How Formula E’s tech war is intensifying

To many, driving is an art form, but it's actually more science than art. That doesn't always sit well with purists, but creating the fastest car and completing a race in the fastest time will always be a matter of mathematics rather than beauty.

The upper echelons of motorsport are becoming more and more an engineer's dream, and Formula E - with its world-first electric-powered single-seater championship - is naturally a part of that.

Guesstimations are largely left in the past these days as simulation tools become more and more advanced. Leaving things to chance when there is the opportunity to trial hundreds of different set-up solutions is the motorsport equivalent of declaring that Britain has "had enough of experts" while trying to win an argument about Brexit.

It's more than just the electric powertrain element of FE that lends itself to in-depth preparation and data analysis. Teams are collecting data all the time, battery and powertrain technology is ever-advancing and the circuits are either barely used or completely new, so there is little or no pre-existing data.

This has led to two technological battlegrounds in FE: hardware and software.

Simply put, the hardware offers the chance for big gains, while the software allows for optimisation of that. Both are underscored by intense simulation, and that's not just in relation to driver-in-loop simulators.

After months of development, the powertrains were homologated at the start of June. That means nearly all private testing would have been conducted with the knowledge that any fresh problems could not be countered by changes to the hardware - the motor, the gearbox and the inverter.

However, it doesn't mean each team will end the year with the exact package they started with. In terms of hardware, they will, but not how they optimise it.

Lucas di Grassi's engineer at the Abt team, Franco Chiocchetti, told Autosport at the end of last season how he would have loved to back-to-back the car that finished the campaign with the one that went to Beijing for the season opener.

Software developments and quality simulation allowed the Abt team to bring small refinements to each race last year, and that has to be the target again for most teams in 2016/17.

"In the first year the driver could make a reasonable proportion difference on saving energy," says Mark Preston, Techeetah team principal. "The margin of error of energy usage around a lap was big enough for the driver to make a difference.

"As software got better some of that got reduced."

This was a phenomenon experienced in the inaugural campaign but one that came to prominence in 2015/16, because individual powertrains threw up different data readings and the teams had to work out how to understand and react to it.

Gradually the power has shifted back to the teams. The more homework you do, the more prepared you are for the coming event. The more analysis you do after the race, the more you learn and the more you apply for the next one.

"It's nothing ground-breaking, it's just more work," says Sylvain Filippi, DS Virgin Racing's chief technical officer.

"Once the season has started you can't change the car; it's just software. It's more about being cleverer on how you use the energy and help the driver as much as you can."

Teams will look at several areas in the build-up to each event.

The old-school motorsport value of a well set-up car is not ignored, because it's intrinsic to performance and strategy.

The downforce/drag dynamic is key because drag is a killer on FE cars, which are power limited at 170kW in the race (227bhp). Plus, a low-drag set-up will mean higher straightline speeds, and more speed at the top-end of the straight is a plus when it comes to strategy, as the motor will not need to be worked as hard.

Low drag means less downforce, and that needs to be offset by the pursuit of mechanical grip - mainly front and rear ride heights, front and rear anti-roll bars, and damper stiffness.

Energy management strategies are, of course, crucial. What is the usage-per-lap target? How much energy can be harvested under braking? Where is best to lift and coast?

Simulation is key to this. It's one of FE's more complicated, and fascinating, battlegrounds - especially as the new-for-2016/17 steering wheel offers greater display functionality. More options are available in-car at any one time, which means a greater reward for more software R&D and coming to each event with more tailored mapping.

Filippi says "I don't know" when asked if it will spark a full-on software war, but admits: "If you have the resources to do the simulation work, it helps you react better to race events on strategy.

"The more you have to react the better."

This is the area where we could still see teams making gains or suffering losses over the 2016/17 season and beyond. But where everyone is starting from is also significant.

THE CHAMPION OPTS FOR EVOLUTION

Renault got the sums right when developing its first bespoke powertrain for last season, and went for a clean-sheet design.

This placed a Zytek motor transversely behind the inverter, which was moved from atop the battery to behind it, all within a carbonfibre casing.

The result was the benchmark powertrain, so it's no surprise that the Renault Z.E.16 is a matter of simple evolution for the upcoming season, with stable regulations and no immediate demand to change the powertrain drastically.

