Mountain bike suspensions systems are set up and tuned based only on qualitative, ‘feeling’ based metrics. The MTB DAQ, sponsored by Dr. Joseph Mello with the Cal Poly Mechanical Engineering Department, is a data acquisition system we developed to improve suspension tuning with a data-driven approach. The system mounts to a bicycle frame and has sensors that measure vibration and travel speed to characterize the bike’s motion. While similar systems exist, they cost in the thousands of dollars range and are very specific to individual bike geometry. Our system was designed to be more affordable to the average rider as well as generalized to many bikes and easy to set up. From our preliminary field testing, we found promising results from data collected with our system at different suspension setups. This project was a team effort between our senior design project team which included Mechanical Engineer undergraduates: Ronan Shaffer, Theo Philliber, Max Ringrose and myself.
Before we jumped into the ideation of the Senior Design Project, our team had to identify our project's scope. This included researching similar products, defining stakeholder's needs and wants, creating a functional decomposition, and establishing engineering specifications that needed to be met. After we got the scope of the project defined, I took charge as our team's project manager. I used a Gantt chart to plan and track timelines of key tasks and goals, while delegated different task to each member of our team.
To begin our Senior Design project, our team initially ideated on creating housing elements for our accelerometers that were external from the main board. We needed to identify where on the mountain bike the accelerometers will be placed to measure data based on the metrics we outlined in our PDR Report. We identified the the potential locations to be on different parts of the frame, so we conceptualized with that crucial detail in mind. We decided on our final prototype design utilizing techniques such as pugh matrices and weight decision matrices.
Our final design for the accelerometer housing consisted of two parts, the accelerometer housing and a mounting platform. The housing would enable the accelerometer board to be screwed into, fixing itself within the housing. It would connect and disconnect to the mounting platform with a squeeze tab. The mounting platform included a slot for a strap and an angled base to provide a surface that can attach to the variable widths of the bike frame. After the initial proofs of the prototype, our team decided to make the housing of the accelerometer one piece to provide an easier and simpler user experience.
During the Winter Quarter, our team concluded the System Design of our Data Acquisition System. This included the Electrical Design of the PCB Boards, the code design used to interact with the gyroscope and accelerometers we chose to integrate into the system, and the housing designs used for the external accelerometers and the main board. Max and I took charge of designing the electrical parts of the board while Ronan and Theo took charge of the housing elements of the system. Because Max and I had limited experience with PCB Design at the time, we both had to research and learn what went into electrical design.
Our team's DAQ System would transfer accelerometer data over the HDMI cables via SPI to the microcontroller. The Gyroscope transferred the angular velocities to the microcontroller via I2C. The microcontroller would then store the data on an SD card. Our primary coding language was python, but the data from the SD card would be converted into a text file after processing it through our MATLAB program. The main DAQ Housing included LED Light to indicate whether it was on and recording, on standby, or off. An LED numeric display would display which recording trial the system wsa currently on.
Our team used an aluminum housing for our main DAQ Board, 3D Printed the accelerometer housings and ethernet cables for data transfer. We utilized the stencil we acquired after ordering our boards to be manufactured by JLC PCB to assemble the electrical components to the boards. We simulated a reflow oven using a hot plate and a heat gun. After fully assembling our MTB DAQ, Theo and I continued to debug and fix the code for both the accelerometer and gyroscope data transfer until we established a working DAQ system.
Our team started the last quarter of our Senior Design Project verifying if we could collect usable accelerometer and gyroscope data from our DAQ. We tested our accelerometers using a shake table in Cal Poly's Vibration Laboratory. After welding custom testing rigs for our accelerometers, our team attached our accelerometers to both a vertical fixture and a horizontal fixture to the shake table to ensure acceleration data was accurate on all axes. With our 3mm rubber pad design under the accelerometer housing, our team noticed a dampening of high frequency accelerations, however this did not affect our results as our team were more interested in the dynamics of the lower frequency accelerations.
After confirming usable data from the Vibrations Lab testing, our team was ready to take the DAQ into field testing. We decided to use one dedicated rider (Ronan Shaffer) and one trail to diminish external variables that could influence the data taken from riding on a trail. We changed the suspension tuning settings, such as the rebound variable and fork stiffness and plotted them against the transmissibility of the two accelerometers. Our goal was to reduce transmissibility as much as possible so the rider would have a smoother experience in general. Many of the test days resulted in unusable data due to bugs in the programming or initial test set ups. By the end of the quarter, our team was proud to say we collected usable acceleration data from our DAQ. Our initial scope for this project was to create a Suspension Tuning System that would collect data recommend suspension tuning settings to create a faster ride or a smoother ride. We soon realized the scope of our project had to be massively reduced due to our limited resources and knowledge as mechanical engineers. This project was headed and completed by four mechanical engineers that were breaking into the field of mechatronics and electrical engineering. Due to this fact alone, our team considered this project a huge personal success.