NHTSA Complaint wrt Regen and TC

[doug]

There's been a NHTSA complaint about a safety concern with how the traction control turns off the regenerative breaking on the Roadster. Does anyone else have this issue? I want to emphasize that this is not a recall. As far as I can tell it's only one owner's complaint for a relatively specific set of conditions. It is, however, important to be aware of this behavior, if it is indeed common.

Make : TESLA
Model : ROADSTER
Year : 2008
Manufacturer : TESLA MOTORS, INC
Crash : No
Fire : No
Number of Injuries: 0
Number of Deaths: 0
ODI ID Number : 10277260
Date of Failure: July 15, 2009
VIN : 5YJRE11B381...
Component: TRACTION CONTROL SYSTEM

Summary:
THE VEHICLE IS PRONE TO 'SUDDEN ACCELERATION' PROBLEMS WHERE THE DRIVER LOSES CONTROL OF THE VEHICLE WHEN THE TRACTION CONTROL SYSTEM ENGAGES AND FORCES THE REGENERATIVE ENGINE BREAKING SYSTEM TO DISENGAGE. THE TESLA ROADSTER USES VERY AGGRESSIVE ENGINE REGENERATIVE BREAKING SO, WHEN IT IS DISENGAGED, THE CAR LURCHES FORWARD UNCONTROLLABLY AND FASTER THAN THE DRIVER CAN REACT BY ENGAGING THE BREAKS. THIS MOST COMMONLY HAPPENS WHEN I DRIVE DOWN A HILL OR AROUND A CORNER AT 20 MPH USING ENGINE REGENERATIVE BREAKING TO SLOW THE VEHICLE AND HIT A BUMP IN THE ROAD THAT ENGAGES THE TRACTION CONTROL SYSTEM. THE CAR WILL SUDDENLY DOUBLE SPEED (40 MPH) FOR APPROXIMATELY FIVE SECONDS AS THE ENGINE REGENERATIVE BREAKING ON THIS VEHICLE DISENGAGES... FASTER THAN I CAN COMPENSATE BY BREAKING. IN EFFECT, THE CAR SUDDENLY AND UNEXPECTEDLY LURCHES FORWARD WHEN THE TCS KICKS IN. I PROBABLY SEE THIS BEHAVIOR AT LEAST TWICE A WEEK ON ONE PARTICULAR STRETCH OF NEARLY EMPTY ROAD, BUT HAVE HAD TWO NEAR COLLISIONS RECENTLY, ONE ON THAT STRETCH OF ROAD FOLLOWING A TRUCK AT A DISTANCE AND ONE ON A DIFFERENT STRETCH OF ROAD MAKING A CORNER WHERE THE CAR NEARLY ACCELERATED INTO AN ONCOMING VEHICLE. GIVEN HOW FEW OF THESE CARS ARE ON THE ROAD, HOW AGGRESSIVE THE ENGINE REGENERATIVE BREAKING IS, AND HOW LITTLE MOST OTHER OWNERS DRIVE THEIR VEHICLES RELATIVE TO MY DAILY COMMUTES, I HAVE TO BELIEVE THAT THIS PROBLEM WILL LEAD TO MULTIPLE ACCIDENTS AND POTENTIALLY FATAL ACCIDENTS OVER THE COURSE OF THE NEXT YEAR. THE ISSUE HAS BEEN COMMUNICATED TO TESLA MOTORS, BUT THEY'VE INDICATED THAT THIS IS STANDARD BEHAVIOR FOR ALL VEHICLES USING ENGINE REGENERATIVE BREAKING AND SHOULD BE EXPECTED. I DON'T COMPLETELY BELIEVE THAT -- OR NECESSARILY BELIEVE THAT OTHER CARS ARE AS AGGRESSIVE WITH THEIR REGEN BREAKING AS THE TESLA IS -- BUT THE BOTTOM LINE IS THAT THIS CREATIVES A VERY UNSAFE DRIVING CONDITION THAT CONCERNS ME AND SHOULD CONCERN ALL OTHER TESLA ROADSTER DRIVERS. *TR



Make : TESLA
Model : ROADSTER
Year : 2008
Manufacturer : TESLA MOTORS, INC
Crash : No
Fire : No
Number of Injuries: 0
Number of Deaths: 0
ODI ID Number : 10277260
Date of Failure: July 15, 2009
VIN : 5YJRE11B381...
Component: VEHICLE SPEED CONTROL

