Tesla Model 3 performance put to the test on Laguna Seca raceway

[COLOR=rgb(39, 38, 36)]Saw this earlier today via link to Jalopnik.
Watched about half of it, up to the point brakes started to be ‘scary’… (watching track is like watching tennis… after a while, you kinda would like to be at the wheel… ;))

Couldn’t help but noticing that he seemed to have a particular challenge with the one (right turning) curve before the blue Cooper Tires overpass… in general, wasn’t impressed by his driving skills…

With regards to brakes, trust it is related to weight and inadequacy of stock brake pads for this purpose… Also wonder if you could get a bit of extra longer brake performance by optimal use of regen…

Not that I’m an expert… :D[/COLOR]

Not sure I understand the Brakes story here. I drive a (beloved) Tesla Model X 100D since last September, and one of the most noticeable uniques is that thanks to its very well tuned regenerative braking you almost never use your real brakes, others that in the last meters to fully stop the car, hence you expect them to never wear, and have to replace the pads only after several years of usage, and the disks never.
Understand in the circuits when pushed to the maxi speed, they may use the brakes more often, in particular if they start with a FULLY Charged battery, as regen is limited when battery is full (or in very cold conditions). But still wearing them in a few tracks sounds un-expected to me.

Thanks for this. I am impressed. I wouldn’t have even considered tracking my (May18?) 3 but hmmm. Upgrade the pads.  Now i don’t need to buy that used Miata.:-) Granted he’s on excellent after market tires/springs, but impressive nevertheless. And 9 laps at 40F with some elevation changes(?) 

This is what I have been waiting for! The Model 3 uses synchronous motor with rare earth magnets (like most other EVs) instead of the Induction motors used in previous models. This motor should run significantly cooler. This will be a real eye-opener when AWD high performance model is available. Bodes well for the Roadster as it should be able to run a full 50 mile race!

Not sure I understand the Brakes story here. I drive a (beloved) Tesla Model X 100D since last September, and one of the most noticeable uniques is that thanks to its very well tuned regenerative braking you almost never use your real brakes, (…)
But still wearing them in a few tracks sounds un-expected to me.

Pat, first and foremost to your X100D, let me just say: lucky you..! You’re in a place most of us can only dream of at this point… 🙂

On the topic of brakes, clearly those were extreme conditions which do not compare with every day use. In addition, if you dig in the original details to be found on TMC (courtesy of M3OC member Macrow69) https://teslamotorsclub.com/tmc/threads/model-3-track-day-laguna-seca.109818/, you’ll find that he has his regen set on low, which certainly did not help (credit for the TMC search to another M3OC user and moderator Garsh… I don’t go on TMC as a matter of principle… ;)).

Just for sharing…

On the topic of brakes.. you’ll find that he has his regen set on low,

Thanks for pointing that out, but why on earth would anyone racing have it set at anything but highest level available? Maybe he forgot about it?

Edit: Apparently it was deliberate. He did not expect the brakes to fail, and wanted to do some follow up laps with the regen on and compare the differences. He just never got to the second run.

[COLOR=rgb(39, 38, 36)]Couldn’t help but noticing that he seemed to have a particular challenge with the one (right turning) curve before the blue Cooper Tires overpass… in general, wasn’t impressed by his driving skills…[/COLOR]

Matt is a very good driver. Those were fast laps on a partially wet track in a car that he’s never had on a race track before. He is very smooth, which is the mark of a skillful and fast driver. Smooth is fast in racing because at the outer limits of tire adhesion, any disruption in any direction will cause a tire to lose traction and thus the car speed. Smooth is fast.

While it’s aesthetically pleasing and good practice to be smooth on public roads, none of this applies in normal street driving, except in emergencies, which most drivers are therefore woefully underprepared for. All drivers should have racing or track experience to better understand what to do in an emergency, and almost none of the drivers on the roads (globally) have that experience. Track days like this, and racing and high performance driving schools would save thousands of lives if more people had that experience, even if they never actually raced.

