We Watched Lucid’s Engineers Tweak the 1200-HP Air Sapphire EV

From the April 2023 issue of Car and Driver.

You can’t let some curbing stand in the way of a good lap time, particularly at Virginia International Raceway, where it’s best to clip almost as many as possible to set a competitive lap. Snake is one such section, and the Lucid Air Sapphire prototype I’m piloting glides over the concrete serrations as if built to do nothing else and then rockets toward the Climbing Esses. More care and selective curb skimming work better here. Next, the tight Oak Tree corner offers another light dose of apex clobbering. The right-hand detour into VIR’s inner loop comes up quickly, but the Sapphire’s humongous 10-piston front brake calipers squeeze trackworthy carbon-ceramic rotors to haul this cruise missile smartly down to turn-in speed without a hint of squirm or squiggle. The car bends through the low-speed horseshoe smoothly with perhaps a hiccup at the exit. Then it’s off to Spiral and the diving left-right combination that comes just over a blind crest.

“Whoa! That was a big twitch.” There’s another slight power interruption when I graze the low curb at the exit of Spiral, but the car reasserts itself through the rest of the Infield. The Sapphire seems to love Roller Coaster, but as I exit Hog Pen and set my sights on the Front Straight, the car walks toward the left edge of the track, robbing me of the confidence to plant my right foot. With a full lap under my belt, it’s time to return to the pits for a debrief. After that, we’ll consider making some changes.

This process isn’t entirely new territory for me. Before journalism, the work that put food on my table was evaluating and tweaking the suspensions of prototypes in development at proving grounds, racetracks, and even public roads when routine driving character was under scrutiny. But my focus was always on tuning the hardware—springs, dampers, anti-roll bars, tires, bushings, etc.—because none of the suspensions I worked on had a line of software code associated with them.

John Culliton, ace driver and senior technical specialist in Lucid’s chassis and vehicle dynamics department, holds a job similar to the one I once had. He rode shotgun so we could discuss the car’s behavior on the fly. Once we untangle ourselves from the prototype’s five-point harness and duck under the full roll cage, calibration and validation manager Esther Unti joins us at the car’s trunk, which contains a raft of data-logging gear and a port into which she now plugs her laptop.

They seem gratified to corroborate my feedback with their own, which is important because it’s crucial to differentiate between what the car is doing and what a driver is doing. This is the first time the Sapphire has tackled VIR’s Grand Course, and a test crew must achieve a certain level of track familiarity before driver tuning can give way to vehicle tuning.

But this is far from the Sapphire’s first tuning session. Quite the contrary. Chassis and vehicle dynamics director David Lickfold and his team have visited numerous tracks and proving grounds. They’ve so far dialed in the specification of the purely mechanical bits—those things I used to tune—and the internal construction of the bespoke Michelin Pilot Sport 4S tires.

Lickfold and company have also settled on a base mechanical valving for the electronically controlled Bilstein dampers, which largely function without any electronic valve adjustment in the base Smooth driving mode. Software-controlled rebound and compression-damping valves add supplemental damping during extreme events in this mode, but they come into their own, automatically adapting with extra damping in Swift mode, then kick that up another notch in Sprint. The team has also landed on a power-steering-assist profile for each mode, as well as the mode-dependent behavior of the Bosch electromechanical brake booster that blends the massive friction brakes with the car’s high potential for regenerative braking.

View Photos

In garage number 3, engineers get a full debriefing on what worked and what didn’t work as they inch closer to a final calibration.

These basic elements feel nicely ironed out, and even though Lucid did not formally lock them down, they stand at near-production status. What does that leave for this VIR session? The engineers are still tweaking the damper software in Swift mode, but the Sapphire’s ability to shrug off curbs tells me they’re close. That mainly leaves something I never had to deal with in my engineering career: managing the torque produced by the Sapphire’s insanely powerful three motors.

A normal Lucid Grand Touring sports at each end a single motor containing a cleverly integrated differential within an impressively compact coaxial motor housing with outputs on the motor’s centerline. Each motor is capable of 670 horsepower, but the GT’s total 1050 horsepower is less than the combined sum of both. That’s because the GT’s 112.0-kWh battery can’t put out as much power as the Sapphire’s 118.0-kWh pack. This difference also clears the way for uneven high-load torque splits.

In the Sapphire, the same 670-hp unit resides up front, but the rear end contains a pair of those motors in a shared housing. Because each rear wheel is individually controlled, the rear axle has a scary 1340-hp potential, and the sum of all three motors is a staggering 2010 horsepower. Lucid will only cop to a total combined output of 1200-plus horsepower, with a definitive number some weeks away. But the difference between whatever that placeholder signifies and 2010 suggests a serious amount of torque-vectoring potential. That brings us back to Unti’s laptop and why Lucid is at VIR.

