Lotus throws the Evija driver under the hypercar
The Lotus Evija X has crashed at Goodwood. It’s barely damaged, and the driver is okay, but there’s important lessons to be learned if only we had the data.
Many cars crash at Goodwood, and the Evija clip has gone viral because everybody loves a smash, particularly a 2.3 MILLION DOLLAR HYPERCAR which is according to some commentators, the fastest car in the world. It isn’t.
Anyway, the reason this crash is of interest is because the Evija is very different to the usual hypercars. First, let’s recap the usual petrol drivetrain – one engine driving the wheels, four in this case, via a series of shafts, cogs, differentials and gearboxes. There would also be some sort of clutch or differential between the front and rear axles, computer-controlled to apportion torque, and there would be a variety of electronic driving aids such as selective braking of each wheel to augment driver control, and maybe some clutch packs to distribute torque between axles. However, fundamentally you have one motor turning rods of metal connected to wheels, and that’s very predictable, and very well understood over many, many years.
The Evija is different. It’s a 2000hp IWD EV, or indvidual-wheel-drive, with one motor at each wheel. No shafts, diffs, clutches. This means it has what I call a software transmission – computers control how much power and torque is sent to each wheel. I explain all that here, with reference to 4x4s and the Cybertruck:
This concept of IWD is new to the car world. There are very few IWD vehicles, and they are very new. And we all know it’s incredibly difficult to program any computer for all eventualities – ever had anything computerised lock up, freeze, glitch? Apple CarPlay, Android Auto? Anything else? Well, that’s software, and now we have a software transmission. I explore the crash implications further in the video below:
Now I’m going to analyse this article on Goodwood’s official website. I’m going to assume that it was written with the cooperation of Lotus as it has information only Lotus could have provided. And by the way, I love Lotus, I own one myself. And I think IWD is the way of the future. Now onwards:
We were all very excited to see the dynamic debut of Lotus’s most extreme creation, the Evija X, at the Festival of Speed presented by Mastercard. A one-off track car intended to showcase the expansive possibilities of the brand’s electric powertrain technology. Things unfortunately didn’t quite go to plan, but not for the reasons you might think…
“not for the reasons you might think” – journalists such as myself, and in fact the livestream commentator immediately pointed out that it looked like a software fault. So the writer begins to ease us away from that viewpoint. If on the other hand you just looked at the crash and thought “what a dickhead”, then the line serves to draw you in, a nice lede doing two jobs.
The Evija X had already been out for its debut run at Goodwood on Thursday morning as part of the Supercar Run, and everything had gone very much to plan. After a successful burnout away from the line it made its way up the Hill with little drama, and everything was set for more of the same when it returned in the afternoon.
Now an emphasis on how well things had gone.
The Evija X had already been out for its debut run at Goodwood on Thursday morning as part of the Supercar Run, and everything had gone very much to plan.
Again, everything was great, no issues, all to plan.
Let’s first take you back to one week ago, when Lotus introduced the driver to the car they would be driving at the Festival of Speed. Planned were a number of testing days at the manufacturer’s test base at Hethel, a chance to familiarise with the car’s systems and astonishing performance. Based at a race track, however, the requirements of both car and driver were fundamentally different to what would be asked of them on the Goodwood Hill.
Really? The idea was to prep for the Festival of Speed, which is a hill climb, and not even much of a hill or one with tight corners. Now Hethel is not a hillclimb circuit, but it’s a racetrack, so you can easily do standing starts and a one-minute dash. This gives practice in starts and cold-tyre running. I’m sure that’s what they would have practiced, runs for the FoS, not continuous lapping. So I don’t buy this line.
The Evija X is no ordinary car. It may be based on the same tub as the road-going Evija, with the same transmission and electric powertrain, but the engineering around those elements is entirely focused on maximising lap time with stronger brakes and slick tyres. It’s a car designed to maintain traction and grip rather than develop a wall of tyre smoke, and in almost every way entirely unrecognisable from its road-going namesake.
I don’t doubt it’s track-focused and agree about the laptime focus not the wall-of-tyre-smoke focus.
Coming up to the Start Line then, the act of turning off the traction control was already an alien concept to both car and driver. It also means there is absolutely zero response from the car in the case of tyre slip. The driver is very much on their own and in full control of the car.
