Taylor Made: 1,137 MPG
My senior design project last semester was to design the body shell for the Eco-Marathon team’s next car. I worked with Varun Punjabi, Jiehao Chen, and Nicholas Mark under TA Sameer Muckatira and faculty advisor Prof. Bruce Flachsbart. The finished car will compete in the Shell Eco-Marathon Americas at Sonoma Raceway in California beginning April 19th.
We worked closely with members of the eco-marathon team, Eco Illini Supermileage, throughout the semester. Illini Supermileage covered work for the engine, drive train, and electrical subsystems. Our project encompassed a steering system design as well as progress made toward a rolling chassis. At the beginning of the project, we divided the workload into three categories and chose leaders for each: manufacturing (Varun), steering (Jiehao), and aerodynamics (me and Nick).
Steering focused on designing a steering system with adjustable toe and Ackerman angles. The design also included optimizing the wheel alignment for least rolling resistance, a process which used a Matlab simulation to test a range of camber angles. At the end of the semester, we had our steering system assembled and mounted on a test chassis along with the wheels that will be used in competition. Some structural components were redesigned from previous years; each of those was 3D-printed and included in the assembly, with aluminum parts to be milled the following semester.
Aero focused on designing the body shell. We used PAC-Car II, a Swiss-made world-record-holding eco-marathon car, as a basis for our initial design. We desired the PAC-Car’s teardrop top view profile, which was shown to be one of the most efficient configurations for eco cars. Shell’s design rules and constraints caused our design to have significant differences from PAC-Car’s; most importantly, we were required to have front wheel steering. The car would have two wheels in front and one or two in the back depending on our preference. Front wheel steering meant that the frontal area of the car would have to allow for turning radius and would be significantly larger than PAC-Car’s.
We used our research and Shell’s rules as a guide for our initial CAD model. We then determined, from measuring the team’s competition driver in various driving positions, that there was a trade-off between driver height and length— i.e. she could be laying close to flat, taking up less vertical space, or sitting up higher, taking up less horizontal space. The trade-off led us to test two body shell configurations: shorter total length with taller driver compartment, and longer total length with shorter driver compartment. After several iterations, we concluded that the former had better experimental performance under our conditions.
We then iterated to optimize other design features such as wheel fairing profile and camber of the underbelly, the latter of which was significant for performance in ground effect. We tested each iteration in ANSYS Fluent, gaining an experimental drag coefficient that we then used to calculate the predicted drag force the car would experience at 25 mph (our estimated top track speed). Because of the long duration each iteration took to complete, we sometimes had separate iterations running on three or four computers at once. One batch of iterations took eight hours to complete while using four computers. However, the effort paid off: our finalized design had a 35.7% improvement in drag force over our initial design.
One of our last touches to the body shell CAD model before moving into fabrication was to create a recess for the front window. The window itself would be added post-fabrication. The team had suggested that we make a window frame in the body shell so that the window would be held flush as opposed to taped or glued onto the shell’s surface. We researched the previous eco car’s and PAC-Car II’s window designs to get a basic profile and then used some of our driver measurements to customize its dimensions to the driver’s range of vision.
Manufacturing encompassed picking the best material for the body shell and performing FEA analysis. Some factors we considered were strength, number of layers/thickness, and cost of the material. We narrowed our choices to fiberglass and carbon fiber and ultimately chose carbon fiber. Another consideration was whether to implement Nomex Honeycomb, a lightweight, structural material that could add strength and rigidity to the carbon fiber layers at little cost of added weight. By the time the body shell design was finished, steering parts had been 3D printed and materials research and testing was done. Additionally, we had chosen to use the VaRTM process (vacuum-assisted resin transfer molding) for creating the carbon fiber body and had made a 1/4th scale model of one design iteration to test the procedure.
After the body shell design was finished, all four of us transitioned to the next manufacturing step: fabricating the shell. We decided to split the shell into two pieces along centerline, basically creating canopy and lower body sections. The mold for the lower body was too large to mill in-house and was sent up to Ingersoll in Chicago at the end of the semester for machining. Since the canopy would not bear structural force, it could be divided into smaller sections and machined on the 5 axis in ESPL. We split the canopy into nose, middle, and rear sections and planned to have fix the nose to the lower body post-fabrication and make the middle and rear sections removable for driver and engine access. We cut, sanded, and glued together layers of very dense foam that was then machined into a mold. The molds were used for VaRTM to create each piece.
Because of our workload, our ME 470 feedback panel did not expect us to have any sections done by the end of the semester. The expectation expressed at our status presentation was that we deliver finalized designs and a scale model. Everyone involved was determined to go beyond for the sake of the team’s timeline to competition, and so we worked with team members to finish all three in-house molds as well as the nose and middle canopy sections. With them we also built the layered foam block that was sent to Ingersoll. We mounted the finished sections on our test chassis along with our completed steering system and hooked up a remote-controlled electric motor to the rear wheel. On presentation day, we surprised our audience by opening the door so that our in-progress car could seemingly drive itself into the room from the hallway. We were one of six groups to win the Outstanding Achievement Award for Excellence in Engineering Design.
Although our project ended with the semester, the team’s work never stopped—they took over everything we had done and seamlessly carried on toward the goal of competing in Sonoma. Illini Supermileage’s current competition record is 1,137 miles per gallon, a record we are confident the latest car will beat. Everyone involved has put forth a ton of hard work and great effort throughout, and I am proud to have been a part of this project.
Photo above: Last year's eco-marathon car, the 4th generation of Illini eco-cars. All photos by Taylor Tucker or the MechSE Department.