The reason the car understeered (slipped away from the inside of the turn) at turn 3 is because there was water there. Matt controlled the car very well when that happened. This is also the mark of a skillful driver.

JeffC makes several good points. One thing not mentioned is that heavy regen on the 3 is currently a bigger problem because it can (currently) only regenerate on the rear wheels. The creates even more serious balance/traction issues and limits how much regeneration you can produce to something well below the limits of braking since hard braking unloads the rears where all the regen is happening whereas the majority of the traction for hard braking is in the front. A dual motor system could in principle be set up to provide track-worthy regenerative braking, and whether that were delivered as single-pedal driving or through the brake pedal is somewhat a user interface issue: there is no reason why you can’t have some regen on lift for track driving with proper adjustment of technique.

That brings us to technical limitations: The power being dissipated under full braking at 100 mph in a car with the weight of the Model 3 is roughly a megawatt. Unless the car is set up to take those kinds of charge rates (the 3 most definitely is not) then it’s simply not possible to provide the deceleration you need at the track by capturing that energy back to the battery. Even when it becomes possible, the large non-conservative forces at work at the track (resistive and aerodynamic) will mean track use really kills EV range. For those that have driven at the track (my track car gets 4 MPG at Laguna Seca, but 27 MPG on the freeway) this will not be a surprise!

BTW, I agree that making making regen on lift optional is a good thing (as GM did with the Bolt – which does very well integrating regen with friction brakes through the brake pedal).

Not sure I understand the Brakes story here. I drive a (beloved) Tesla Model X 100D since last September, and one of the most noticeable uniques is that thanks to its very well tuned regenerative braking you almost never use your real brakes, others that in the last meters to fully stop the car, hence you expect them to never wear, and have to replace the pads only after several years of usage, and the disks never.
Understand in the circuits when pushed to the maxi speed, they may use the brakes more often, in particular if they start with a FULLY Charged battery, as regen is limited when battery is full (or in very cold conditions). But still wearing them in a few tracks sounds un-expected to me.[/QUOTE] [QUOTE=»Michael Russo, post: 13136, member: 5998″]On the topic of brakes, clearly those were extreme conditions which do not compare with every day use. In addition, if you dig in the original details to be found on TMC (courtesy of M3OC member Macrow69) https://teslamotorsclub.com/tmc/threads/model-3-track-day-laguna-seca.109818/, you’ll find that he has his regen set on low, which certainly did not help (credit for the TMC search to another M3OC user and moderator Garsh… I don’t go on TMC as a matter of principle… ;)).

Correct. Street driving and track driving place extremely different loads on all of the vehicle systems, especially brakes and tires. One of the most common modifications when taking a street (production) car to a race track is to upgrade the brake pads to racing pads. Tire/wheel upgrades are probably second most common. Upgrading the front brake disks is also common. (Rear brakes don’t get as much power due to load transfer to the front wheels under braking.)

When going all out on a race track, all of the acceleration available from the motor is used whenever possible. All of the brake power is used whenever possible. All of the lateral and longitudinal acceleration from the tires is used. This is very hard on all of those systems, but especially brakes, which on a street car are almost always undersized for racing.

Matt upgraded his wheels, tires and springs, but not his brake pads. He’s already searching for better brake pads.

Let’s not get carried away… these NSX and 911 lap times… they are in the wet (marked w on the fastest laps website). Those cars are 1:3X cars at Laguna in these conditions. For comparison, a Ford Fiesta ST has run 1:49 there and 1:5X is typical for hot hatches and Miatas.

Thanks for pointing that out, but why on earth would anyone racing have it set at anything but highest level available? Maybe he forgot about it?

Edit: Apparently it was deliberate. He did not expect the brakes to fail, and wanted to do some follow up laps with the regen on and compare the differences. He just never got to the second run.