Summary:
THE VEHICLE IS PRONE TO 'SUDDEN ACCELERATION' PROBLEMS WHERE THE DRIVER LOSES CONTROL OF THE VEHICLE WHEN THE TRACTION CONTROL SYSTEM ENGAGES AND FORCES THE REGENERATIVE ENGINE BREAKING SYSTEM TO DISENGAGE. THE TESLA ROADSTER USES VERY AGGRESSIVE ENGINE REGENERATIVE BREAKING SO, WHEN IT IS DISENGAGED, THE CAR LURCHES FORWARD UNCONTROLLABLY AND FASTER THAN THE DRIVER CAN REACT BY ENGAGING THE BREAKS. THIS MOST COMMONLY HAPPENS WHEN I DRIVE DOWN A HILL OR AROUND A CORNER AT 20 MPH USING ENGINE REGENERATIVE BREAKING TO SLOW THE VEHICLE AND HIT A BUMP IN THE ROAD THAT ENGAGES THE TRACTION CONTROL SYSTEM. THE CAR WILL SUDDENLY DOUBLE SPEED (40 MPH) FOR APPROXIMATELY FIVE SECONDS AS THE ENGINE REGENERATIVE BREAKING ON THIS VEHICLE DISENGAGES... FASTER THAN I CAN COMPENSATE BY BREAKING. IN EFFECT, THE CAR SUDDENLY AND UNEXPECTEDLY LURCHES FORWARD WHEN THE TCS KICKS IN. I PROBABLY SEE THIS BEHAVIOR AT LEAST TWICE A WEEK ON ONE PARTICULAR STRETCH OF NEARLY EMPTY ROAD, BUT HAVE HAD TWO NEAR COLLISIONS RECENTLY, ONE ON THAT STRETCH OF ROAD FOLLOWING A TRUCK AT A DISTANCE AND ONE ON A DIFFERENT STRETCH OF ROAD MAKING A CORNER WHERE THE CAR NEARLY ACCELERATED INTO AN ONCOMING VEHICLE. GIVEN HOW FEW OF THESE CARS ARE ON THE ROAD, HOW AGGRESSIVE THE ENGINE REGENERATIVE BREAKING IS, AND HOW LITTLE MOST OTHER OWNERS DRIVE THEIR VEHICLES RELATIVE TO MY DAILY COMMUTES, I HAVE TO BELIEVE THAT THIS PROBLEM WILL LEAD TO MULTIPLE ACCIDENTS AND POTENTIALLY FATAL ACCIDENTS OVER THE COURSE OF THE NEXT YEAR. THE ISSUE HAS BEEN COMMUNICATED TO TESLA MOTORS, BUT THEY'VE INDICATED THAT THIS IS STANDARD BEHAVIOR FOR ALL VEHICLES USING ENGINE REGENERATIVE BREAKING AND SHOULD BE EXPECTED. I DON'T COMPLETELY BELIEVE THAT -- OR NECESSARILY BELIEVE THAT OTHER CARS ARE AS AGGRESSIVE WITH THEIR REGEN BREAKING AS THE TESLA IS -- BUT THE BOTTOM LINE IS THAT THIS CREATIVES A VERY UNSAFE DRIVING CONDITION THAT CONCERNS ME AND SHOULD CONCERN ALL OTHER TESLA ROADSTER DRIVERS. *TR

 

[DaveD]

..., AND HOW LITTLE MOST OTHER OWNERS DRIVE THEIR VEHICLES RELATIVE TO MY DAILY COMMUTES,...

Well, I've just passed the 10,000 mile mark on my odometer, and my Roadster is my daily driver. I go up and down plenty of twisty hills, under wet, as well as dry, surface conditions. I've not once had this described behavior happen to me. I'm not qualified to speculate on what might be wrong with his particular car, nor am I able to judge his driving ability.

Is his reference to "traction control system" meant to be referring to the ABS?

 

[doug]

It does actually make sense for the traction control to reduce regen. Perhaps it shouldn't completely kill it like this owner describes. I can see how it might be a bit disconcerting (as it is when you lose regen because of a full battery). I don't, however, think it should be a problem if the driver is aware of it.

