High speed braking instability - ABS/ESP bug?

[SammichLover]

I'm not saying aerodynamics are not significant, I'm saying devices (spoilers) fitted to production cars rarely have a significant effect on downforce...

That has to do with production cars typically not traveling at 140mph+ like the OP is discussing, especially in the US. Because of the force = velocity^2 nature of aerodynamics, and how modern vehicles that are driven on highways usually not being anywhere near design limits for traction and stability at Interstate speeds, a spoiler that doesn't amount to a hill of beans effect at 70mph on the Interstate becomes a real factor for keeping vehicles on the road when clipping along at 120mph, much more above that.

 

[bradhs]

I've felt the issue that OP described a handful of times. Always at high speed, in a slight turn, slam the brakes and keep turn radius. The car feels like it is dancing through pulsation of brakes and possibly powering the front and rear motors in an attempt to keep you going on a path that if feels you should be on.

Recently, under heavy acceleration through a turn that had very little grip the cars electronics actually got the car to oversteer slightly to where I had to correct it by reducing my turn radius.

I chalk them both of up to the software needing some fine tuning... In all cases while it does feel weird I had confidence that the car was not going to do something drastic.

 

[22522]

Something does start to happen around 118 MPH..

 

[22522]

Something does start to happen around 118 MPH..

That is down steep hills as an additional parameter. At high altitude.

Up steep hills saw no issue.

 

[Tarzan_fi]

I think P3D had a significant weight shift when unloading throttle, or rear regen is very high. I had a relatively high speed corner during a test drive which started winding in. I lifted the throttle rather quickly and I started to brake as it tightened. The car unloaded the rear rather harshly and it felt unexpectedly light. I just reduced brakes a bit, turned back the steering wheel a bit and when it stabilised I progressively increased braking again.

It felt as if the front was too soft or rear had too much rebound damping. I was able to repeat it a few times. I am not sure if it is because I had the regeneration on standard (or at least I had not touched it, as I didn't know about the option), which led to immediate weight shift and perhaps higher than optimal rear braking rather than me controlling brakes, or what was the reason. Brake balance may have also been too rear biased in that vehicle, and there can be many other reasons, including me just being unfamiliar with an EV.

But the rear was definitely lighter/more loose than I expected. I would consider getting the dampers re-valved if it wouldn't get better with low recovery and me getting more familiar with the car.

 

[Dangerous Fish]

After listening to Randy Pobst talking about how he helped develop track mode originally and a more recent experience of some instability he's felt in high speed corners when lifting off the throttle, I'm now of the opinion the Model 3 may not have been extensively tested at high speeds during development which surprises me somewhat but I suppose surprises are what Tesla is all about...

More work needed by either Tesla or some independent test drivers to work out what's happening and get some software updates in the pipeline.

 

[SammichLover]

After listening to Randy Pobst talking about how he helped develop track mode originally and a more recent experience of some instability he's felt in high speed corners when lifting off the throttle, I'm now of the opinion the Model 3 may not have been extensively tested at high speeds during development which surprises me somewhat but I suppose surprises are what Tesla is all about...

More work needed by either Tesla or some independent test drivers to work out what's happening and get some software updates in the pipeline.

What are the Autobahn equivalent proving grounds in North America? I wouldn't think typical road tracks here could do it, because of the scale required?

 

[Dangerous Fish]

What are the Autobahn equivalent proving grounds in North America? I wouldn't think typical road tracks here could do it, because of the scale required?

If Randy Pobst has reported feeling the same sort of instability whilst driving Buttonwillow (I'm making an assumption, but it sounds like it could be connected) then I don't think it would take a 125mph+ curve to reproduce the issue.

Somewhere like this has the facilities to permit some high speed braking/lane changing tests

Track & Facilities – Fowlerville Proving Ground

And there's a photo of a Model S on the site, so maybe Tesla did test the Model 3 there as well?

TRC would also be a good place to test Home - TRC

 

[SammichLover]

What kind of radius does the Model 3 need to maintain, say, 130mph?

 

[stickman]

I’m thinking this is tire related. I had the abs kick in at 130mph on the track (slight bend, I shouldn’t have braked) and that’s with P4Ss. If you’re doing that kind of speed on street tires and brake into a turn all bets are off, IMO

 

[Dangerous Fish]

What kind of radius does the Model 3 need to maintain, say, 130mph?

The way the OP reported the issue, he was braking from 125mph in a long radius bend. You would only need a track or airfield where you could reach 125 and enough track width and run off area to turn in and brake in order to simulate the same situation and that would be possible on many tracks and airfields I've driven on in Europe and certainly would be in the U.S. too. Some of the tyre manufacturers have very big dynamics/handling areas they use. We're talking massive expanses of tarmac where you could cone off a test scenario like this. It's not beyond the scope of automotive testers and it's something Tesla needs to do if they haven't already,

I’m thinking this is tire related. I had the abs kick in at 130mph on the track (slight bend, I shouldn’t have braked) and that’s with P4Ss. If you’re doing that kind of speed on street tires and brake into a turn all bets are off, IMO

Tyres and pressures could play a part in it, yes but if the tyres are correctly pressured and in good condition (and it's dry) you should be able to brake gently at 125mph with some steering input and not be anywhere close to the ABS kicking in to prevent wheel lock-up. The ABS system is not working correctly if you felt that happening. It might be the software over-braking one or more wheels because it's got confused and is trying to maintain stability, but that's not the same as the ABS kicking in.