Taking the last question first, Matt was forced to use regen at the end because the friction brakes had worn out. Regen was the only way left to slow down in a short distance coming back into the paddock.

Taking the first question, he did not forget about using regen. Heavy regen on throttle lift is inappropriate for track use and also inappropriate for emergency use on the street for basically the same reason. (Racing reaches the limits of adhesion in ways somewhat similar to what happens during an emergency on public roads and can be useful for understanding what happens then and why it happens.)

Many Tesla fans (of which I am most definitely one) will be confused by this, but Tesla made a major mistake by putting heavy regen on the throttle. Briefly, this makes something called trailing throttle oversteer much worse, and is therefore an ergonomic, safety, human factors and vehicle dynamics error. Trailing throttle oversteer is what happens when a driver suddenly lifts off the accelerator hard enough to transfer load to the front wheels. Reducing the corresponding load on the rear wheels also reduces their traction and the rear wheels can start to yaw faster than the fronts, leading to the back end «spinning out». There are several names for this: oversteer, «getting loose,» etc. One reference: https://en.wikipedia.org/wiki/Lift-off_oversteer

Putting heavy regen on the throttle makes trailing throttle oversteer worse if it transfers even more load to the front wheels when an untrained driver lifts off the throttle suddenly in an emergency, as almost all do. It’s human nature and basic psychology for someone to lift violently off the throttle or jam on the brakes in an emergency, if they don’t freeze up and do nothing, which is also very common.

I fully understand several reasons why Tesla did this (and better and deeper than most people). In their simulations, for untrained drivers, Tesla found that heavy regen on throttle is more energy efficient during non-emergency street driving. «Single pedal driving» has been touted as a feature from the beginning, reinforcing this idea to the point where it has become orthodox (at least at Tesla; see below). Orthodoxy is not always right.

And yes, I do understand the benefits of single pedal driving. Personally I find left foot braking much better for cars with no clutch, as do most racers I know. https://en.wikipedia.org/wiki/Left-foot_braking

Additionally, regen on throttle was carried over from AC Propulsion’s cars, including the T-Zero on which the original Tesla Roadster was based. Alan Cocconi, the engineer behind AC Propulsion and therefore at least indirectly responsible for Tesla existing, is a friend of a friend. Alan is also the engineer who designed the motor controller electronics on the AeroVironment Impact, which led to the GM EV1, the other major inspiration for Tesla existing according to interviews with Elon and Martin Eberhard.

The better alternative is to fully integrate regenerative braking with the brake pedal, as GM did on EV1, Toyota the Prius and their original RAV4 EV (which we still drive today), VW eGolf, Nissan Leaf and most other car makers. This integration is very hard to do well; the original Prius did not to it well; later generations did.

As Elon correctly notes, things should be reasoned from first principles. Unfortunately I’m not sure he is aware (enough) of the first principles of ergonomics, psychology, human factors engineering, and vehicle dynamics that I mention here. And yes, I have tried to contact him about this. Zero response. I did also talk with a Tesla engineer a few years ago who said at the time that Tesla engineers were split roughly evenly about this issue internally.

Tesla should correct this error and fully integrate regenerative braking with the brake pedal, or at least make it a user selectable option. The above is only a very brief summary.

I challenge my fellow Tesla and EV fans to keep an open mind about this, honestly consider the alternatives, and always be respectful of thoughtful people who have different ideas. Groupthink can be a negative. Respectful consideration of other ideas is usually a positive.

Please note that I have been an EV driver and supporter for 20+ years (since EV1), and I am buying a Model 3 since it’s the best car since the EV1. I fully support all EV makers especially Tesla, and I’d like this error to be corrected.