If it was an “alien concept” to both then there are two possibilities. First, there was no plan for a burnout on start. This seems unlikely, as Gavan Kershaw spun the wheels on his run. This screenshot is from his start. That is, shall we suggest, more tyre smoke that optimal for a quick run.
Presumably the traction control was off for that, or at least in a reduced mode. The crash-run driver is not known for certain, but Lotus would not have let just anyone behind the wheel so we know it’d be someone able to handle a performance car, not a driver to whom lack of traction control is an “alien concept”. I also find it unlikely, but not impossible, a pro driver went against an order not to disable the traction control. And Lotus might have been able to enforce it on, or partially on. This paragraph leaves me with questions, and I feel it is designed to distract.
As we saw on the Evija X’s first run, that doesn’t make it an undrivable beast, but it does make it far less predictable. You’d certainly want to be working from a consistent base when the time comes to put your foot down.
Delicately walking the line between not saying the Evija X is terrible with the electronics off, but definitely starting to point that finger of blame.
Before that fateful second run, however, the build-up was stifled by a stalled Solus GT, which meant the Evija X had to circumvent the stationary car in order to line up, at a slight angle, on the Start Line.
This would not have helped; you want to launch straight, and have equal weight on both wheels on a given axle. However, the car does look lined up straight, and drove straight for a solid second before it yawed out of control. Not sure I buy this either. Look at the still below, seems like pretty straight lines down the centre of the road before things go wrong.
That didn’t stop the driver from fully committing when given the signal to go. Data traces show that they immediately jumped to 100 per cent torque demand from all four of its electric motors. That’s an instant injection of 1,704Nm (1,257lb ft), not limited by a rev limiter, which within a few metres had spun the rear wheels up to 170mph and the fronts to 150mph. The resulting burnout was quite something, while it lasted.
A demand of 100% torque is not the same as a supply of 100% torque. Modern cars torque-limit in many conditions such as lower gears. Nevertheless, that’s a lot of torque. But the car gets off the line okay and straight. It doesn’t immediately spear off. Oh, and “That didn’t stop the driver…” the finger of blame is pointed a little more directly. Nice work. It would also be interesting to compare how much torque was demanded by Gavan on his run, and in fact to compare the data traces across both runs.
Now why were the rear wheels at 170mph and the fronts at 150mph? That’s destabilisation right there, it’s actually how radio-control drift cars are set up to drift, the rear wheels turn faster than the front. Rock crawl buggies over-drive the front axle for climb stability. I would suggest that the software drivetrain allowing the rears to spin that much faster than the fronts is an error, and would have made the car very hard to handle. I suspect this contributed to the crash. One of the commenters on my video made a great point; I noted there was next to no steering input to recover, and he thought maybe if the car thought it was doing 150mph it’d have speed-sensistive drive-by-wire steering which would reduce steering wheel angle. Possibly.
The reason it didn’t last was quite a simple one. The driver unfortunately lost control.
And here we go, having led up to this point and prepared the ground, the Dump Truck of Blame is backed up and unloaded squarely on the driver. Who to be fair, might be at fault in part. But no error on the car at all? Nothing? How’d it accelerate backwards then?
“It’s computer software issue this one, it isn’t driver error” said commentator Harry Metcalfe, in reference to the invidual-wheel-drive powertrain, and I agree.
The Evija X is capable of reaching a top speed of 240mph, and the energy involved in that power delivery was simply too extreme.
Top speed is not a reliable indicator of how hard a car is to control.
As the driver begun to lift off the throttle to try and search for some traction, the balance of the car shifted and pitched it into the bales.
I don’t follow this. Why would that pitch the car into the bales? Lifting off is a front-rear pitch down, not left-right.
In total, the incident from the start to the impact was about 1.5 seconds. The driver was unfortunately unable to maintain control.
And another jab at the driver. For pro racing drivers. 1.5 seconds, even 0.5 seconds is an eternity. Cars are usually easier to control at slow speeds burning out than at high speeds correcting oversteer.
In plenty of cases, much like the Evija X, the car itself is not designed for these tyre-shredding moments either.
