4th December, 2016

SEM 2016 Day 6: Optimisation

We woke up on Saturday, day 6, sitting at 19th place of 49 battery-electric prototype competitors on a score of 157 km/kWh. Of those 49, just 20 had completed a successful run. We were happy to be among the select group that had completed 8 gruelling laps, but we intended to be much higher on the board. Friday’s successful second run was the result of a low-gear setup for robust hill-climbing without much regard to efficiency. To move up the scoreboard would require some risk-taking and a move into unknown territory. In SEM, only the best of 4 attempts counts, so the 157 km/kWh couldn’t be taken away from us. We had some freedom to experiment. After much debate we opted for a slightly higher gear, which should allow us to build up more momentum on the flats and downhill. This would mean a less stressful entry to the big hill, but less traction or torque if we needed it in a tight spot.

One of the things that makes SEM great is the camaraderie among competitors. Many teams stopped by to ask a favour. 8 mm drill? 1000 pF capacitor? Don’t suppose you’ve got any imperial bolts? Most times we couldn’t oblige, but when we could, we did. We were blessed with the little generosities of other teams. Slovakians gave us a sheet of 1 mm polycarbonate for a window repair.  And always there’s the eagerness to explain designs and share experiences, even from teams fighting for a 1% margin over their rivals at the top of the leaderboard.

Keith got to know our paddock neighbours from Southampton University, competing in SEM for the first time. They were battling through technical inspection and not expecting to get on track that day. They had trouble with their drive electronics (perhaps connected to the fact that their electronic engineer was in Romania, getting married as we spoke) and they had two very nice tyres that they wouldn’t get to use. They lent us the tyres, and they got our spare propulsion battery and some of Barry’s considerable power electronics expertise. We bought one more tyre to make a full set. For the more financially anxious members of the team, this was much more palatable than buying three of our own. You pay fancy prices for fancy tyres. Thus a team debate was settled, and the Geec acquired a set of tyres with the lowest rolling resistance of any in the world. Thanks Southampton!

 Sean and Barry checking the drivetrain, Sean and Hugh fitting tyres, Sorcha in position for Arduino hacking.

Sean and Barry checking the drivetrain, Shane and Hugh fitting tyres, Sorcha in position for Arduino hacking.

But it’s never that simple. The slimmer tyres left the car lower with the rear bodywork riding just 13 mm over the ground, risking bottoming out on the small ramps on the track. We did a quick paddock tour looking for cars with similar ground clearance and discussed the issue with a French team in our impeccable French. Vingt-cinq millimetres? Eh oui. Avez-vous des problemes comme ca? Il y a du… bump? (At this point, making hand gesture for car slamming onto road.)

We couldn’t escape the conclusion that the body was too low to the ground. To drive it like that would risk big energy losses  and damage to the body at every bump, but there was no way to move the body higher on the chassis. It would have to be modified. For the second time in a week, Sean cut his beautiful creation, the carbon-epoxy Geec 2.0 body. This must have been distressing, but the change was cosmetically invisible, and the functional aerodynamic effects were almost certainly negligible. It was a price worth paying for the new tyres, with rolling resistance at least 10 times lower than the first set.

Laura is visualising, probably.

Laura’s pre-race preparation. She is probably visualising.

Around 4.30 pm, Laura launched from the start line with the Geec’s characteristic genteel acceleration. 0-100 km/h never, 0-25 km/h in half a minute or so. The team huddled on the hillside. Laura curved into view with controlled momentum and simply breezed up the hill, despite the higher gearing. The first and hardest obstacle of the run was cleared. As we counted the 5 minutes until her next appearance, utter chaos developed on the hill. Two cars faltered and stopped, within metres of each other, at the base of the slope. At this moment Laura swung around the bottom corner and immediately faced the pointy tails of the two stationary cars, blocking most of the road. She had no choice but to brake, almost to a dead stop. As she crawled around the first stopped car and applied power again, the Geec came to a standstill and rolled back.

Until two days before, the Geec had never climbed anything steeper than the NUI Galway car park with its barely noticeable 1% slope. Starting the Geec is a specialist skill at the best of times, and now a hill-start on a 5% slope – in wild traffic – was needed to rescue the run. We watched a series of forward lurches, backward rolls, promising starts and sudden stops. In the cockpit, Laura was easing the pedal down, watching the current readout, trying to coax enough torque from the motor to overcome gravity without pulling enough current to trip the electronics. After a few iterations on the accelerator, brakes, and battery controls, she found the winning recipe and began to accelerate up the road. The Geec lived to fight another lap.

Weaving uphill

She’s weavin’ like a loom on the mountain of doom. See that train? That’s what horizontal looks like.



Laura and the Geec (bottom left of picture) rounding the magnificent London Olympic Stadium and ArcelorMittal Orbit.

Laura was far behind schedule by now but she recovered the situation with a string of 4-and-a-half-minute laps, and eased off to the target steady pace of 5:00 to 5:10. The final lap presented another obstacle course of stationary carbon fibre and aluminium. This time she found a line, weaved through at speed to the crest, and cruised downhill home to the finishing line.

Waiting for the score check

Waiting for the score check at the end of the run.


The dreaded Joulemeter, the box provided by Shell that watches every Joule of energy coming out of battery, and determines our score.

The dreaded Joulemeter, the box provided by Shell that watches every Joule of energy coming out of the battery, and determines our score.

Again, the agonising wait for a score. Barry, Shane, Sean and Laura emerged from the start/finish tent with long faces: 153 km/kWh, slightly worse than the previous day’s score. Analysis went on late into the night. The optimisation attempt hadn’t succeeded, but we couldn’t yet tell if that was all down to bad luck with traffic on the hill, or if there’d been an engineering miscalculation. The onboard data showed us current draw over the whole run and highlighted where energy had been wasted. Not surprisingly, that second lap was an entropic disaster zone, with many bursts of high current as the car stopped and started. Encouragingly, Hugh found that some of the later laps showed much better energy scores than the run as a whole. Meanwhile, a tired team began to plan for the final attempt on Sunday. To be continued!

Josefine Kristy Web Developer

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Curabitur vestibulum eget justo sit amet varius.

Website: http://www.spikeincor.eu