When it comes to professional single-seater racing, Formula E cars are the quietest, the wackiest and the slowest on the stopwatch – although some folk do appreciate not needing ear plugs while admiring cars that look as though they have burst from a sci-fi comic.
That’s subjective. But how slow in fact are they? This year on the Monaco Grand Prix circuit, Formula E lap times were comparable to those of the Formula Regional European Championship, a series pitched below Formula Three. That is despite the Formula E car exploiting 50% more horsepower.
Genuine racing – the small footprint of the Formula E car allows overtaking even at Monaco
On the other hand, the Formula E semi-spec car is nearly 20% heavier than the spec Regional car and has much less grip. That is due to the relatively low downforce imposed by the regulation bodywork in the interest of aero efficiency, plus the enforced use of spec all-weather recyclable tyres.
Of course, the higher the power relative to the available grip, the greater the spectacle. That is especially so for cars that have a small footprint and as such can actually overtake within the tight confines of Monte Carlo. It all made for a hugely entertaining 2023 Monaco ePrix.
It was all the more entertaining for the battle embracing a number of manufacturers: Nissan, Porsche, Maserati, Jaguar, NIO, Mahindra and DS. Let’s not forget that Formula E these days comprises an FIA World Championship. And step by step it becomes more technically advanced.
2023 was its ninth season and the first for its Gen 3 car. That has a spec chassis and a spec front-axle harvesting system, with manufacturers allowed to develop a bespoke rear powertrain solution. To explore its technology Gemma Hatton spoke to four engineering insiders.
The car package
In the case of Gen 2, while rules banned torque vectoring, it was possible to use two motors and thus not require a differential. Two motors are no longer permitted though, so there has to be a differential at the output from the gearbox. The totally spec Gen 1 car had a Hewland five-speed gearbox, but since the advent of Gen 2 it has been feasible to design the powertrain around a single gear ratio.
A single ratio removes the complexity of a shifting mechanism and a clutch. It is inherently the lightest and easiest to package; it also reduces transmission losses and intrinsically enhances reliability. We put it to our insider engineers: is this now the norm for Gen 3?
One of them said, “With every gear ratio you have, it’s about half a percent of efficiency loss. From the Hewland most teams went to two or three gears, and then as the next step it’s all been single-speed, once everyone realised the efficiency gains. Nobody has multiple speeds anymore.”
Another added, “We don’t know the precise architecture of everybody’s layout but it’s most likely they all just use spur gears, either one set or two sets, depending on how fast they want to run the motor relative to the wheels.”
Jake Hughes was fastest in qualifying for the Monaco ePrix, driving for the Nissan-powered McLaren team
Yet another remarked, “I believe everybody’s powertrains are now pretty similar, the technologies inside are pretty similar. However, currently some teams have a dual-stage gearbox. With a single stage the motor directly drives the crown wheel, which is attached to the diff, whereas some teams have a motor onto a reduction gear onto the diff.
“If you want to spin your motor really fast, you need to have that reduction gear stage to obtain the required rear wheel speed. It is still a single-speed transmission but you’re in a different efficiency area with different inertias, and the whole motor design is different. I think Jaguar and DS use the dual-stage approach.
“There’s a compromise between the mass of the motor and that of the transmission. Clearly, you can have a lighter motor when you have a lower torque and higher speed machine, but then the gearbox will be heavier as it needs the second stage.”
Early in the development of Gen 2, Renault housed its motor and (single-speed) transmission in a carbon fibre composite structure including the inverter, whereas most other manufacturers kept the original layout of the inverter on top of the battery housing. Have others since taken this integrated approach?
One of our experts remarked, “Ideally the motor and inverter should be integrated, because the closer you can get everything, the less hardware there is. The reason most people don’t integrate is because with the vibrations the busbar connections are a massive weakness.”
Another noted,“You’ve got a heavy motor and a heavy inverter, and the motor position is almost fixed; if you start moving it [away from the rear axle] you will introduce transmission losses. The inverter you can move forwards. The Formula E car has inherently a rearward weight distribution, so if you have the inverter more to the front, it could be beneficial for your overall vehicle dynamics.”
Our first respondent added, “However, the phase cables are a consideration. You get losses in those, so the further you put the inverter away from the motor, the more those losses are, and you’re also adding cable weight.”
To that our second respondent mused, “I think it’s more the case that mass is the driving factor; to get the combined mass of both items down comes before packaging considerations. But you also need to consider serviceability – if you have an inverter issue and you need to switch it, ideally you want to have it separate.”
Another of our experts added, “If you have to change the inverter and it’s linked to the motor whereby you can’t separate them, you’ll take a sporting penalty that’s double what is necessary to replace it.
