Recent Projects: Transportation
Front Wheel/Drive System
We addressed inadequacies of the 2012 race car. The deliverable was to develop the drivetrain without adjusting the suspension. This will be the first time Dartmouth Formula Racing has used electric batteries in an all-electric configuration. We analyzed and purchased a motor kit and then divided tasks by drivetrain components — throttle, battery container and housing, motor mounts, and wiring — to put it together in parallel. The mechanical side ran finite element analysis to test stress and strain put on the vehicle. The electrical side learned how the components interacted and how to safely protect both the car and the operator. We were able to get the all-electric car running by the end of the term.
Hot-Wiring Of The Future: Exploring Car CAN Busses
Team: Chris Hoder, Theodore Sumers, Grayson Zulauf
Sponsor: Siege Technologies
Advisor: Sergey Bratus
Cars' electrical systems have evolved into complex networks of sensors and microprocessors, referred to as Electronic Control Units (ECUs). A typical modern car contains between 50 and 70 ECUs, which communicate with each other over the car's Controller Area Network (CAN) bus using the mandated CAN protocol standard. Siege Technologies believes malicious parties could exploit these unauthenticated intra-vehicle networks, and thus sponsored our early-stage research into the security flaws present in cars' CAN busses. Our deliverables were a proof-of-concept demonstration of vulnerabilities in CAN networks by compromising one or more safety-critical car systems, a methodology for uncovering and exploiting security weaknesses in the CAN bus of any vehicle, and a software package to implement that methodology on any vehicle. We purchased a 2004 Ford Taurus, reverse-engineered the manufacturer-specific protocols used on its CAN bus, developed hacks on safety-critical ECUs, and produced a generalized methodology for characterizing and exploiting a vehicle's CAN bus. As we designed experiments, we developed the software to implement them. The final result is an intuitive interface allowing the user to view, store, and analyze raw CAN data, quickly implement our experimental methodology, and use our code base to implement their own vehicle-specific hacks.
Lightweight Vacuum Power Flywheel Enclosure
Team: Awais Malik, Cody Engle-Stone, Zachary Kowalski
Sponsor: Ron Muller
Advisor: Ulrike Wegst
Our deliverables were: SolidWorks models and analyses of viable flywheel and enclosure systems and an optimization and feasibility study of the flywheel and enclosure system. Based on findings regarding measures to effectively contain a flywheel's worst-case failure, our recommended design for the enclosure has two layers: an inner "burst liner" and an outer "pressure vessel." The burst liner acts as a brake in the event of a failure. The pressure vessel surrounds the burst liner and maintains the vacuum inside the enclosure. We designed this redundancy because carbon-fiber flywheels are capable of rapidly dissipating a large amount of energy when they fail, and if any oxygen enters the high-temperature enclosure, an explosive combustion reaction may occur. We developed material selection case studies to justify our choice of carbon-fiber-reinforced polymer for the flywheel, low-alloy steel for the pressure vessel, and titanium alloy for the inner burst liner. We created an easily customizable Matlab program that outputs viable flywheel systems, given constraints in system energy, energy per flywheel, mass, volume, maximum rpm, number of flywheels, or specific flywheel dimensions.
Walvisstaart Propulsion System
Rising fuel prices and environmental concerns have incentivized research of efficient marine propulsion systems. Taking cues from fish and other marine life, flapping foil propulsion shows promise to significantly increase fuel efficiency. Walvisstaart's flapping foil propulsion system is designed for vessels more than 100 meters long. We investigated the viability of a Walvisstaart-style system for commercial fishing boats between 10 and 20 meters long. The primary deliverable was a functional, scale-model prototype of such a design. We developed a computer model of the Walvisstaart system and used it to design a prototype. We measured thrust and torque created by our device at various advance coefficients, and used these data to calculate the characteristic non-dimensional parameters of a Walvisstaart system. Both our theoretical and computational models agree that a Walvisstaart-style propulsion system can achieve higher hydrodynamic efficiencies than a conventional screw propeller.