Car suspension tuning is a very subtle and difficult black art. It's all about what's happening dynamically in response to transient inputs (go around a corner, hit a bump, etc.).
I believe Lotus does a lot of engineering consulting helping other manufacturers dial in suspensions. I wouldn't be surprised if Tesla had them consult.
Even some of the big car companies have trouble dialing in suspension on their own (I'm looking at you underdamped Japanese cars). European brands tend to be really good at this, and can take the same raw materials that everyone else is working with and create superior setups. Porsche somehow manages to start with the 911 where they stick the engine way out behind the rear axle (high polar moment of inertia, poor front/rear weight distribution) and make it handle beautifully. VW cars also tend to have great suspension tuning. Being able to keep the weight low to the ground in the Tesla contributes to a low center of gravity and I imagine must really help enormously.
Also the body under-tray is beautiful, like something you'd expect to see on a purpose built race car, probably helps a lot with the aerodynamics. I imagine more and more cars will get similar under-trays to improve aerodynamics and fuel efficiency.
It's interesting there is strong anti-dive geometry built into the car, probably to avoid problems with throttle lift oversteer caused by inexperienced drivers. I wonder if the current Model S handling leaves anything to be desired when driven to it's absolute limits and how the car would handle with some of the anti-dive geometry removed. Removing this geometry is a common mod on some cars.
"Porsche somehow manages to start with the 911 where they stick the engine way out behind the rear axle (high polar moment of inertia, poor front/rear weight distribution) and make it handle beautifully."
Actually there are some benefits to the arrangement. The high polar moment of inertia actually should make any spin slower to start and easier to catch than a mid engine arrangement. Under hard braking the weight shifts forwards so the weight distribution changes to more balanced than with other arrangements. Under power the weight is over the rear wheels for power out of the corner.
A 911 behaves differently to other cars but I'm not sure the arrangement is actually worse in most dynamic scenarios. Understeer into the corners is the main dynamic weakness when pushed really hard and can be compensated for by trail braking into the corners. I guess you can get into trouble if you go into a corner too fast, panic brake and lose the rear end but if you are used to the set up you can make use of it.
[I've only driven a 1980's 911 so maybe I'm over extrapolating to all rear engined cars and I've never tried a mid engined car.
This book is really good at explaining how do get the most out of different weight distributions and has some fun stories (I've got the first edition but it probably isn't much different):
http://www.amazon.co.uk/Porsche-High-performance-Driving-Han...]
i've owned an NA2 NSX and a 997 911 GT3 and the NSX was much easier to lose control in. it felt unstable in general.
911 grips harder as you turn harder, very strange to get used to but once you 'get it' it feels great. obviously there is a limit but you are not going to hit it on the street in a car like a 911.
i haven't driven any modern mid-engine cars like the R8 or boxster though, i'm sure it's vastly improved from the NSX, which was basically designed in the 80s, so no fault to it. it was a great car. can't wait to see the new one.
right now i drive a performance sedan but my next car will be a "regular" 911, probably a 991 S.
as for the tesla... eh. doesn't give me that "gotta have it" feeling. i still want gasoline and a clutch. call me old fashioned.
i would be interested to know the % of current tesla owners that have ever owned "real" sports cars, not just "sporty" cars. it doesn't strike me as the same crowd.
>as for the tesla... eh. doesn't give me that "gotta have it" feeling.
It's actually really easy to test drive Model S. Click big red button on the website, they call you, and boom - test driving Model S Performance.
While I don't think it's time to buy EV yet, I highly recommend test driving Model S Perf - it feels, accelerates and handles beyond amazing for car of it's size.
the model S performance competes against M5, E63 AMG PP, Audi RS5/6/7 and can't beat any of those in 1/4 mile or 0-60 times. i would disagree with "beyond amazing", it's at the rear of the pack in terms of acceleration and handling.
if you've never actually driven a fast car, i'm sure it's very impressive, but objectively, it's not fast compared to its competition.
it appeals to a different crowd, the kind that wants EVs. i bet most people who considered/bought the S perf have never seriously considered an M5, E63, RS, or even Cadillac V (since we're talking American cars here). all of those handily beat the S in every benchmark.
