Another gearbox discussion

[power]

I think its great they are getting away with only one gear for simplicity sake and reduced costs and weight. If you look at their power curve for the Roadster, it starts to fall off at 6000rpm. It doesnt end up with lot of power at 14000rpm, so all your accelleration potential is down at 60mph and falls from there. A CVT/IVT is not so much for a gas motor, as it is for any kind of power system WITHOUT constant power. since there is a power peak at 6000rpm, a CVT would allow the full 280HP to be available for acceleration all the way to the max speed potential. (Which is probably in the 170mph range). Part of the confusion I speak of, is that regarding the "Flat" torque curve. This, by definition, means that HP is rising proportional with RPM, it means that even in the lower rpm, accelerative forces are left on the table, not fully utilized. Just think how fast the Tesla Roaster would be if it was constant power, not constant torque! (via use of a CVT, or some sequential, close ratio, auto transmission). I can show you several high performance combusion engine torque curves that look very similar to the electric motor, right up to the point where it redlines. Then, the eletric motor performs its magic.

If Tesla decided to race one of these little gems, It would do very well in the club ranks as is. Picture how light it could be with all the interior gutted. A modification of a second gear would help it with accelerations in areas where it would most be mostly used . (i.e. after 60mph, more like 70 to 120mph). Even if the car only had a 120mph top speed, it would be fine for tracks here in northern california. I bet it would surprise a bunch of racer mainstream.

MK

I think Tesla should just stick with "direct drive" single ratio gear boxes for all models. The eMotor is so flexible that you can get away with that and it helps efficiency, and reduces cost and complexity. Also it eliminates one more thing to think about while driving.

If they need a bit more top speed they could work on better cooling and perhaps even a bit more high RPM capability without having to resort to the complexity of a shifting mechanism.

CVT/IVT is great/useful when you have a gas motor with peaky torque characteristics, but the eMotor is so flat that it doesn't seem needed.

 

[just-an-allusion]

Part of the confusion of the gear box discussion is a basic misunderstanding of what really accelerates a vehicle. Tesla has designed a single speed gear reduction that uses a near 8:1 ratio. Coincidentally, this is a very common final ratio for most high performance piston engine cars in 1st gear and usually reach a speed of 50-60mph. Because of its near flat motor torque curve, (which is no different than many piston engine's torque curves at near the same power levels), both a piston engine in 1st gear vs a Tesla Roadster with similar peak HP power plants will accelerate at near the same rate up. (If we are talking about comparable weights). The Tesla Roadster then still is able to accelerate up to 120mph still in the same single gear.
What's more amazing about the Tesla Roasters power plant, is not so much the flat torque curve to 4500rpm, but the slow decay of power to 14,000rpm. As we know, this limits the top speed and acceleration of the vehicle to 120mph. If the Tesla Roadster was able to have a second gear that was selected somewhere well past its peak HP was reached, its top speed would be extended as well as increased acceleration to that top speed.

Acceleration is proportional to power at any same vehicle speed as seen by this simple Newtonian identify:
Acceleration=Power/(mass x velocity)

Yes, everyone understands/remembers simple physics (doug), my primary problem is everyone's efforts in trying to correlate a motor's performance and functionality to that of an engine's, or vice-versa, when it's such a clear case of apples to oranges/they are two distinct, separate entities unto themselves with their own individual methodologies of operation and performance characteristics, i.e., while a "gear shift" may be a perfectly appreciable method of operating an engine at various speeds, the same doesn't hold true for a motor and any efforts on any car makers behalf to equate/integrate the actuating mechanisms of the one into that of the other earns them a lesson in futility, often had at quite considerable expense.

So, the idea is to maximize power at any vehicle speed. In the end, the way it will probably be done, will be to use a IVT (Infinitely variable transmisson). This way, max power would always be used when max acceleration was required, and some other rpm could be used that was in the motor's peak efficiency range for a given cruise speed.

Here is where I begin to see signs of a deeper intellect, i.e., if any form of a "transmission" is a requisite for an electric powered vehicle at all...it's a new form of powering our vehicles and as such should have the metaphorical table cleared/the chaulk board wiped clean of any vestiges of information relative to the "old school" method of powering our vehicles and an effort to begin anew should be the guiding philosophy of the day...but I like where your heads at.

 

[TEG]

I have driven a few EVs which had retained the old transmission from their post conversion days. The wide/early torque of the electric motor generally means you can just leave it in one gear and not bother with shifting. Sure, you could probably get a little bit better performance if you put in a CVT that could handle the torque, but does the Roadster really need it? I think it has sufficient performance to out accelerate just about any road car at street legal speeds.