This powertrain gives Culliton and Unti a lot of tuning freedom, but they must account for many different performance-driving scenarios. In a straight line, there is nothing stopping them from sending all the battery can muster to the rear axle—except sanity and physics. You’ll always get a quicker and more stable launch when more than two tires put torque to the ground. To that end, they tell me that about 75 percent of launch torque goes to the rear. Even at the drifting end of the spectrum, they still send a bit of torque to the front “to make slides more predictable and achieve higher sideslip angles.” Cool.

VIR’s Grand Course lies between those two performance extremes, with a variety of corners. That’s what makes it such a great location for our annual Lightning Lap test and a useful place to sort out the Sapphire’s complex torque vectoring, which goes way beyond simulating the differential action that comes naturally to a single-motor drivetrain. Torque vectoring in an EV with two motors on one axle can generate force more quickly, precisely, and smoothly than commonplace brake-based torque vectoring. And the degree of split is highly variable because one motor can regenerate power while the other drives forward.

Unti’s laptop, loaded with Lucid-developed control programs, manages all of these settings. In fact, some 90 percent of the lines of code within the Sapphire were written in-house. The application Unti is working with today allows her and Culliton to alter parameters that subsequently rewrite underlying code accordingly. For example, the wheelbase module reallocates torque to mimic the feeling of a physically shorter or longer wheelbase, depending on conditions.

Culliton and Unti punch in a few tweaks to address issues I encountered. For high-speed exits, such as when coming out of Hog Pen, they up the level of rear torque vectoring to add a bit of rotation as I transition from part throttle (they use the word “throttle” as opposed to “accelerator” or some other unsatisfying term, so I will too—this once) to Front Straight acceleration, so I get past the feeling that adding more will increase understeer. In the dramatic Spiral transition, they soften vectoring and lessen the turn-in moment so as not to overwhelm the rear tires. And in the scenarios where the back end twitched on corner exit, they add a bit more front torque bias for rear stability.

They didn’t tell me any of this before I went out in the Sapphire again, lest I anticipate the effects. But the adjustments worked a treat. The two minor corner-exit wiggles disappeared, and the car was much more stable when I crested the low-speed transition in Spiral. Best of all, I was better able to feed on the power as I exited the final turn. The phrase “lack of confidence” disappeared, replaced by “my God, this thing is quick.”

There was a side effect, though. As is often the case in tuning, when you improve one thing, you lose out somewhere else. That happened in Roller Coaster, where the Sapphire became a bit twitchier. A mistake on my part might have brought that about as I sailed a few feet past my usual turn-in point and tried to bring it back. The car seemed to try to help, but here two wrongs did not make a right. Back to the pits for another think and software tweak.

At this point, my time was up, but Culliton, Unti, and Lickfold were going to remain at VIR for another round or two before packing up and moving on to another testing locale. It’ll be great to see where the Lucid Air Sapphire ends up when they’re finished, but things look very promising indeed. Let’s hope we can get our hands on one for the next Lightning Lap.

Sapphire vs. Plaid

It’s tempting to look at the Tesla Model S Plaid and the Lucid Air Sapphire through the same lens. After all, each has a tri-motor powertrain that produces upward of 1000 horsepower. But there are fundamental differences.

For one, the Sapphire was part of the Lucid Air game plan from the start and will appear in the model’s second year of production. David Lickfold told us his team knew of the need to put down all that massive power and torque when they initially conceived the suspension. The same is true of the goal to make the car utterly trackworthy, which means its 10-piston front calipers and carbon-ceramic brake rotors are no afterthought.

Meanwhile, the Plaid emerged in the Model S’s 10th model year. It’s doubtful that Tesla planned the Plaid’s existence that far out, especially since the company built its brand around the induction motor that Nikola Tesla invented, a stylized cross section of which is Tesla’s logo. But in 2019, Tesla quietly started abandoning induction motors in the Model S, switching to permanent-magnet synchronous ones, and the Plaid wouldn’t have been feasible if that hadn’t happened.

What’s more, the Model S Plaid shares tires and brakes with the less powerful variant. All- season 19-inch tires remain standard, with 21-inch summer rubber an option. The Plaid is mainly a motor swap. A $20,000 carbon-ceramic brake upgrade kit is planned for later this year, but it should have been part of the Plaid package from the beginning.

Technical Editor

Dan Edmunds was born into the world of automobiles, but not how you might think. His father was a retired racing driver who opened Autoresearch, a race-car-building shop, where Dan cut his teeth as a metal fabricator. Engineering school followed, then SCCA Showroom Stock racing, and that combination landed him suspension development jobs at two different automakers. His writing career began when he was picked up by Edmunds.com (no relation) to build a testing department.