Authors: Nathan Quinlan, Maeve Duffy, Rory Monaghan, College of Engineering and Informatics, NUI Galway How efficient is your car? Most family cars need four to six litres of fossil fuel to cover 100km (45-70 miles per gallon, in more familiar measures). Hybrids run at the better end of that range, and perform more consistently across various driving situations. Small advanced diesel engines approach three litres per 100km or 100 miles per gallon (though with some hidden costs, as we have recently learned). Why should we care about vehicle efficiency? How about that fact that Ireland depends on imports for 90 per cent of the energy it uses, at an annual cost of nearly €7 billion? A total of 40 per cent of our energy is spent propelling planes, trains and automobiles, which are 97 per cent oil-driven and are responsible for 20 per cent of national greenhouse gas emissions. We can achieve major impacts on Ireland’s energy security, greenhouse gas emissions and national energy bill by expanding the public transport infrastructure, using indigenous sustainable fuels such as renewable electricity and biofuels for transport, and improving the efficiency of transportation. In the summer of 2013, a group of NUI Galway engineering students and staff decided to see how far transport efficiency could be pushed. With the support of Shell E&P Ireland, we entered Shell Eco-Marathon (SEM), a global competition which had never seen an Irish competitor. SEM finds the world’s most energy-efficient cars built by students, which happen to be the most efficient cars built by anybody. In three events (Europe, Asia and Americas) teams test their cars on a 10-lap 16km test course. It’s a race where speed doesn’t count, as long as the average speed is at least 25 km/h. The winner is the one that uses the least amount of energy. Engineering student projects are usually low-stakes simulations of professional tasks. This one is different: an SEM team is a small business that has to deliver a complex innovative product on a rigid deadline. Success is measured in an unforgiving and public way in international competition. As the student team began to appreciate the scale of the challenge, dreams of glory gave way to realism. Our officially stated aims for the 2015 competition were to get through pre-race technical inspection and complete a run to register a score. In private, we allowed ourselves a slightly loftier goal: not to come last. The NUI Galway team opted to enter the battery-electric prototype class, named itself the Geec (Galway energy-efficient car), and got to work.aaacar6 Right: The team at the launch of the Geec in April 2015. L to R: Maeve Duffy, Nathan Quinlan, Tarek Nigim, Sorcha Tarpey, Eoin Clancy, Kevin Dunne, Hugh McSweeney, Maryrose McLoone, Cian Conlan-Smith, Niamh Keogh, Cian Lyons, John O’Connell, Paul Mannion, Daniel Fahy, Rory Monaghan, Barry Flannery, Erin Kelly. Not pictured: Adam Fleming, Mark Kelly, Wisdom Agba, John Mannion, Joseph Arkley, Fearghal Kineavy, Seán Bolton.

On the drawing board

The rules of the SEM prototype competition class let designers pursue efficiency at the expense of all other characteristics except safety. This leads to some extreme designs, most of which have two steerable front wheels and a single driving rear wheel, with the motor or engine in the rear. This arrangement allows a classic teardrop aerodynamic envelope, and is less susceptible to rollover than the two-rear-wheel alternative. Rollover haunts the design analysis. The car should stay as close as possible to optimal speed for efficiency at all times, but the threat of rollover dictates the maximum cornering speed. It is a function of wheel-to-wheel dimensions and weight distribution, which are also entangled with aerodynamics, chassis loading, braking and traction. This makes for a complex problem that can only be solved as the detailed design crystallises. In parallel with this high-level layout calculation, design must minimise the car’s requirement for energy and its various losses. This requires an understanding of the flow of energy from the on-board source to the wheels and beyond.

From battery to wheel

Efficient travel means minimising energy use per unit distance, or more precisely, the instantaneous ratio of power P drawn from the energy source to speed V. Equation (1) below summarises the most important influences on this measure of efficiency, and gives the engineer a to-do list. P/V=(ma + 1/2 C_D ρV^2 A + C_RR mg)/(η_c η_m η_t ) (1) [caption id="attachment_27083" align="alignright" width="300"]aaacar5 A member of the team during a test run[/caption] The top of the fraction on the right-hand side represents energy required to move the vehicle, which will be discussed below. The bottom describes losses inside the car. Power P flows from the battery through a controller, a motor, and mechanical transmission, with efficiencies ηc, ηm, and ηt, respectively. Power is reduced to ηcηmηtP at the wheel. This multiplication of efficiencies is punishing: even if each stage converts 90 per cent of energy, less than 73 per cent reaches the wheel.