I've driven 20+ BMW M3s and Porsche 911s. The Model S spanks them. It's a surreal driving experience. Totally quiet. Instant torque. Amazing handling for a large car.
It's really something you need to experience before you criticize it.
so in other words no M5, no alpina B7, no E63. probably no Caddilac V, no LS7 ZR1, no Audi RS either. in other words, you have no idea about other cars in the Sperf's class, just some bullshit cheapie base-model cars that are about $30k cheaper. but you are a tech dude with money who likes the idea of electric vehicles, so you like the Sperf, because it APPEALS TO YOU. and there's nothing wrong with that.
"instant" torque is cool and all, but no match for "more" torque. which also happens to be "instant" when you move beyond a cheapass 3-series M or base 911 with no balls (try a GT2, when you start wearing big boy pants) or V10 in that price range.
I've driven a 911S which I'd say is a pretty respectable car. I've driven also supercharged M3s as well. But you're right, I haven't driven those other cars and I'm not making a judgement call on them.
All I'm saying is that in my opinion based on driving the cars is that I prefer the Model S Performance over the supercharged M3, 911S because of superior handling, torque and feel.
Just my two cents, no need to get offended that a battery car can give a decent ride.
Also, I know there are people who have traded in the M5's for a Model S Performance and have been quite happy.
haven't test driven one but it can't beat my current car in performance numbers, so ...
and the roadster just absolutely does not appeal to me. i don't like roadsters in general but it's just too feminine and it looks hideous with the top on.
Generally speaking I'm right there with you, but thrown by your use of Japanese cars as an example of underdamping.
I haven't driven a lot of European cars, but compared to most offerings from Detroit Japanese damping is fantastic.
I've driven late model cars by two of the big three that felt like they forgot to put dampers in at all,
you just bounce and float around on the springs. I assume this (combined with size) is why American cars are so often referred to as "boats".
The car definitely does handle wonderfully for a non-professional driver. There are quite a few complaints about severe tire wear due to the negative camber for people who consistently ride it very rough.
I suspect they engineered the throttle lift oversteer and such specifically for the wider market they were looking to hit. Not just car aficionados, but also people who were EV or green or geeks and such.
For comparison the McLaren MP4-12C [1] has probably the most trick setup you'd see this side of a early 90s Williams F1 car [2][3]. Notice on the 12C how the rebound/compression dampers from each corner are cross connected instead of a roll bar, and the z spring at the back to resist high downforce.
Considering the Williams F1 active system was developed for the 1992 season, I'm kinda disappointed that no fully active system has ever hit a production car. The best we've seen is active damping on several cars. Bose had a pretty cool prototype [4], but never hit prod. The neat thing about electromagnetic damper/springs is that they can regen for a EV.
If your unfamiliar with car mechanics, then this is a good serious of short video tutorials [5].
that article is wrong, most of the cars listed there have semi-active systems with just the damping being electronically adjustable (see my below comment). That's pretty sloppy work by edmunds, considering all car press releases describe in detail(in marketing speak ofcourse) exactly what systems they have and how they roughly work
Pretty much all systems marketed as 'active' aren't really 'fully active'. Remember from mechanics/Differential Eqns, a suspension system is basically a damped harmonic oscillator. The two main constants are the springing and damping. The majority of cars just have electronically variable dampers. ie the same oem parts from bilstein/dephi/bosch with some silly marketing acronym on like Porsche PASM, MagneRide etc.
Some have variable strength roll bars, but roll bars are really just a 'hack' anyway. Very few have used a variable springing system (Merc ABC). An air suspension can change springing and ride height, but's it's slow.
In addition to low latency damping/springing adjustments/ride height, a fully active system in able to apply force into the suspension so the attitude of each corner is under complete software control, milli by milli. You can have a perfectly smooth ride, but no roll. All current non-active designs are a compromise between sporty and confort (high vs low springing/damping constants, and movements in attitude due to accelerating/braking/downforce/cornering/road surface.
I think it's a symptom of "Tesla = auto upvote on HN" syndrome. If this had been "Ford Focus S Suspension Walkaround" (or "Porche 911", for that matter) I doubt we'd be seeing it on the front page.
