Formula SAE
General Car Design
I started and lead the University of the Pacific's FSAE team to the FSAE Michigan I.C. competition in May 2022. My main roles were Treasurer and Project Manager to ensure that parts where getting ordered and chosen for the right reasons. In my senior year, I also took time to train and develop the next generation of students to manage the team. It took me 4 years to figure out all the different sources of funding available, so I made sure to pass on what I had learned to help set up the team for success with future car development and competitions.
We used a GSXF 650 motorcycle engine, with a steel tube chassis and suspension, using the internal components of a Miata differential for the drive train. About 80% of the fuel system, shifting system, coolant system, and electrical system were custom designed and created. The brakes and pedals were purchased from Wilwood, using flexible hydraulic lines. The engine was turbo charged with a 45mm turbo due to the 20 mm diameter air intake restriction requirement. The over all weight was around 850 lbs.
In reflection we learned some lessons about weight in the design process. Some areas we could have done better to cut down mass included the chassis (unnecessary triangulation and overbuilt for mounting components), the power train (used oversized hardware and excess material for mount/tensioning system), suspension (custom uprights with half the mass, 1/4" instead of 3/8" hardware, smaller A arms), and driver seat (integrate it instead of buying standard overbuilt 3rd party chair).
GSXF 650 Engine
Chassis and Suspension
Center of gravity and cross section of vehicle
Limited Slip Differential
Team getting ready for practice with all Tech Inspections passed
Finished Vehicle
Chassis
The design of the chassis had to fit the driver (within a margin), the pedal box, the engine, fuel system, and air intake. It also had to allow for connection points of the suspension. Given that this was UOPs first year competing, I decided to stick with a simple tubular chassis, made out of 1010 hot rolled steel with as few bends as possible. It was also in the best interest of my senior design team and I, to make it simple to fabricate. The outline of the bottom and mid layer are the same, with only 3 bends in the main hoop. In the future it would be advantageous to reduce the number of triangulated nodes.
The tedious part of the fabrication were notches of each tube to ensure close contact for welding between all the tubes. I use TIG welding to connect all of the tubes which took about a month. In retrospect, I believe MIG welding it would have been better due to the time trade off, and because when set up right, MIG can provide the same structural integrity as TIG.
The torsional rigidity of the chassis was tested in 2 ways, using an FEA analysis in SolidWorks and a real life test. For this I pined the chassis at 3 points (location of the suspension) and mounted a mass to the last suspension point. With variable loads I measured the deflection off of the reference surface to determine the largest torsional rigidity which was in the vertical axis of the vehicle (2927 Nm/deg).
Power Train
The power train started with the GSXF 650 engine. I decided to use this engine because I had previously bought the engine (for very cheap, sub $500) to rebuild an older Formula style chassis that I found on campus. This meant the electrical system was easy to transfer and I had experience with all the auxiliary systems, like cooling, fuel, ECU and shifting. Plus that way we had 2 engines available incase one had any issues.
The power train continues with the gear selection and for this there are options to upgrade to an automatic shifter. It would have required modification to the clutch basket and this affects the space constraints, and the engine housing (including clutch and gearing) was already interfering with the chassis. I also bought an electronic shifting kit which used a solenoid to actuate the shifting linkage, and while doing so cut the spark to the engine to unload the transmission. This allowed me to use paddle shifters on the steering wheel which I 3D printed and fabricated to send the signal to the kill switch and solenoid.
Next in the power train is the sprockets and chains. I used the OEM sprocket which had 15 teeth and 525 chain to ensure that the chain was not a weak point so there wouldn't be any adjustments needed at completion. For the differential I chose a Miata and used only the internals. This allowed me to mount the sprocket directly to the differential and aligned the output shafts with the rear uprights. This led to me use Miata uprights and CVs with a custom shaft welded between the CVs. One challenge with using the Miata differential was that when using only the internals, there is no enclosure for the lubricant. The housing for the limited slip components was made out of cast iron and only had a few holes. I asked my welding professor to patch the holes where he used Iron-Nickel to join the hot rolled steel plugs and cast iron housing. I also had to seal the out puts where the CVs plugged in. For this I used a combination of thick grease, O-rings, and gasket maker. This allowed me to fill the housing with differential fluid and it worked well in the end.
For the gear ratio I decided to stick with the OEM gear ratio from the bike. This is from crank to ground, so I accounted for the wheel size and determined the drive sprocket size to match the ratios. This was the worst decision for the performance of the vehicle because the motor cycle has a max speed 135 mph, and FSAE vehicles max needed speed is 80 mph, with an average of 35-40 mph. With a larger drive sprocket and and smaller engine sprocket more power could have been utilized. During the competition we shifted between 1st and 2nd gear almost exclusively even though the engine has 6 gears.
Finally, the team decided to add a turbo due to the 20mm diameter restriction required by the rules, as well as an oil system to cool and lubricate it which pulled and returned directly from the engines oil reserve. The goal was to add a custom fuel tech harness and tune the engine to have the right amount of fuel for at every RPM range. However, this did not work at the time because the fuel tech could not control our spark plug coils. They required a specific kind which we did not have or a stand alone controller. At this time we were in March and ran out of time to troubleshot. So I made the decision to use the OEM harness and use the default tune which ended up working well. At low RPMs we ran slightly fuel rich, but at high RPMs we made around 0-5 psi boost equating to running naturally aspirated.