The team at work

For simplicity of construction and robustness, a 24-V permanent magnet, brushless DC motor was chosen, with a 24-V, 20-Ah Lithium-ion electric bicycle battery power source. The speed-torque characteristic of a permanent magnet DC motor is quite flat, varying typically by only four to five per cent over its rated torque range. Therefore, the most effective and efficient means for controlling its speed is by varying the motor armature voltage. In our car, this was achieved by designing a voltage step-down buck converter whose output voltage was controlled by a PWM signal generated by an Arduino microcontroller in response to an input signal sent by the driver through an accelerator type foot-pedal. During starting, this enabled a gradual increase in motor voltage, thereby limiting its current until the motor drove up to speed. [caption id="attachment_27084" align="alignright" width="300"]aaacar3 Three minutes of data from the final competitive run, illustrating the on/off drive-and-glide strategy used to minimise switching losses in the DC converter[/caption] Once the car was up and running, its speed was controlled by applying the full battery voltage to the motor in pulses of 10-40 seconds, with no voltage applied between pulses, as shown in the graph. In this way, the average motor drive voltage was varied to produce a corresponding variation in speed. Switching on and off of the battery voltage was controlled by the same foot-pedal as used during starting, except that during driving, the pedal was pressed either fully on or fully off. When on, the applied voltage provided increased motor current and therefore acceleration, while the speed would drop by a few km/h during battery-off intervals. This approach was chosen to maximise efficiency of the step-down buck converter, ηc, with the elimination of high frequency switching losses in the main MOSFET and freewheeling diodes. For the car, it provided controllable operation over the range of speeds required for different track conditions of cornering and straights. The motor-to-wheel gear ratio is a key variable. Ideally, the motor should always beat the speed where its efficiency, ηm, is maximum. For a permanent-magnet DC motor, this is also where torque is lowest, so a car geared too low may struggle to accelerate up to and above its design speed. A car geared too high has to drive at higher speed to maintain efficiency. The Geec ran with a single-stage chain-drive with a motor-to-wheel speed ratio of 7.9.

From wheel to car

The driving rear wheel works against three forces that resist the car’s motion, appearing on the top of the right hand side of Equation (1). The first is inertia, the product of mass m and acceleration a, which must be minimised by keeping speed as steady as possible. The second is rolling resistance, due to energy dissipation in the viscoelastic rubber of the tyre. (It is quite distinct from friction between the road surface and the tyre, which opposes wheelspin.) The rolling resistance force is depends on CRR, a tyre property, and weight mg. Mass dominates both of these unwanted forces.The robust and simple Geec 1.0 weighs in at 82kg, while competitors ranged from 140kg to an impressive 27kg. The third force to overcome is aerodynamic drag, which is typically half as large as rolling resistance at the low speeds of SEM [1]. Drag is determined by the car’s speed V, air density ρ, frontal area A and the drag coefficient CD, which is a function of shape. Thus, speed should be kept low and the body of the car must be small, ideally just big enough to envelop a driver lying feet-first. Aerodynamics conflicts with the need for wheel-to-wheel width to prevent rollover.

Design, build, test, learn, repeat

The team converged on a low-risk, conservative design philosophy. Thus, the first Geec has a steel frame, a relatively high driving position, a heavy-duty chain drive, and a simple body. With a low centre of gravity and wide-set exposed front wheels, we calculated that all three wheels would stay firmly on the ground through any conceivable manoeuvre. [caption id="attachment_27085" align="alignright" width="300"]aaacar2 After two years of preparation, the moment of truth[/caption] Development provided students with a taste of engineering that cannot be experienced in a lecture hall. There were good days, as when the car first hummed across a yard at walking pace, or hit race speed. There were bad ones, like the wet weekend when the high-performance motor faded and finally died, teaching some hard lessons. Nonetheless, students learned that the need to integrate all components into a fully functioning car required continuous team communication; driver safety and performance introduced ergonomic, user-centred design; a range of switches and sensors was needed to ensure adherence to electrical safety rules; while an active sponsorship drive resulted in significant publicity, providing students with opportunities for high profile media communications.