I work for a suspension manufacturer in the aftermarket industry for trucks, and a lot of it isn't that extraordinary. The air springs/shock absorber combination that you see there are becoming more normal across wider applications, while already coming equipped on luxury vehicles for quite some time now.
The part that impresses me though is the steering knuckle (they comment on it towards the beginning), and its aluminum hollow construction. We build modified steering knuckles for trucks that provide additional lift height, but out of ductile iron in a cast mold, same as OEM specs/mfg process. So i'm kind of at a loss how they would make it hollowed out, while maintaining stability. Definitely expensive, i know billet aluminum knuckles for corvettes run about $1500/piece.
Basically welding using a combination of joining pressure with a very high speed spinning steel bit that heats the aluminum joint via the friction created by it spinning on its surface. That spinning bit travels along the seam heating, and then it's pushed together when it's no longer quite solid, and spun together by the bit. Some interesting discussion of the crystal structure formed and the associated mechanical properties, as well.
When I look at this, I cannot how complicated it must be to build a world class automobile. Its all the more impressive that this is the first ground up(vs the roadster being an electric Lotus) car from a start up. Impressive indeed!
More or less what I was thinking while reading the article. How do you even staff up for something like this? I guess a lot of recruiting in Detroit or something.
"Huibert Mees is the Chassis Systems Technical Fellow responsible for the design and engineering of the Model S chassis systems. He joined Tesla from Ford Motor Company, where he was responsible for all the chassis systems and vehicle structure of the Ford GT supercar. Huibert has been working in the automotive industry for over 20 years, and was responsible for the rear suspension design of the Lincoln LS and Jaguar S-type as well as numerous advanced chassis design concepts at Ford."
Both SpaceX and Tesla are largely built on skimming the superstars out of the entrenched players. There are plenty of people that are really really good at their jobs at Boeing, Lockheed, Ford, General Dynamicsm, etc. They are generally underpaid, undervalued, and given little freedom relative to their skillset. But they stick around in those companies because that's where you have to be in order to work on really cool stuff like the Ford GT 40 or a space rocket.
Tesla and SpaceX offer the same cool stuff to work on, without the crappy low-value coworkers that never get fired, legacy systems holding you back, and the opportunity to do things as you believe they should be done.
This is part of why the old companies will have a hard time noncompeting. Not only are SpaceX and Tesla disrupting their business, but Elon Musk is stealing from their top tier talent pool that could defend against disruption.
If the market is any indicator, it believes Tesla is going to be successful. I bought several thousand shares at $17/share right after the IPO. TSLA closed today at 54.55/share.
How specifically did he recruit them? Did he just show up at the factory one day? (I'm really interested in the detailed logistical steps of how one starts a brand new car company)
If you want to take Tesla as an example for how to start a brand new car company:
1) Spend 6 years at an insanely high burn rate with zero revenue for research and development to create a marketable car
2) Have enough brand loyalty to delay delivery to your high-earning clientele multiple times
3) Hustle Daimler into investing to do payroll at the height of the financial crisis, with Daimler themselves having no money
4) Hustle the DOE into funding your fully electric sedan while having only a taped together basic model to show for it, with your competition (Toyota) being granted billions just to refit a factory
Interestingly, my friend who was interviewing with Tesla said that their HR was among the worst he ever talked to. Specifically, they didn't return emails and were quite inefficient.
Not as impressive (well, not to me) as how unnervingly clean the bottom of that car is. The roof of my car would be dirtier than that about 1 day after washing it.
I suppose this must be what owning a car in California is like.
Sounds awesome - I think some MIT guys are working on something similar with a liquid polymer that hardens with a laser. Are there any inherently very expensive parts of dmls, or is the expense due to low volume/high markups? What are the big barriers to a consumer-level version?
Reading about suspensions reminds me of the inerter (J-damper) [1] used in F1 racing. It was discovered by analogy with electrical circuits; it was the missing mechanical part corresponding to a capacitor.
So HN Tesla owners, how happy are you with the car itself, not considering the ideas behind electric cars and what not, just with the car and driving experience and charging, worrying about running out of juice ...?
It's fantastic. Best thing I've ever driven, regardless of environmental impact. It's silent and odorless, responds instantly and precisely, and it seats 5 adults + 2 kids comfortably.