So why do it, on the earlier run? Did the driver disobey orders or go too far? This implies the driver did something he was not authorised to do. Remember, the previous driver shredded tyres, and this crash run was not timed – a burnout is not the fastest way to get off the line. Some have said it was to warm tyres; no, not a timed run, and the burnout was way beyond tyre warming, and too far off the start line.
The good news of course is that nobody was hurt. The integrity of the cockpit was unaffected and the driver walked away. The car in fact was, aside from a single bodywork panel, entirely undamaged, and would have been able to run again if not for post-incident protocol which requires the car to be properly checked over.
Letting everyone know the car was not wrecked, written off etc. I think that’s good, testament to the engineering and the haybales!
We’ll appreciate the efforts of these drivers a little more from now on, as they deal with the challenge of controlling 2000PS track monsters like the Evija X in alien circumstances, within the emotionally charged confines of the Goodwood Festival of Speed.
“Alien circumstances”, again. I don’t think so.
IWD is a new technology with huge potential for car dynamics, and equally huge potential for disaster if things go wrong, so it would be good if Lotus was as open as possible about what happened with the Evija X so we can all learn, rather than throwing their driver under the hypercar. As it is, the void left by the lack of information is being filled with rubbish from the anti-EV crazies, and the pro-EV zealots are happy enough to throw it all onto the driver. That’s not how we move forwards, learn, and start to realise the full potential of IWD and EVs.
Update 22/07/24
I’m lucky my videos and posts attract informed commentary, and so far, nobody who appears to have an automotive engineering or performance car background seems to think it’s the driver’s fault. Here are two sample comments, first from @grantminshull7689 on YouTube:
The lack of front wheel steer during the fast yaw to the right is probably the result of insufficiently powered electric steering given the enormous centrifugal force generated at each front wheel, spinning an unusually fast RPM at low road speed. (More interesting debate can be had as to why the same system would probably work better at higher speeds)
The sudden change in direction appears to happen as the driver withdraws from full-throttle. I’m suggesting that the left rear snatched grip (a subtle difference in road surface is enough) whilst decreasing it’s RPM and still producing a massive torque, while the right rear was still spinning much faster than road speed and producing next to no forward drive. Following this the sensor-control equipment appears to be either faulty / badly set / badly designed or any combo of this. It will be challenging to find a set-up possible to counter such a violent snatch in such a narrow road to avoid a crash. Better software would avoid this in the first place.
A note on conventional diffs. The inertial properties of a conventional mechanical differential system naturally mitigate against this sort of effect when one wheel begins to snatch – the sudden de-acceleration of one wheel is countered by the acceleration of the other wheel. In other words energy would be transferred into the rotational speed of the opposite wheel instead of yawing the car so violently to the right, as happened.
Lotus are being disingenuous to the driver in my opinion. His reactions looked incredibly quick and actions visually correct. There is a lot of development work yet to be done on stabilisation systems of such powerful cars. Lotus owes this man an apology.
And from Greg Milligan on Facebook:
For the Lotus explanation to hold, it would need uneven traction side to side and one side to grab traction.
The video clearly shows all 4 wheels with complete traction loss and spinning at high speed.
The lurch to the right is extreme and sudden, and happens while the left side wheels are still spinning at high speed with very little traction and all 4 wheels are on a similar surface.
That alone makes very little sense.
Next, the path followed by the car is wrong…if that were the case it should curve rapidly towards the wall, hit, and the momentum of the vehicle weight should cause it to pivot around the contact point with the wall as the rest of the car slides around.
This again is not what happens.
The pull to the right is extreme, and the car backs off the wall, pivots around its centre, before then going backwards into the wall…it behaves much like a Bobcat loader spinning in place, and it appears that at least the rear right wheel is turning backwards whereas it should still be spinning forwards at high speed.
Everything points towards the rear right wheel either instantly stopping or even reversing, pivoting the car into the wall then pulling it out, around, and backwards into the wall again.
I’ve watched probably thousands of crash videos, ranging from dashcam highway videos to motorsport on TV and in person, as well as the odd spin in karts myself, burnouts gone wrong, asymmetric traction from broken axles etc…
The physics are all wrong for what is claimed by Lotus.