 

[TEG]

THE VEHICLE IS PRONE TO 'SUDDEN ACCELERATION' PROBLEMS WHERE THE DRIVER LOSES CONTROL OF THE VEHICLE WHEN THE TRACTION CONTROL SYSTEM ENGAGES AND FORCES THE REGENERATIVE ENGINE BREAKING SYSTEM TO DISENGAGE. THE TESLA ROADSTER USES VERY AGGRESSIVE ENGINE REGENERATIVE BREAKING SO, WHEN IT IS DISENGAGED, THE CAR LURCHES FORWARD UNCONTROLLABLY AND FASTER THAN THE DRIVER CAN REACT BY ENGAGING THE BREAKS. THIS MOST COMMONLY HAPPENS WHEN I DRIVE DOWN A HILL OR AROUND A CORNER AT 20 MPH USING ENGINE REGENERATIVE BREAKING TO SLOW THE VEHICLE AND HIT A BUMP IN THE ROAD THAT ENGAGES THE TRACTION CONTROL SYSTEM. THE CAR WILL SUDDENLY DOUBLE SPEED (40 MPH) FOR APPROXIMATELY FIVE SECONDS AS THE ENGINE REGENERATIVE BREAKING ON THIS VEHICLE DISENGAGES... FASTER THAN I CAN COMPENSATE BY BREAKING. IN EFFECT, THE CAR SUDDENLY AND UNEXPECTEDLY LURCHES FORWARD WHEN THE TCS KICKS IN. I PROBABLY SEE THIS BEHAVIOR AT LEAST TWICE A WEEK ON ONE PARTICULAR STRETCH OF NEARLY EMPTY ROAD, BUT HAVE HAD TWO NEAR COLLISIONS RECENTLY, ONE ON THAT STRETCH OF ROAD FOLLOWING A TRUCK AT A DISTANCE AND ONE ON A DIFFERENT STRETCH OF ROAD MAKING A CORNER WHERE THE CAR NEARLY ACCELERATED INTO AN ONCOMING VEHICLE. GIVEN HOW FEW OF THESE CARS ARE ON THE ROAD, HOW AGGRESSIVE THE ENGINE REGENERATIVE BREAKING IS, AND HOW LITTLE MOST OTHER OWNERS DRIVE THEIR VEHICLES RELATIVE TO MY DAILY COMMUTES, I HAVE TO BELIEVE THAT THIS PROBLEM WILL LEAD TO MULTIPLE ACCIDENTS AND POTENTIALLY FATAL ACCIDENTS OVER THE COURSE OF THE NEXT YEAR. THE ISSUE HAS BEEN COMMUNICATED TO TESLA MOTORS, BUT THEY'VE INDICATED THAT THIS IS STANDARD BEHAVIOR FOR ALL VEHICLES USING ENGINE REGENERATIVE BREAKING AND SHOULD BE EXPECTED. I DON'T COMPLETELY BELIEVE THAT -- OR NECESSARILY BELIEVE THAT OTHER CARS ARE AS AGGRESSIVE WITH THEIR REGEN BREAKING AS THE TESLA IS -- BUT THE BOTTOM LINE IS THAT THIS CREATIVES A VERY UNSAFE DRIVING CONDITION THAT CONCERNS ME AND SHOULD CONCERN ALL OTHER TESLA ROADSTER DRIVERS.

Nitpick: It should say "brake" not "break".

Also, for the roadster I would say "motor regenerative braking", not "engine regenerative breaking".

The language also seems to suggested bursts of acceleration, but from my understanding the action of stopping the regen would not cause the car to "lurch" or have "sudden acceleration" rather it could be an effect of no longer slowing. Yes, if you are going down hill there could be some gravitational acceleration, but I envision that would be minor and not as dramatic as described.

By the way, if you head down a hill with a nearly full charge in range mode, then switch to standard mode does the regen suddenly disengage since the computers consider the pack "already full" at that point?

 

[bolosky]

It does actually make sense for the traction control to reduce regen. Perhaps it shouldn't completely kill it like this owner describes. I can see how it might be a bit disconcerting (as it is when you lose regen because of a full battery). I don't, however, think it should be a problem if the driver is aware of it.