The carbon casing is now far slimmer and no longer houses any gearshift, as the motor now has a fixed single-ratio drive to the rear wheels.

While light and simple, this set-up demands a lot from the motor in terms of torque and efficiency. With no means to select a lower ratio, a lot of current needs to pass through the motor to produce torque.

This can be an inefficient use of the limited energy available from the battery (28kWh), and could blunt acceleration off the start and out of slow corners.

This is negated somewhat by the fact that last year Buemi and Prost rarely shifted down to first during the race, and the motor's rev limit and efficiency was sufficient for a good top speed.

ONE MOTOR OR TWO?

The DS Virgin set-up is a total departure from its twin-pancake motors and aluminium casing from last season, which was large in diameter and slower revving but delivering far more torque.

Now the car sports a single axial motor, mounted transversely and driving through two gears. This is a more efficient and lighter set-up, and converges on the Renault offering.

It's also a similar architecture to that adopted by Mahindra and Dragon Racing, which are effectively using the same set-up - the Mahindra-developed powertrain with Magneti Marelli motors.

Italian electrical expert Magneti is extending its Formula 1, World Endurance Championship and road-car knowledge of hybrid technology into the FE arena.

The composition is effectively two smaller motors on the same shaft and within the same casing, and the reduced electrical losses through the motor make the unit more electrically efficient. This drives a two-gear transmission that fits within a carbonfibre casing.

It's notable that the thermal efficiency of the silicon-carbide-switched electronics makes the cooling requirement for the package far smaller. Smaller radiators mean less drag, improving car performance.

NextEV continues to run two large pancake motors, which run slower but with far more torque. This means they can run without multiple gears.

Last year's motors underperformed and the rear subframe assembly was too flexible until makeshift reinforcements were added. This stiffened the rear, but came at a great cost in extra weight.

Now Omnigear makes the motors and inverter in house with support from Rational Motion, and the units are mounted transversely across the chassis in the same layout as DS Virgin last season.

This assembly features a differential between the two motors and the entire assembly is mounted inside a neat carbonfibre casing. Motor performance, stiffness and weight should therefore be resolved for season three.

CONSERVATISM AND COMPROMISES

It's surprising that every other team retains some aspect of the longitudinal season-one layout, albeit with different hardware.

Abt's 2015/16 package was a Schaeffler-developed motor mounted longitudinally and driving a three-speed transverse gearbox. These parts and the inverter were mounted inside aluminium bellhousing and gearcase castings.

Abt found its package so effective that it's hardly changed the layout, the car retaining the gearbox and motor.

Closest to the set-up from before is Venturi, which races a McLaren Advanced Engineering developed motor and inverter, now mated to a two-speed gearbox.

While the McLaren motor appears to be very similar externally to last year's, the inverter is a new design - somewhat smaller and with cable connections in different places, but still mounted on top of the battery.

Andretti made a false start to the 2015/16 season as the powertrain it had been developing with Houston Mechatronics failed to run reliably and the team had to abandon its development and run the season-one powertrain.

For this year a new powertrain has been developed around the Magneti Marelli motor, retaining the season-one layout with a longitudinal motor and two-speed gearbox. Weight has been saved with the use of carbonfibre and magnesium for bellhousing/gearbox casings.

WHAT WILL JAGUAR BRING TO THE TABLE?

This leaves new entry Jaguar as the other manufacturer with what can be described as a season-one layout.

It has adopted hardware listed as being its own build, albeit with Williams Advanced Engineering input along the way.

Perhaps more was expected of Jaguar's first electric powertrain, but having only signed to enter the series in December there was not the time to develop a new-generation layout before crash tests and initial testing were required.

With this compromise, what was designed as a four-speed gearbox is being run with just a pair of gears.

It's a compromise forced by timescales, but not a huge handicap for the team in its first year. Jaguar does not lack for resource and this is an area that will inevitably help it in the short- and long-term.

Not only will that give it the best possible platform for each event, but it will also pay it back handsomely in terms of development of the next powertrain.

If Jaguar has joined the likes of Renault and DS in sparing little expense in the pursuit of FE success - the suggested ballpark is more than €10million - then its first attempt at a powertrain will, at the very least, be respectable, regardless of the short lead time.

What it does over the coming months is one of several fascinating behind-the-scenes battles to be fought alongside the on-track squabbles that will begin in Hong Kong on October 9.