Also the idea of trying to gear the Roadster differently to make it into a 170MPH vehicle seems misguided to me. I think the current eMotor would overheat, and the battery pack probably wouldn't be too happy trying to maintain the power required to achieve that sort of speed. At the moment I think heat build up of the eMotor is the most limiting factor, but the health of the Li-Ion cells probably aren't too far behind. They don't like to be discharged too quickly.

 

[power]

I think anything that Tesla can do to compete against the supercars is not misguided. Certainly, they need to work out a few other things first before they start looking at revolutionizing racing

I dont think overheating would be an issue if the car was to spend a little more time at its max power range. Its not uncommon for these type of motors to have some type of full power rating. There is also no difference in (to the motor and drive electronics) in accelerating from 40 to 80 mph , (operating in the peak HP range) than to operate from the 80 to 160mph. ( I just corrected an assumption by looking at the Tesla power/torque curves. peak hp happens at near 8000rpm not 4500rpm as I first stated.)

A simple 2 speed gear box wouldnt be that much of a challenge. Im sure some of the sequential, no accelerator lift, racing transmission technology could be used here!

Certainly there could be considerable performance gain potenatial for the roadster past 80mph where the power is falling off past its peak power pretty rapidly. In a racing environment, this is where the performance needs to be. The roadster would be anemic by losing near 100 hp compared to its piston engine competitors in this speed range. A simple gear box could drop the rpm back down to just below its peak HP range

MK

I have driven a few EVs which had retained the old transmission from their post conversion days. The wide/early torque of the electric motor generally means you can just leave it in one gear and not bother with shifting. Sure, you could probably get a little bit better performance if you put in a CVT that could handle the torque, but does the Roadster really need it? I think it has sufficient performance to out accelerate just about any road car at street legal speeds.

Also the idea of trying to gear the Roadster differently to make it into a 170MPH vehicle seems misguided to me. I think the current eMotor would overheat, and the battery pack probably wouldn't be too happy trying to maintain the power required to achieve that sort of speed. At the moment I think heat build up of the eMotor is the most limiting factor, but the health of the Li-Ion cells probably aren't too far behind. They don't like to be discharged too quickly.

 

[TEG]

There is also no difference in (to the motor and drive electronics) in accelerating from 40 to 80 mph , (operating in the peak HP range) than to operate from the 80 to 160mph.

Actually there is... At 80+ MPH you are using a lot more power to fight wind resistance, so the available acceleration power will be less. Therefore you would spend a LOT more time doing the 80 to 160mph acceleration thus causing a much longer period of full power being applied. I think the vehicle is designed for shorter bursts of acceleration, not prolonged full power as would be required by some racing and autobahn use.

I think Tesla basically proved their point with the roadster, and they would like to move on into more mass production vehicles, not into specialty racing. Certainly some Roadster customers have inquired about racing opportunities, but I don't think that is where the company has their focus right now.

 

[vfx]

A simple 2 speed gear box wouldnt be that much of a challenge

:biggrin:!!

 

[graham]

A simple 2 speed gear box wouldnt be that much of a challenge

:biggrin:!!

Hee! I have seen that movie before... it didn't end well...

 

[just-an-allusion]

I think Tesla basically proved their point with the roadster, and they would like to move on into more mass production vehicles, not into specialty racing. Certainly some Roadster customers have inquired about racing opportunities, but I don't think that is where the company has their focus right now.

"Racing" is never where Tesla had their focus.

The impetus behind the ideology of Tesla was/is to produce a real world, market viable alternative to the ICE, the next step, that would propel the World in to the future of vehicle powering...and perhaps even beyond the inherent limitations of the commuter arena.

Basically, to prove that it could be done, which they have, unlike GM who chose instead to (attempt) to prove that it couldn't (If only they had the benefit of hindsight, though they should have known better).

 

[Brent]

A simple 2 speed gear box wouldnt be that much of a challenge. Im sure some of the sequential, no accelerator lift, racing transmission technology could be used here!

Just to clarify what some other posters have alluded to...it turns out that it IS a challenge to make a two speed gearbox for an electric motor. TM nearly sunk itself financially over botched designs, and is still not out of the woods. A google search should turn up loads of info on the problem.

I believe the biggest challenge stemmed from the high RPMs at which the e-motor operates. Most regular (non-racing) autos' gearboxes don't deal with RPMs in the 13,000 range, so the Roadster's required new designs, none of which seemed to work out. Either the gearbox wasn't reliable long-term or it didn't shift quickly enough or it wasn't cheap enough. When TM announced that it was skirting the problem by using upgraded electronics, you could almost sense the sigh of relief.