A niche benefit: I live in the mountains and it handles going down much better than gas cars, where you have to downshift and use engine braking to avoid burning the brakes. It uses 12 miles of range going up, and gains 4 back going down so the average is close to a flat road of the same distance.
I'm curious about how it handles downhills. I would've thought that without engine braking you would have to rely exclusively on the brakes and would thus burn them up even more. Does the Tesla have another mechanism? Or just insanely efficient brakes?
Have you put kids in the back much? Based on their comfort and the space they take up, do you feel that the S is enough for a family car, or will the X better fill that niche?
My kids absolutely adore "the cave". They wave to people sometimes at red lights. ;)
They won't be able to sit back there forever. My daughter probably has 2 years and my son maybe 4 or 5. When they are both done with the seats, we can have them removed and get even more trunk space.
One thing that I ding Tesla a bit about... They don't tint the hatchback window the way they do the panoramic roof glass. This was a mistake as the back seats get hot in the direct sunlight and there isn't currently air conditioning vents back there.
Enough owners have raised these concerns that I believe they will likely correct one or both of them in future revisions of the car, and we'll see if they make it one of the proposed hardware retrofits.
In the interim, we had the back windows after-market tinted and the kids have no complaints.
It's almost certainly safer than the late 1960's station wagon we had when I was a child, with the backwards-facing seats in the cargo area. The rear crumple zone on it was everything behind the axle.
Yeah, it is the most expensive car I've ever owned, and it is also the first car that I have ever truly loved. I enjoy every minute I am in it, and I can drive it around town doing errands and such without wasting gas which you can't do in some other very expensive luxury sedan or sports car.
"Those batteries sure take up a lot of space. I wonder what their physical volume is -- compared to a 15 gallon gasoline tank."
Its more realistic to compare the volume and weight of the ICE, transmission, driveshaft, rear end, and fuel tank with the electric motors, drive train, and batteries for two cars of similar range and performance.
About 26,472 cubic inches. Fifteen gallons is 3,465 cubic inches. (Taking the wheels off, if the car were rectangular its dimensions would result in around 696,581 cubic inches of volume, so the battery pack takes up about 3.8% of that)
Rotors don't have to be changed as frequently as pads. It's probably a lot easier this way vs dismantling the whole thing.
Thanks for that info! So the batteries take up about 9 times as much space as a gas tank?
If you do not replace the rotor when you replace the pads, a machine shop has to grind the rotor, so its surface will be flat to work effectively with the new pads. If you don't, the pads won't make full contact with the rotor, and maintaining the brakes would make them operate worse until they wore in.
Brakes in electric cars receive vastly less wear than conventional vehicles, feels like literally an order of magnitude. They really only come into play at all during final stopping and emergency braking. The separate parking brake design on the Tesla is an interesting related detail.
In my experience so far (Prius, 110000 miles) it is more like 5x. And the Prius brakes are tiny things, especailly compared to the huge brakes on the German cars I've owned in the past. The Tesla S is comparable to them in weight and handling, and the brake system looks comparable too. From the testing and reviews we know it has excellent stopping distance, so it looks to me like the best of both worlds. I wouldn't be surprised if it gets 10x the lifetime out of brake pads, depending on driving conditions of course.
Ah, you're right. I'm cheap and lazy but this car deserves doing it right. Perhaps the pads are slipped in after the assembly is put together so you don't have to juggle with them while putting it on? Alternate idea: a chunk of the caliper missing reduces unspring weight?
I thought this was absurd so looked it up. Apparently most cars these days have metal pads, so both the pads and rotor wear equally. I'd never heard of that, all the small city cars I've had always had pads made of rubber/ceramic material.
A lot of manufacturers sell cars with ceramic because it's less annoying to the driver. Performance brake pads are going to be metal (or a combination of alloy and something else) because they bite harder and pull heat away from the rotor.
I've worked on many domestic and foreign cars, and I've never seen slide-out calipers. Most have a window so you can check pad and rotor thickness, but none have a slide-out section.
Even if you don't replace the rotor, you still have to remove have it turned (grinding the contact surfaces flat again), so the new pads make good contact, instead of being pressed against the ridges and valleys left in the rotor from the worn-out pad.