The Formula E field is unleashed for the 29-lap race on the Monaco Grand Prix circuit
“Currently some people integrate the inverter fairly directly into the motor package. Potential disadvantages of this approach include a direct thermal interaction between the motor casing and the inverter. You have to consider that in respect of how you thermally manage the entire package, although cooling-wise it is possible to do both together.
“Also, you might want to isolate the inverter mechanically in a different way to the motor, to ensure you don’t effectively cause vibration problems for it. For sure, integration can bring advantages but it also brings some disadvantages, and it’s not clear-cut what everybody does. We know some people integrate closer than we currently do but we’re not really too concerned about that.”
Motor and inverter
The Gen 1-spec McLaren motor was radial flux and relatively small yet high-revving. In general, an axial flux motor tends to be inherently torque-dense, whereas a radial flux motor tends to be power-dense. The five-speed Hewland transmission was necessary to keep the McLaren motor in its most efficient operating range across low and high speeds.
Opening up development saw both axial and radial flux architectures coupled to transmissions of between one and five speeds. DS opted for a pair of axial flux YASA motors stacked on a common shaft, with drive sent through a single-speed transmission. But the overall package was heavier than a single motor solution.
One of our experts summed up the current situation thus, “I am fairly sure everybody uses radial flux these days, effectively because thermally they’re easier to look after. And they are pretty short, and more efficient at peak levels as well, the kind of use we make of them.”
In 2016, Marelli introduced a six-phase brushless motor plus twin three-phase inverters. Each inverter controlled one of two stators interacting with a single rotor. This created the effect of six-phase motor operation but with the simplicity and low switching frequency requirement of a three-phase inverter.
One of our experts explained, “An inverter converts DC [battery] energy into ‘three-phase’ AC; the motors are always three-phase. Some of the more advanced machines might be what people call ‘six-phase’ but it’s generally two three- phase circuits in the machine.
“If you run that you effectively distribute the current across twice as many electrical switch modules. So it means that in theory, you’re halving the load and you can get even more power. So in the case of Formula E [with power capped] it’s in the best interest of the efficiency of the system.
“But it’s going to be at a mass penalty, because you’ve got basically double the inverter volume. If you can live with that, in Formula E it’s key to maximise efficiency and not to drop off the battery towards the end of the race.”
Another of our insiders added, “Some people run two, three-phase motors superimposed on one another, but most don’t. Instead we have the true six phases, controlling every phase independently. “So rather than having two, three-phase motors permanently locked in, so that they run synchronously, you have independent control. I don’t think anyone does anything outside those two options. But to be fair, we don’t know for certain because everybody’s powertrain is different.”
In terms of motor speed, the message from our experts is that these days the range is in the 20,000-27,000 rpm area. “I don’t think anybody goes faster than 27,000 rpm,” one said.
Another one explained, “An electric motor becomes more efficient as its speed goes up. However, efficiency at high speed is offset somewhat, as the torque output falls as speed goes up. Adding a second transmission stage becomes necessary if you’re running the motor above 20,000 rpm, and that will cost up to 1% of efficiency, but in theory a higher-rpm motor can as an overall package more than compensate for that loss.”
Another of our experts said, “Personally, I don’t believe you can offset the efficiency loss of the two-stage gearbox.”
Yet another added, “It isn’t as straightforward as saying the machine gets more efficient the faster you run it. There are quite a lot of interacting parameters, which is why everyone has slightly different [efficiency] numbers, although they’re becoming more aligned than perhaps they were 4 years ago or so.
“A lot depends on the interaction between the control unit and the motor, and that in turn depends on how many phases you run, within the inverter and outside, how many control phases are in the motor, what kind of pole arrangement you have in your motor – all sorts of things. It is in the details where everyone does something slightly different.”
Another of our experts noted, “There are a lot of challenges with making electric machines go fast, and that comes more into the motor drive technology. Your PWM switching frequency is effectively the switching rate that synthesises the AC waveform.
The winner of the Monaco ePrix was Nick Cassidy, driving for the Jaguar-powered Envision team
You need a certain number of PWM switches per waveform, so the faster you want the machine to go, the faster the PWM switching frequency needs to be.
“The higher frequency you can have there, the better the synthesising of the current waveform, so the motor will have fewer switching losses reducing its efficiency. However, every time you switch there is a loss period within that. You have to manage those losses, otherwise the inverter or motor becomes too hot.
“So, silicon carbide switching technology has become more prevalent [replacing IGBTs] and is now standard across the Formula E grid, running upwards of 40 kHz switching frequency.”
The battery
The Gen 3-spec battery has pouch-type cells from Total SAFT, and is designed and assembled by Williams Advanced Engineering. One of our experts noted that, compared to Gen 2, the Gen 3 battery is a similar size but lighter and, using a different cell technology, its power density has been significantly increased.