 

[just-an-allusion]

Perhaps it is that the whole issue at the core of all of the engineering turmoil and financial expenditure that Tesla (as well as many other prospective EV manufacturers) has undergone is due to everyone's inability to think of the drive train as a motor (that it is) instead of as an engine (which it is not).

 

[TEG]

I believe the biggest challenge stemmed from the high RPMs at which the e-motor operates. Most regular (non-racing) autos' gearboxes don't deal with RPMs in the 13,000 range, so the Roadster's required new designs, none of which seemed to work out.

Along with the dizzying high RPMs, their 2 speed gear box had to match a dramatic shift in RPMs going from 1 to 2. For a more traditional 5 or 6 speed transmission each gear change is accompanied by a somewhat small change of ratio and slight change in engine RPMs. For Tesla they were trying to make it shift very quickly while switching eMotor RPMs from like 12Krpms down to something like 6Krpms in one big jump. Also when the eMotor slams on the torque again it can really torture the gears given that there is no flywheel buffer to smooth things out. They mentioned a multitude of problems they tried to overcome, but my point is it wasn't just a typical exercise.

If they weren't so focused on supercar acceleration they might have been able to make things reliable by making a more leisurely shift mechanism.

 

[vfx]

I'll throw in that they had a quantity issue as well.
The numbers they needed to make were well under the minimums the majors were even willing to look at especially for a unknown company.

Whomever ended up making the gearbox it was going to be far more expensive than any other because of 1. difficult large shifting ratios and, 2. no high volume price breaks.

 

[power]

I guess it depends on what your definition of "a lot" is. . In our racer, 20mph segments from 60mph to 120 are anywere from 2 seconds to 4 seconds when you measure the time it takes to get from 100 to 120mph for example. In a racing application, I dont think 7 or 10 seconds full throttle is the issue for acceleration times at most tracks. It is a question of is there enough dwell time to cool the motor and electronics. If the Roadster has done 12 second drag runs at full power, certainly there are very few tracks that would demand more power for a longer period of time. In otherwords, on the track, full power times are rarely over 10 seconds at a time. (more like 7) The key issue, can the batteries, motor and electronics handle a full power, up to 10 second 50% duty cycle for 30mins.

Tesla is are folks that are racers/performance car enthusiasts. Im sure they are thinking about a way to make a stand in the racing community eventualy to validate their product at the highest standards. We (they) might have to wait for the battery technology to evolve a little more.

Mk

Actually there is... At 80+ MPH you are using a lot more power to fight wind resistance, so the available acceleration power will be less. Therefore you would spend a LOT more time doing the 80 to 160mph acceleration thus causing a much longer period of full power being applied. I think the vehicle is designed for shorter bursts of acceleration, not prolonged full power as would be required by some racing and autobahn use.

I think Tesla basically proved their point with the roadster, and they would like to move on into more mass production vehicles, not into specialty racing. Certainly some Roadster customers have inquired about racing opportunities, but I don't think that is where the company has their focus right now.

 

[power]

I dont really think that is the issue at all. The drive train doesnt know or car what is driving it. The torque/hp curves are not all that different. Even the engine vs motor speeds are not all that different. However, someone did bring up a good point. sure, motorcycle, porsche GT3cup cars have sequential gear boxes that shift from 9 to 13,000rpm, BUT, the dont drop the RPM in a fashion that the Roadster would have to. However, there is nothing stopping Tesla or any other EV company to have 2-3 gears up top in a sequential fashion. That would solve that issue. still no need for a clutch as we are talking only having gear shifts at the 80mph + range.

I dont think that there would be a need for more than a single shift. a shift at 9000rpm dropping the rpms down to a common 7000rpm would give the tesla a shift at 90mph and provide the extra performance up to near 120mph and slightly beyond. This would be sufficient for most all road racing applications of the car.

Mk

Perhaps it is that the whole issue at the core of all of the engineering turmoil and financial expenditure that Tesla (as well as many other prospective EV manufacturers) has undergone is due to everyone's inability to think of the drive train as a motor (that it is) instead of as an engine (which it is not).

 

[power]

Good point as I mentioned about the dramtic drop in rpm required for a dual speed gear box. However, there is no rule saying that there couldnt be 2 additional gears after 80mph to select

So, then the motor would be shifted from more like 9,000rpm ending up at near 7000rpm which is the rpm drop of most all race cars and motorcycles. remember, the idea here is keeping the engine in the max HP range which the motor provides on either side of 8000rpm.