The 20->40 mph behavior he describes isn't consistent with the regen just being turned off by traction control. The only thing accelerating the car would be gravity. To go from 20 mph to 40 mph is (of course) a 20 mph speedup, which is about 9 m/s. Gravity accelerates you at one g (of course), which is 9.8 m/s, so to go from 20mph to 40mph would require about a second of freefall or many times that long rolling down any sane slope.

I guess his problem is that he's driving over a cliff.

He's also got a 5s reaction time.

(OK, I really think that he's not measured this and he made up ridiculous numbers, and that there is a real issue where the car starts rolling downhill faster when TC cuts the regen. But, it's probably not a big issue, and I bet it's not going to generate lots of accidents like he says.)

 

[Tdave]

I've seen the TC light illuminate going over a bump in the road while regen braking is occuring. I see this happen more than once a week. There are enough poorly maintained roads around here for me to see it that often. Perhaps my adjustable suspension set to very firm may also exacerbate the frequency of this condition.

HOWEVER, I've never experienced any noticable reduction of regen causing any sort of change in expected speed. The condition is momentary, a fraction of a second. If regen is being disengaged for me, it's reengaging quickly enough for me not to notice.

 

[doug]

The missing piece of information here is the throttle position. Is his foot completely off the pedal or is it partially depressed? To what percentage? (I wonder if the car's software records the conditions each time the TC engages.)

His language does suggest an unfamiliarity with driving dynamics. Notice also that there is a separate but identical complaint about the "VEHICLE SPEED CONTROL". So perhaps he thinks the car is actively accelerating. Of course the relevant quantity here is "jerk" which is the change in acceleration. If you have a negative acceleration that is suddenly reduced in magnitude, that results in a positive jerk which would be perceived as a forward "lurch" in the non-inertial reference frame of someone in the car.

At any rate, it is reasonable for the owner to make this complaint if he thinks there is a legitimate safety concern.


.

 

[Mark06GT]

I've had something similar to this happen to me twice, but in neither case was it as severe as this person describes. Both incidents were while traveling at low speed and hitting a bump. In both cases the car was not charged fully so regen had been working up until that point. In neither case did I feel like the car was out of control, but I did feel like the car lurched forward a bit after hitting the bump. I dismissed both incidents because I never felt like the car was uncontrollable, but maybe he’s on to something.

 

[J in MN]

When the 2G Prius first became popular, there were similar complaints from some drivers about its sensitive traction control. They would try to irresponsibly enter a roadway in front of fast oncoming traffic, and then try to say that the car is "dangerous" because of traction control. (The TC would usually activate because of loose gravel or snow, or bumps causing the wheels to lift off.)

I believe these are irresponsible drivers unable to adapt.

 

[mpt]

I experience re-gen being disabled almost every day on a steep downhill turn as the car struggles to grip over wet leaves and broken pavement.

Three points:

1. The momentary release of braking force does not unsettle the car, nor does it increase in speed in any way apparent to me.

2. If the regen were NOT to release, an inbalance in the rear braking force caused by the lack of traction at one side would produce lateral forces de-stabalising the rear.

3. Without re-gen, in my previous car, I'd slow down on friction braking alone; the ABS DOES trigger at the same point releaseing braking power and resulting in the same momentary response.

What a smart idea to have regen braking abs!

Dismiss.

 

[mpt]

...Oh, and, my 05' Honda Insight does the same but it doesn't automatically restore regen so I do pick up speed... eventually... after a hundred yards of no regen. Which is a bigger problem... I miss out on that lovely downhill recharge :-(

 

[Tdave]

Sounds like what we have here is nothing more than a loss of traction and therefore inability to maintain the same braking force. ICE cars with ABS will behave the same way. It doesn't matter what causes the loss of traction. That's an argument between the tires and the road. The Roadster's choice is to either lock up the wheels and lose control, or do what it does and reduce braking to maintain control. Decades of ABS experience says the latter is the right choice.

Scotty: "Ye cannot change the laws of physics, Captain"

 

[toutizes]

Ahhh, this explains a similar puzzling behavior I've been seeing several times.

I'm arriving a bit fast towards my driveway and brake medium hard to slow down. Then I hit a bump, always the same bump, and for a very short moment it feels like the brakes don't work at all. Then they bite again and everything's fine.

The moment is very short, a heartbeat maybe, but it's always enough to send a jolt of adrenaline in my body. Think "Oh Sh*t!"

