms function. Additionally, it specifies essential maintenance procedures such as regular cleaning intervals, lubrication schedules, and inspection guidelines to promptly identify wear or damage. The manual also lists all necessary tools required for routine maintenance and disassembly, including specialized equipment for precise tasks such as bearing replacements. A shortened version of the Item List is included in the manual, with the full version available online to include CAD files (created by TRI) and links to recommended suppliers. Furthermore, the manual addresses potential troubleshooting scenarios, offering solutions to common issues that may arise during the clock’s operation based on obstacles the team has encountered. 

Enclosure Proposal

To ensure long-term protection of the Jacobs Hall Clock and its internal mechanism, a comprehensive refurbrishment of the existing enclosure is proposed. The enclosure's steel frame is currently affected by rust and surface degradation from decades of weather exposure near the coastline. To resolve this, the entire steel frame should be removed and transported to a qualified metal fabrication facility, where it will undergo sandblasting to remove corossion. Following surface preparation, the steel should be hot-dip galvanized to provide high corrosioin resistance. Once galvanized, the frame should be power-coated using an outdoor-grade finish to ensure long-lasting protection against moisture and sun damage. To begin this process, a local metal fabrication studio has visited the clock and a quote is passed on to TRI to complete during the following summer. 

Additionally, the aluminum window frames must be replaced due to misalignment and seal failure. New aluminum framing with powder-coated finishes should be installed. high-quality weatherproof gaskets and rubberized seals must be applied around each window panel to prevent dust and moisture infiltration. All seams between frame segments should be sealed with industrial-grade silicone caulking that remains flexible and UV-stable over time to account for thermal expansion or contraction. 

This restoration process, if executed properly, will not only protect the clock from environmental wear for over 30 years but will also enhance its appearance and reduce long-term maintenance costs. 

Final Results

The final performance statistics of the restored Jacobs Hall Clock gearbox demonstrates a strong return to operational performance. After the full cleaning and replacement of critical components, the clock mechanism achieved a verified runtime of 19+ hours, with limited testing. Tick detection using the IR break beam sensor shows tick accuracy exceeding 90%, a dramatic improvement form the initial condition below 50%. These results confirm that the restoration successfully addressed the most severe sources of friction and mechanism degradation. The low minimum runtime and high standard deviation are most likely caused by the precise configuration of the pallets that are still to be finalized. Inspection revealed that the escapement wheel is slightly off-center on its bracket and shaft, creating a high point at each revolution. This introduces small variations in pallet contact due to inconsistencies in relative distance. Despite these final adjustments that need to be resolved, the gearbox is in a highly functional state and ready to be reinstalled into the professionally refurbished enclosure once completed. The project demonstrates both significant mechanism improvements and long-term sustainability, ensuring that the Jacobs Hall Clock can resume operation as a fully restored engineering landmark.