It isn't over-engineering because they are just Brembo calipers with Telsa stamped on them. Brembo has a huge share of the performance/racing brake market.
On a regular car you might have to have your rotors turned, but on something like this the rotors are more heavy duty and I think resist warping. Plus, on something like this you could run a lot softer pad (for track days) and need to change them out after a single weekend of use. It is still a good idea to sand the rotor surface, by hand, just so the pads will seat correctly.
Edit: And I do agree, I've never seen a "normal" car with the slide out pads... although they always look like you can do that until you get close and realize it is going to be a little more work than you thought. :)
This isn't quite my field, but I've been told you only need to have a rotor turned if it's warped or scored; otherwise you can just change out the pads, then apply the brakes heavily to conform them to the rotor surface.
Some pads have a thin piece of soft spring steel that drag on the rotor when the pad is thin, producing horrible screeching noises without damaging the surface.
Almost all performance brakes have this easy drop in pads to make changes faster at the track (I'd change pads 2-3x a track day in my evo). They bought an off the shelf brake setup from brembo. Nothing special here for people in the know.
They have a Model S sans bodywork (just the lower frame, drivetrain, and suspension) at the Tesla Store here at the 3rd Street Promenade in Santa Monica. If you get a chance to stop in a Tesla Store, it's definitely worth it to see the intricacies of the suspension.
Pretty pictures, but I just like how I can push the buttons on my screen and make the car go up and down. ;)
Actually, when I first read this doc a couple of weeks back, it showed me the reason for something I had noticed. If I have my foot on the brake while changing the suspension level, I've noticed that when I let off the brakes, the wheels shift a bit. Turns out that for the front tires, it doesn't lift and lower straight up and down. There is a bit of horizontal motion as well.
"21x8.5 inches +40mm" on the wheels. Gotta love the auto industry's ambivalent relationship with the metric system! Too bad they didn't try a little harder, they probably could have expressed one of the dimensions in micro-furlongs.
Why is the regenerative braking limited to .25g? Even if they can't charge the battery faster than that, couldn't they dissipate the extra power some other way?
They effectively do dissipate the extra energy already - as heat, from the physical brakes. If they can't charge the battery faster than .25g of breaking would provide, there's no reason to put increased stress on the engine (it's much more expensive than a physical braking system).
Some electric trains actually do this, they dump the energy as heat in a bank of resistors.
The motor on a tesla is on the rear axel, but most braking power is dissipated from the front axel due to the weight transfer. The question is if you also had a front axel motor (like the Model X), how much regen could you do. I remember reading somewhere that induction motors don't regen that well at low RPMs, hence you need to supplement with friction brakes. Anyone with a better understanding care to comment; would it be possible to build a car with no friction brakes, just 100% regen? I suppose you could always apply a counter torque to the motor.
So the regen force would still exist (?), just it would be really small, and so taking a long time to fully stop. Would it be possible to apply a reverse torque to the electric motor to decelerate faster. For example, I'm thinking about a hypothetical race car that could have no traditional hydraulic braking system to save weight, a large amount unsprung.
I believe Lotus does a lot of engineering consulting helping other manufacturers dial in suspensions. I wouldn't be surprised if Tesla had them consult.
Even some of the big car companies have trouble dialing in suspension on their own (I'm looking at you underdamped Japanese cars). European brands tend to be really good at this, and can take the same raw materials that everyone else is working with and create superior setups. Porsche somehow manages to start with the 911 where they stick the engine way out behind the rear axle (high polar moment of inertia, poor front/rear weight distribution) and make it handle beautifully. VW cars also tend to have great suspension tuning. Being able to keep the weight low to the ground in the Tesla contributes to a low center of gravity and I imagine must really help enormously.
Also the body under-tray is beautiful, like something you'd expect to see on a purpose built race car, probably helps a lot with the aerodynamics. I imagine more and more cars will get similar under-trays to improve aerodynamics and fuel efficiency.
It's interesting there is strong anti-dive geometry built into the car, probably to avoid problems with throttle lift oversteer caused by inexperienced drivers. I wonder if the current Model S handling leaves anything to be desired when driven to it's absolute limits and how the car would handle with some of the anti-dive geometry removed. Removing this geometry is a common mod on some cars.