Race day

SEM is an extraordinary festival of engineering, innovation and energy. It attracts some 50,000 engineering nuts of all ages as well as 3,000 competitors in 197 teams. The cars and their creators are housed in the paddocks, which close at midnight; at 11.45 pm the huge hall is still full of the sound of motors whining, tools clinking, and horns tooting (a horn is a mandatory safety feature, but also useful for spontaneous celebratory non-music). Six short hours later, the teams return from the nearby campsite to get back to work. Friendships were made with neighbours in the paddocks and technical tips exchanged. Tyres were kicked and composite bodies poked and prodded. Above all, the atmosphere in the paddocks provided an incredible jolt of competitive energy to the team. It became essential for the Geec to belong at SEM, not just be there. [caption id="attachment_27086" align="alignright" width="300"]aaacar1 In action on the streets of Rotterdam[/caption] The first half of the week was consumed with rebuilding the car after its journey by shipping container, and preparing it for the dreaded technical inspection. Our conservative design philosophy paid off handsomely, with the Geec relatively easily passing ten rigorous tests including electrical and mechanical safety, braking capability, driver emergency evacuation. Objective 1 had been achieved. After a day of practice laps, competition began: up to four attempts over three days to cover the 16km course with minimum battery energy. As the Geec pulled out on to the track for its maiden competitive voyage, hearts were in mouths. We were sharing the track with 29 other cars of all shapes, colours and sizes. Some 35 minutes later, the Geec pulled into the finishing tent with an official score of 172 km/kWh, equivalent to approximately 4,800 mpg. The Geec had completed a competitive run; objective two had been achieved. The following day the Geec logged two more clean runs, in which improved driving strategy saw steady efficiency increases to 202 km/kWh. The rollover calculations proved correct: the Geec remained firmly stuck to the ground. Steely nerved drivers Maryrose McLoone and Niamh Keogh were well attuned to the car’s limits and could maintain momentum through the most critical corners, where aerodynamically cleaner machines were forced to coast down and accelerate back up, working against inertia at off-optimum motor speeds. The secret third objective - don’t finish last - now seemed secure, and there was space for a gamble. The team had theorised that a 10-tooth sprocket on the motor shaft would enable higher efficiency, but with an enforced change of motor just before shipping the car, we never got the chance to test it. The change of reduction ratio from 7.9 to 9.5 would slow the car close to the target average of 25 km/h, but a run below that speed would be void. In a final roll of the dice, we pumped our tyres 30 per cent above rated pressure. The result was a jump of almost 100 km/kWh, up to 287 km/kWh or 8,000 mpg. At this efficiency, the Geec would use 13 cents worth of electricity to drive from Galway to Dublin (if it had a big enough battery). Even among the exotic and extreme engineering of Shell Eco-Marathon, this was a score to be proud of: the Geec, on its debut, finished 23rd of 51 competitors in the battery-electric category.

The future

[caption id="attachment_27088" align="alignright" width="300"]aaacar7 An early design concept for Geec 2.0, with an aerodynamic simulation[/caption] Geec 2.0 is now in development to compete at SEM 2016 in London. It will be a lighter, more compact car, with an aluminium chassis and wheels fully enclosed in pods under a raised body. Computational fluid dynamics (below) has helped us to balance aerodynamics and stability. After building Ireland’s most energy-efficient car, the team will travel with bigger ambitions and hard-won experience. This will be the beginning of a long campaign to build the world’s most energy-efficient car. We are indebted to the supporters and partners who made the first Geec possible: Shell E&P Ireland, Wood Group Kenny, Belcross Enterprises, Sinbad Marine, CBS Bearings, Smurfit Kappa, Quicktec Computers, Enform Plastics and Maxon Motor. This year we welcomed new partners ÉireComposites and Ard Precision who are deeply involved in the 2016 and 2017 cars. We are also fortunate to have the support of RMS, ANSYS, CADFEM Ireland, Pat Rynn Engineering, Enform Plastics and Easy Composites. However, more resources are needed to compete in London at SEM 2016. If you would like to support the Geec with funds, materials, or services we would love to hear from you. After beginning as a tentative experiment, the Geec has been embraced as an educational initiative and wholeheartedly supported by NUI Galway and its College of Engineering and Informatics. For final-year students, it comprises the major project and the core of their academic programme. Younger students can build up experience and take on increasing responsibility as they progress. For all participants, it’s an uncompromising leap into the deep end of technological innovation. We look forward to a sustainable future for this project, and hope to see more like it energising the landscape of engineering education in Ireland. To follow the Geec’s journey to SEM Europe 2016 in June, please visit,, or @theGeec on Twitter. References J Santin, Christopher H Onder, J Bernard, D Isler, P Kobler, F Kolb, N Weidmann, L Guzzella, The World's Most Fuel Efficient Vehicle: Design and Development of Pac-car II, vdf Hochschulverlag AG, Zurich, 2007