Our third trip to Shell Eco-marathon Europe felt different from the first two. The Geec 1.0 had been the pathfinder, a rugged machine built to get around the track at all costs. It did a lot better than that, finishing 23rd out of 51 competitors in Rotterdam. The Geec 2.0 was an entirely new car with a low profile and a wide track for good handling, fully enclosed in an aerodynamically designed body. Like many cars on the new London track in 2016, it struggled to climb the steep hill without overheating or tripping electronics. Rotterdam, in hindsight, was a walk in an exceptionally flat park. But knowledge built up as the days went by, and the team came home with 21st place, a massive accomplishment for a risky design and an exhausting build that was done in barely half a year.
So in Autumn 2016, a new team inherited a prototype with huge untapped potential, thanks to the work of about 30 people over 3 years. This was the first Geec team that didn’t feel the need to start again with a brand new car; the Geec 2.0 had been built to a fundamentally sound design that left room for step-by-step development. The Geec 3.0 was built around its predecessor’s aluminium structural chassis and its carbon-fibre aerodynamic skin, more or less intact, with some systems replaced or upgraded (more about that in future posts). Because we weren’t building from scratch, we had more time and headspace for testing and analysis. We developed a dynamometer to bench-test motors, giving us deep knowledge of the drivetrain performance. This fed into a new computer model where we could simulate the car on a detailed virtual circuit. To top it off, we were returning to a track we’d already fought and overcome, with two race-hardened drivers.
The first days in London went smoothly. For the first time, we cleared technical inspection in a single clean sweep. The car weighed in officially at 50 kg, compared to 68 kg the year before, and 87 kg before that. Practice sessions gave us reasons to be optimistic. Niamh and Laura completed a full race-length run each of 16 km over 10 laps. Last year’s nightmares on the hill were fading away. The days were long but they went fast. The paddocks were open from 6 am to 10.45 pm and we used every hour to improve the car in dozens of little ways.
We found that the course designers had been busy setting new problems. There was now a hairpin corner of more than 180 degrees, positioned close to the lowest point in the course – just where you’ll be fast, if you’ve looked after all the energy you had at the summit. The hill was still there, but with a shorter lap it had to be climbed 10 times instead of 8. More time would be spent on the hill, pushing the motor hard at low speed, high current, and low efficiency. On the plus side, the remaining 1.4 km of the circuit was nearly all downhill. The models suggested we would hardly need the motor at all, except on the hill.
How SEM works
After several days of inspection and practice, every passed team gets up to four attempts over the final weekend. To score, the car has to complete 10 laps (15.7 km) in 39 minutes or less. The race marshals measure the energy consumption by checking their official joulemeter, installed on our battery, before and after the run. The score is reported as distance travelled per unit energy used, in kilometres per kilowatt-hour (km/kWh). The team with the biggest score gets a big trophy.
When the session is on, cars queue up to go through pre-race checks, then feed onto the track from a slip lane. There can be up to about 30 cars on track at a time. They’re all trying to run as close as possible to the minimum allowed average speed, but it’s not an orderly procession. Speed goes up and down as cars handle gradients and corners, and every team has a different strategy. So although cars aren’t racing head to head, they are manoeuvring and passing. SEM rules require drivers to be considerate, in general, and allow others to drive at their preferred speed without being blocked. This would turn out to be important for us.
Scores on the doors
The car was set up with the same robust motor we’d used last year and a high-ish gear. Laura took the Geec out for the first competitive run on Friday afternoon, and it was gloriously uneventful. We were on the board with a solid score of 190 km/kWh. It was fun to see ourselves in 12th place on the screens, but we slipped down as more cars completed their first runs.
The next move was a change of motor to a lower-powered and lighter version of the first one. Niamh drove run 2 without drama and came home on 241 km/kWh, beating our 2016 best of 236. The shorter lap made for some crowds on track, but last year’s hard-earned driving experience came into play. We’d bagged a couple of safe and sensible runs, and set a new best with two runs still in hand. That score of 241 couldn’t be taken away; only the highest score would count in the final rankings. All the design work told us the car had more to give, and now we could afford to take some risks.