P.S. Tesla needs to correct this for the very powerful new Roadster if they expect it to be taken seriously as a performance car.

tknightnelson brings up several important points. First, keep in mind that typical street driving sees a small fraction of the power and forces of racing and emergencies on the street. For street use, regen on the rear wheels (for example for a rear motor EV like the initial Model 3 or early Model S) is only a slight efficiency loss over having it on the front wheels. Tesla increased the overall efficiency of their cars using dual motors, and part of the reason is almost certainly due to being able to get regen on the front wheels, where it’s more efficient. (It’s more efficient because there’s more load and therefore more efficient coupling of power on the front wheels during braking, as I suggested above. Dual motor is also more efficient power coupling to the road during acceleration, even under light street driving, which is why dual motor Teslas are more energy efficient than single motor.)

So ideally, regen would be on the front wheels due to greater load on the front wheels during braking. One the reason for Tesla using regen on throttle lift may be the (single) rear motor (of early cars), but even with only a rear motor, regen could be integrated on the brake pedal instead of the accelerator since the forces and power levels of normal street driving are so low. They are low enough to not upset the balance of the car if (rear only motor) regen were put on the brake pedal. For racing, regen is much better on the front wheels.

Keeping in mind the first point, braking on a race track or in emergencies on the street can generate far more power than EV motors or batteries can handle, so at least some friction brakes must be used. As tknightnelson mentions, the power can approach the megawatt range, which friction brakes can handle briefly, but today’s typical EV motors and battery packs can’t.

Also, due to high loads on the race track, as tknightnelson correctly notes, rear wheel only regen power as on the single motor Model 3 and early Model S, would be limited due to heavy load transfer to the front under braking. It might have a tendency to unbalance the car, if not implemented very well. I’m sure part of that is a reason Matt did not use Model 3’s regen on the track.

It’s definitely the case that the high power, high loads and high losses of racing will reduce the range of EVs AND combustion cars alike. Going all out on a race track burns through fossil fuels as much as it does EV battery charge. Matt’s Model 3 heavy battery charge usage at Laguna Seca was similar to a combustion car’s heavy fuel usage.

Regen could be useful on the racetrack to recover some braking energy, but it would be limited by battery and motor power capacity, and it must be integrated correctly to be usable and not upset the balance of the car. Heavy regen really does belong integrated on the brake pedal for that, particularly due to trailing throttle oversteer being made potentially much worse by putting it on the throttle, as Tesla does under «standard» regen. The same applies to emergencies during street driving, for related reasons.

(As we all know, regen is no where near to 100% efficient (since that would be a perpetual motion machine), but it does usefully recover some braking energy that would otherwise be wasted as heat by the friction brakes.)

tknightnelson brings up several important points. First, keep in mind that typical street driving sees a small fraction of the power and forces of racing and emergencies on the street. For street use, regen on the rear wheels (for example for a rear motor EV like the initial Model 3 or early Model S) is only a slight efficiency loss over having it on the front wheels. Tesla increased the overall efficiency of their cars using dual motors, and part of the reason is almost certainly due to being able to get regen on the front wheels, where it’s more efficient. (It’s more efficient because there’s more load and therefore more efficient coupling of power on the front wheels during braking, as I suggested above. Dual motor is also more efficient power coupling to the road during acceleration, even under light street driving, which is why dual motor Teslas are more energy efficient than single motor.)

So ideally, regen would be on the front wheels due to greater load on the front wheels during braking. One the reason for Tesla using regen on throttle lift may be the (single) rear motor (of early cars), but even with only a rear motor, regen could be integrated on the brake pedal instead of the accelerator since the forces and power levels of normal street driving are so low. They are low enough to not upset the balance of the car if (rear only motor) regen were put on the brake pedal. For racing, regen is much better on the front wheels.

Keeping in mind the first point, braking on a race track or in emergencies on the street can generate far more power than EV motors or batteries can handle, so at least some friction brakes must be used. As tknightnelson mentions, the power can approach the megawatt range, which friction brakes can handle briefly, but today’s typical EV motors and battery packs can’t.