I first thought that my was foot being jerked off the break pedal on the bump, but I made sure to keep the pedal pressed and it was still doing it.

This explanation makes sense. To test it we could check that it does not happen in performance mode (the regen is less there, right?)

Feels like a bug to me. Maybe traction control should not disengage regen if you're braking?

 

[malcolm]

Sounds like a fault.

In order to engage regen - just like engine braking on a conventional car - you have to have your foot off the accelerator/gas.

A bump to one wheel is likely to cause reduced rotation of that wheel, which would indicate to the TC that the powered wheels are spinning too quickly in comparison.

In both cases, why/how would more power be applied to the drive wheels?

Momentary loss of brakes shouldn't increase speed as dramatically as described.

 

[raymond]

The question is how the car can distinguish between a bump and a patch of ice. Encountering a patch of ice while braking (explicitly or regen) you loose traction, and the only way to keep control of the car is to release the brakes (stop regen) until you regain traction.

The effect of a bump is identical to a patch of ice, so whatever action the car chooses it should work in both situations.

No matter what, regen only applies to the rear wheels. If the rear wheels lift off because of a bump, you will loose regen braking. The question is how long it takes for the car to see it has landed and re-engage regen.

How long do the wheels loose traction because of a bump? How long does it take until regen is re-enabled? The latter should preferably be significantly less than the former.

 

[toutizes]

But, but, if you hit a bump with a front wheel, and this causes TC to shut-off regen to the rear wheels, won't you feel a noticeable decrease of your braking power?

If this is right, I stil don't see how it would be different in a regular car. The regular car TC would start the ABS pumping the brakes and you'd temporarily loose braking power too.

Could it be that the regen on/off/on/off cycle is much slower than the ABS brake on/off/on/off?

 

[mpt]

The TC light comes on whenever an issue occurs, it flickers when you go over speed bumps at an anlge or when one wheel hits a sunken manhole cover. Same on regen; it's really quick, 100mS-500mS tops. As soon as traction is restored the light is already out.

For those who've not driven a roadster flat out, the TC has to be experienced. Unlike ICE vehicles where the lag between control and reponse is wide enough to fit a bus through, the TC on the roadster is invisible; on wet, greasy leaves you can floor the accelerator and the car rockets off with the TC light on solid but with the car accelerating wildly; there's no pulseing, juddering, surging, etc. so much power under such fine control; it's a marvel to behold really. Nice job by Drew Baglino and team - Tesla Motors - Engineering

This is why I (and all owners I expect) find the idea of a complaint in this area to be quite incredible.

 

[Roger Reid]

But, but, if you hit a bump with a front wheel, and this causes TC to shut-off regen to the rear wheels, won't you feel a noticeable decrease of your braking power?

If this is right, I stil don't see how it would be different in a regular car. The regular car TC would start the ABS pumping the brakes and you'd temporarily loose braking power too.

Could it be that the regen on/off/on/off cycle is much slower than the ABS brake on/off/on/off?

As has been explaned to me the ABS and TC are separate. TC can be turned off. ABS cannot be turned off. In the case of this complaint, if the driver relied only on the regenerative braking going down a hill, in my opinion, inattentive driving.

On the other hand it behooves all Tesla drivers to test and know the limitations of your brakes. Take the time to bed in the brakes. Push the peddle as hard as you can in order to get the ABS to start pulsing. The brakes are powerful but don't be afraid to push hard and don't let the regen braking lull you into a false sense of security. If you are letting the regen braking slow you down, have your foot at the ready in case you need the brakes. Drive safe and use common sense.

 

[Norbert]

It sounds like the TC is already very good, but one could imagine that a car could have a sensor to detect when a wheel doesn't have weight on it, and handle that case differently (fully re-engage as soon as the weight is back on), than if it looses traction even with weight on the wheel. Not sure if that would make sense technically, though.

 

[BlackbirdHighway]

I believe I have experienced this effect. I was slowing down using regen on a wet road and and suddenly felt a sharp surge forward exactly like the car was hit from behind. I quickly looked in the rear view mirror, and the car behind me was too far away to have caused it. I was on a level road, so I can imagine it would be much more pronounced while going downhill.

I've seen the TC light coming on many times when hitting bumps in the road, but only one time felt the push forward during regen in over 6000 miles.