Also, there is no issue with not having a flywheel mass, infact that works to an advantage in this example. the motor or gears doesnt know what the load is when the power is supplied and applied to the gears. Its driving the entire mass through the drivetrain, including any inertia of a flywheel or not. (in otherwords, think of the tires as the flywheels. Again, there is no more torque with this electric motor than there is with a car with an equivilant power rating. as I mentioned earlier, older BMW M3's for example, have a flat torque curve to about 5000rpm and a similar peak HP.

mk

Along with the dizzying high RPMs, their 2 speed gear box had to match a dramatic shift in RPMs going from 1 to 2. For a more traditional 5 or 6 speed transmission each gear change is accompanied by a somewhat small change of ratio and slight change in engine RPMs. For Tesla they were trying to make it shift very quickly while switching eMotor RPMs from like 12Krpms down to something like 6Krpms in one big jump. Also when the eMotor slams on the torque again it can really torture the gears given that there is no flywheel buffer to smooth things out. They mentioned a multitude of problems they tried to overcome, but my point is it wasn't just a typical exercise.

If they weren't so focused on supercar acceleration they might have been able to make things reliable by making a more leisurely shift mechanism.

 

[graham]

It is possible that you could be interested in just the existing 2 speed XTrac, if you can convince Tesla to sell it to you. It is only good for 10,000 miles, but I understand most track cars end up swapping parts more frequently than consumer cars. Maybe that is good enough assuming you can continue to get the replacement parts?

 

[doug]

It is possible that you could be interested in just the existing 2 speed XTrac...

Is that what was in the first 40 or so production cars (just locked into 2nd gear), or did they have some other interim transmission? I forget. If it's the former, they may have some slightly used spares floating around fairly soon.

 

[vfx]

What Graham said.

It's not like there was transmission failures were after only a few days of driving downtown.
The long term durability testing is what killed them. I thought I read that they would fail at 5000 miles and Graham said 10,000. Either way a street car should last 100,000 or so before changing out such a major part. In the racing world even 5000 miles is a few years of going around a 2 mile track 30 minutes at a time.

 

[TEG]

Take it for what it is worth, but Jeremy had told me that TM doesn't intend to sell any used parts (e.g.: old transmissions, DT1.0 eMotors, etc.) I gather it isn't worth the liability and possible customer support hassles. I don't know if that stuff will be recycled or just "quietly disappear" somewhere, but don't count on it being offered up for sale to the general public.

With regards to the flywheel (or lack thereof), I don't think people are getting my point (or maybe I am wrong about it), but I think one thing a flywheel does is smooth out torque delivery. Lets say the PEM has some high frequency noise and the torque delivery fluctuates very rapidly with rapid, subtle variations in torque. A flywheel slows down the rotational changes so the gears don't experience those "fluttering fluctuations". With no flywheel the eMotor may be able to "micro-jackhammer" at the gear teeth during shifts. This is just my own (perhaps half baked) pet theory.

JB alluded to the fact that there were multiple "modes of failure" for the DT1.0 transmission, so we are probably only "scratching the surface" here.

 

[Joseph]

"Is that what was in the first 40 or so production cars (just locked into 2nd gear), or did they have some other interim transmission? I forget."


The forty some interim cars use the two-speed transmission locked in second gear with all the first-gear guts removed. It operates like a one-speed transmission since you can't shift out of it. Because it uses the old PEM, it can't produce enough torque so that's why the interim cars only do 0-60 in six seconds.


Here is the complete story from ABG:

"Tesla VP Darryl Siry contacted us to clarify the issue with the original X-Trac gearbox. Apparently the problem was not one of actual durability of the transmission itself. Tesla evidently spec'ed out a two speed unit with no clutches. The design intent was to do clutch-less shifting and manage the torque output of the motor during the shifts. Unfortunately the rotational inertia of the motor made this plan unworkable as the torque output couldn't be changed fast enough. According to Siry there were never actually any mechanical failures of the X-Trac transmission, it was simply a matter of Tesla not being able to get their control strategy to work adequately with the hardware. Subsequently Tesla contracted Magna to develop a dual clutch two speed gearbox. Unfortunately, this is the design that had durability issues leading to Tesla's ultimate decision to revise the motor for more power and go with a single speed unit for production. The vehicles running with the interim single speed gearbox are actually using the original X-Trac box with the lower gear locked out (for prototypes) or removed entirely (early production cars). - Sam"

Tesla now on the receiving end of a lawsuit, Magna sues for breach of contract - AutoblogGreen