Also, due to high loads on the race track, as tknightnelson correctly notes, rear wheel only regen power as on the single motor Model 3 and early Model S, would be limited due to heavy load transfer to the front under braking. It might have a tendency to unbalance the car, if not implemented very well. I’m sure part of that is a reason Matt did not use Model 3’s regen on the track.

It’s definitely the case that the high power, high loads and high losses of racing will reduce the range of EVs AND combustion cars alike. Going all out on a race track burns through fossil fuels as much as it does EV battery charge. Matt’s Model 3 heavy battery charge usage at Laguna Seca was similar to a combustion car’s heavy fuel usage.

Regen could be useful on the racetrack to recover some braking energy, but it would be limited by battery and motor power capacity, and it must be integrated correctly to be usable and not upset the balance of the car. Heavy regen really does belong integrated on the brake pedal for that, particularly due to trailing throttle oversteer being made potentially much worse by putting it on the throttle, as Tesla does under «standard» regen. The same applies to emergencies during street driving, for related reasons.

(As we all know, regen is no where near to 100% efficient (since that would be a perpetual motion machine), but it does usefully recover some braking energy that would otherwise be wasted as heat by the friction brakes.)

Indeed… all agreed. Some finer points about what’s possible *in principle* and when we could get there:

Aerodynamic losses are the piece you can’t do much with. Those losses can be minimized through careful design, but not recovered like the losses to inertia which can be partially recovered through regeneration. As we all know, EVs get worse «mileage» at freeway speeds than city driving, unlike ICE cars that don’t recover any kinetic energy, and that’s precisely because of the inability to recover aerodynamic losses. Those losses rise exponentially (squared in distance but cubed in time) so at track speeds they become huge – 4 times larger per mile at 120 mph then at 60 mph. So, just driving a constant 120 without ever having to bleed off speed means 1/4 the range and there is nothing you can do about that.

Things get worse when you you have to move energy around, even if you don’t waste it with friction brakes. The losses there relate to electrical resistance: in the charging system and the internal resistance of the battery. Proper design can enable higher regen rates, where many of the design challenges are the same as those that must be resolved to deliver that amount of power for acceleration. That is to say, when you have a 1000 kW (or 2000 kW) car that weighs <4000 lbs., many of the systems will be designed to take similar power in regen needed at the track. Whether or not they can do this in constant cycle (at the track you are using either max acceleration or max braking perhaps as much as 50% of the time) is a big issue, but one the can be resolved with proper design, especially w.r.t. cooling to remove the heat generated by putting huge current through the motor, charging system and battery. We are starting to see designs now that have numbers in this range for power output, so I would expect an EV "supercar" to be a possibility before too many more years pass. To be sure, Tesla will be intending to solve these problems with the Roadster.

Agree with everything you said too. Electrical, mechanical and chemical losses apply twice for regen; first when converting the motion energy from kinetic to electrical to chemical for deceleration, then second when converting it back in the other direction for acceleration. It’s always better to avoid the losses and coast whenever possible, obviously only for street driving.

(As an aside, that’s another reason I’m against regen on throttle; it generally makes it harder to coast. I know it is possible to coast on Teslas since there’s a dead zone in throttle position between acceleration and deceleration provided for coasting. A different (and better IMO) solution is to put regen on the brake pedal and have an optional coast down mode as EV1 did. That was slightly geeky, but did work.)

The cool thing is that some of the efficiency improvements learned for super high performance cars like the new Roadster may also apply to lesser cars to some extent. Some obvious things like larger diameter power wires for lower resistance won’t translate to lesser cars though. Higher voltage (if used) may not either. (They may not be cost effective to use in lower performance cars.)

The best way to reduce aerodynamic losses is to design for low drag. The cost of low drag design is only applied once, but multiplies over all the time driven at higher speeds over all of the cars ever built for that design. It’s a very large energy efficiency multiplier. 😀 Tesla came near to the EV1 with Model 3, but could not quite match its coefficient of drag.