Shows the Miata LSD before modifying
CAD of the differential mounts and tensioning system
Shows the drive sprocket mounted, including the sealing O-rings, pillow block bearings, and oil fill port
Mounted differential including spacers for tension and chain
Shows the CVs and the Uprights
Shows the shifting solenoid and mount (silver) and the oil scavenging pump (white and orange)
Electrical
We started with the goal of using the fuel tech ECU and a custom wiring harness to be able to tune the engine to achieve max efficiency, while maintaining power. This was due to the 20mm diameter air restriction, and the use of the turbo to counter the reduction of airflow since we where using a 4 cylinder engine with 656 CC. Around March we determined that we would not be able to solve the problem that the fuel tech ECU could not control the spark plug coils. We would have had to get a stand alone spark control system as we determine later in the year, but at the time we did not find out the root cause of the issue. Given the time left I decided to use the OEM harness since I had previous experience with it.
I had to extend the connections for the fan, brake light, and gauge cluster. I had to create a custom connection for the fuel pump and level, because on the motorcycle they where integrated into one assembly and one plug. For the pump I used a standard high flow automotive pump using 12 volts and for the fuel level I bought an equivalent resistive float for the size of the fuel tank that we used. I also integrated the wiring for the shut off systems required by FSAE such as the master switch, emergency cockpit switch, inertial switch, and brake overtravel switch. The auxiliary systems I wired in were the oil pump and the electric shifter. The kill switch for this had to be spliced in, directly to the spark plug wiring, and all of these systems had to be control from the dash of the vehicle.
For details, see the wiring diagram below. Most of it is OME, but all components that are not in use (ABS, lights) were removed, and the safety systems, and auxiliary systems where added.
OEM harness with extended gauge cluster wiring
Dash where 7 is emergency shut off, 8 is paddle shifters, 10 is power to fuel and fan, 11 is the scavenge pump, 13 is boost gauge
Body
This was my first experience working with carbon fiber at a larger scale. We were fortunate to have someone donate 5-10 roles of fiberglass and carbon fiber, along with all the pumps and equipment to vacuum form the body. We had to get the wool, bags, and airproof tape. This process worked well except we used the wrong resin at first giving us floppy parts at higher temps 80 F. Once we got the right resin the process worked great.
The molds I modeled in Solidworks to match the contours of the chassis as close as possible. Once it was modeled I created a positive mold and cut it into sections to be cut on a CNC with max 4" of z axis travel. I got pink foam from home depot at 2" thick and once they where cut out I glued them together to form the full mold. The final step was applying a layer of fiber glass and epoxy to create a solid mold that would not absorb the resin, especially under vacuum. This also allowed us to remove any imperfections, giving us a clean final result with the carbon fiber.
The body was mounted with tabs on the chassis using nut inserts, grommets to protect the carbon fiber, and screws to hold it in place. This method work surprisingly well and was easy to remove when needed.
One of the FSAE competition design judges expressed positive reviews of our quality composites process. See timestamp 2:40 in the video linked below.
Failed nose cone with wrong epoxy
Process of prepping for carbon fiber, mold in the back ground
Completed and mounted body and nose cone
Full body with vinyl, cut outs for the suspension are visible
Misc
The pedals use a basic 3 pedal assembly from Wilwood, one for the clutch, one for the rear and front breaks (2 master cylinders), and one for the throttle. I created a custom mounting bracket for the brake over travel switch, and throttle cables. The pedal assembly was mounted on a set of old adjustable seat rail, which allowed for quick and easy adjustment of the pedal position.
The exhaust was fabricated out of the OEM exhaust and cut up and re-welded to fit the tight requirements of our firewall and turbo mounting flange. I retroactively CADed the exhaust in SolidWorks and the heat shield.
I 3D printed the intake restrictor and ensured it met spec. I proceeded to coat the inside and outside in the high temp epoxy because the intake ran right in front of the engine with little protection. The intake was printed to optimize for mid range air intake since the gear ratio meant that the engine would operate across all RPM ranges (the longer restrictor is optimized for high RPM but it is too tall).
The suspension was a challenge because it was designed with the wrong mounting points in mind. As a result, most of the geometry was off. On top of that, once we built the chassis, mounted the engine and tested the center of gravity, it was determined that the track width is too narrow and would have resulted in the car tipping over in the roll over test. This was because of all of the suspensions A arms had been cut and extended. We targeted 15cm roll center in the front and 20cm roll center in the rear.
The paddle shifter was added at the very end of development because I had trouble getting the solenoid to actuate the shifter linkage both up and down gears. It would only work in one direction. I eventually found the root cause to be that the solenoid was not operating at it's full stroke. Once this was accomplished, I had to create an easy way for the driver to hit the shifting buttons. I looked for some designs and found a hinge and magnet combination that created a compact and effective package. The mounting ring was welded to the wheel, and the switch, hinge, and stationary magnet holder were bolted to the mounting ring. The actual paddle was mounted to the hinge. This mechanism worked great.
Pedal Box
Exhaust with pin whole (fixed)
Intake restrictor
Rear Suspension
Front Suspension
Paddle shifters
Conclusion and Future Work
I loved starting this club and executing on so many different designs which all lead to one great vehicle and competition. I am also so proud of the team I accomplished this with! This recount is only what I worked on and there are so many parts and systems that would not have come into existence had it not been for the combined effort of the whole team.
As for the future for of the UOP team, they returned to the completion in 2024 with a lighter vehicle and passed tech and competed in the endurance event. I hope next year will be even better.
For myself, if I ever get the opportunity to compete in FSAE again I will be shooting for the top. There are so many ways to improve.
This is the assembly and completion video from the FSAE team at University of the Pacific 2022
This website uses cookies.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.