“Those specifying the cell for Gen 2 deliberately made a conservative choice, and since then there have been 4 or more years of cell development,” he said.
He added that where the Gen 2 battery had to be run in a tight temperature band, the Gen 3 battery can run a bit hotter. “For sure, you can still drive the car in a way that will result in you having some limitation due to the thermal limitation in the cells.
But whereas with Gen 2 that was critical regarding race strategy it’s so much less of a topic now that it’s almost peripheral. “Also, the fact that the battery can run hotter means the energy you need to release to go through the cooling system is a bit less.”
The spec battery control software constrains the unit in such a way that each one performs across the season in the same way as all the others. “That was the same with Gen 2,” remarked one of our experts. “There have been more challenges in that respect with the Gen 3, because it’s technologically a more challenging product, but [the supplier] has done a pretty good job of dealing with it on a fair basis through the season.
“There have been very few instances where you could argue there’s been an unfair advantage by somebody having a battery pack that was in a better condition than the others.”
Regen
The introduction with Gen 3 of a spec 250 kW front-axle harvesting system came with the elimination of conventional front and rear friction brakes (aside from an emergency-only front brake provision).
One of our experts noted, “Front and rear harvesting means that rather than having a paddle phase, as we used to call it, whereby the driver would deliberately use the rear powertrain to accrue some energy – such as in the first phase of braking, after coasting – that now doesn’t need to happen.
“So the way the balance happens into the corner is very different. One complication is that there’s no locking differential on the front axle, so there’s a limit to what you can achieve on corner entry when the inside tyre starts to unload. From that point of view, there’s been quite a lot of ‘driver-style change’.
“Also, because the front and rear [harvesting] efficiencies are different, quite a lot of complexity has been added to the braking strategy. The current powertrains have been pushed to efficiency levels that are quite incredible, including about 99% efficiency in rear axle regen.
“That [front-to-rear regen efficiency disparity] does affect energy management strategy throughout the race though – I can say that without giving too much away.
“You knew where you were with Gen 2; if you didn’t want to ruin your rear tyres, and they were hot, you couldn’t regen as much as you’d like. Now, it’s not quite as straightforward. In terms of braking, it has become a bit more like a real racecar but during races, with the [front and rear] power caps, it is further complicated.”
Another of our experts noted, “With Gen 2 we would sometimes say to the driver, ‘You need to try and slow the car as much as you can from the rear axle only’, because that was part of our overall energy optimisation strategy. Now, because we’ve got the front axle regen, it means he can brake almost like a normal car.
“But the faster you’re going, the more likely you are to hit the [regulation] power limit of each motor. From an energy point of view it’s most efficient to sit on the power limit of each motor. To take an extreme case, if your rear powertrain is much more efficient than your front then you would probably bias your regen strategy to the rear motor, providing that the tyres are capable of achieving it.
Genuine glamour and race-able cars: Formula E at Monte Carlo
“If not, and you can take the braking performance loss, it could be that you are net faster to prioritise the rear motor to do most of the regen and minimise the contribution from the front motor. What you lose from braking effectiveness in terms of how quickly you can slow the car, you might gain back from being able to regen more from the regulation energy point of view.”
Software
When it comes to Formula E software, one of our expert witnesses noted, “For each race we have a submission deadline for our primary software to be on the vehicle controller that the FIA allocates to us during the event. And we do a software update every event. It’s a continuous war between the teams, in much the same way as aero in Formula One, because our hardware is locked.
“Software is our main freedom, so that’s where a lot of our resources go, just trying to make progress in every small area we can, event by event. Then, depending on the event, we will also try to prioritise different topics.
“The energy strategy is one of the main ones we work on event by event – how we decide what power is deployed when, and the driver must know what to do with his controls to implement that.
“Effectively we aren’t allowed to drive the car for him, which is what we’d like to do, at least some of the time. So there’s a lot of interaction between the prompts he receives about what we want the car to do, and how he trains to do that in a driver-in-the-loop simulator.
“If the driver drives in a manner that was not practised before the event or in the practice sessions, it causes a totally different set of operating conditions. Steam comes out of our engineer’s ears when that happens!”
Conclusion
Formula One claims that Miami and Las Vegas are glamorous new venues, but genuine glamour is Monaco. But these days can you get genuine racing around the streets of Monte Carlo?
Yes you can, and it’s called the Monaco ePrix. And when it comes to the enabling powertrain technology, unlike long-term frozen Formula One, this year there is plenty of fresh interest to be found in Formula E.
Acknowledgements
RET would like to thank Lewis Butler, technical director at Mahindra Racing; Jonathan Peters, chief engineer at Intertek; Joseph Lysaght, head of electronics and controls at NIO 333 Racing; and Ash Willoughby, energy management and controls engineer at NIO 333 Racing, for their kind assistance with this revealing insight into Formula E.