I suspect what's happening is an interaction between braking, regen and steering inputs confusing the Vehicle Dynamics Controller and it's then applying the wrong amount of braking and motor torque to the wheels for that situation. That may feel like ABS kicking in to the driver of course.

Randy talked about the chassis dynamics engineers (Lars Moravy +1) adjusting the regen balance between the front and the rear on the AWD cars to dial out inherent understeer and oversteer in track mode as Randy drove on the track, but it seems he wasn't asked to help with the non track mode regen settings or the RWD Model 3. He also says that he only had 1 day of testing with Tesla to develop track mode! A day is just not enough. He now admits there's further work to do after feeling more instability when driving 3s at Buttonwillow again recently.

If Tesla has only done one day of Model 3 chassis testing at high speed it's quite worrying really, especially for our German friends with their unrestricted stretches of road where they are expecting to use car to it's full potential.

This does need escalating. Or build some RHD cars and send them over here so we can test them!

What would also be good to find out is whether this issue happens with track mode off in either regen setting and if it happens to all varations (RWD, AWD, P3D).

 

[diamond.g]

Do we know if turning on Track mode makes it worse or better?

 

[Dangerous Fish]

Do we know if turning on Track mode makes it worse or better?

I don't believe we do and that would be useful to find out, but I'm not suggesting anyone should test any of this on public roads!

 

[stickman]

Tyres and pressures could play a part in it, yes but if the tyres are correctly pressured and in good condition (and it's dry) you should be able to brake gently at 125mph with some steering input and not be anywhere close to the ABS kicking in to prevent wheel lock-up. The ABS system is not working correctly if you felt that happening. It might be the software over-braking one or more wheels because it's got confused and is trying to maintain stability, but that's not the same as the ABS kicking in.

I suspect what's happening is an interaction between braking, regen and steering inputs confusing the Vehicle Dynamics Controller and it's then applying the wrong amount of braking and motor torque to the wheels for that situation. That may feel like ABS kicking in to the driver of course.

That’s a whole lot of conjecture and speculation. You have no way of knowing if the ABS is working 'correctly'. In my case I felt the tires slide the first time through that corner (no brakes, ~130mph) and also braking into the first serious corner, so I've no reason to doubt there was tire slip when I applied the brakes there on the second lap. After that session I dropped pressures and waited until the tires were hotter (track hot) before going for it and had no further ABS activation during the rest of my sessions. I have no idea what kind of tire temperature is attained from extended high speed cruising but I doubt they get track hot. The OP said he was traveling ~120mph and applied 50% brake pressure. It does not seem unreasonable that the car would detect some wheelspin in that situation and engage ABS/VSC.

I agree that maybe Tesla should do more high speed testing for selling the car in Germany but everyone saying "this is a bug" without seeing any data is silly.

 

[Dangerous Fish]

The OP said he was traveling ~120mph and applied 50% brake pressure. It does not seem unreasonable that the car would detect some wheelspin in that situation and engage ABS/VSC.

Wheelspin caused by what? How are the wheels spinning? You'd have to have seriously bad set of tyres on your car to provoke the ABS with 50% brake pressure on a dry road, even at 120mph. 50% is not a lot of braking you know. You have to stand on brakes to get the ABS working, regardless of the speed.

I agree we need more data and more reports, but common sense can rule out some possible causes at this stage.

 

[Daniel in SD]

Wheelspin caused by what? How are the wheels spinning? You'd have to have seriously bad set of tyres on your car to provoke the ABS with 50% brake pressure on a dry road, even at 120mph. 50% is not a lot of braking you know. You have to stand on brakes to get the ABS working, regardless of the speed.

I agree we need more data and more reports, but common sense can rule out some possible causes at this stage.

What the heck does 50% brake pressure even mean? I assume you mean a ~0.5g stop (half of maximum deceleration rate) because 50% of the maximum brake pressure should be more than enough to engage ABS. There's a lot of margin in a brake system!

 

[stickman]

Wheelspin caused by what? How are the wheels spinning? You'd have to have seriously bad set of tyres on your car to provoke the ABS with 50% brake pressure on a dry road, even at 120mph. 50% is not a lot of braking you know. You have to stand on brakes to get the ABS working, regardless of the speed.

I agree we need more data and more reports, but common sense can rule out some possible causes at this stage.

I used the wrong term, I meant tire slip. But you're wrong. You don't need bad tires, and you certainly don't need to "stand on" the brakes to activate ABS, you just need to exceed available traction enough that the speed sensors detect slip. I had Pilot Sport 4S, with less than 1000 street miles on them, and I used a lot less than 50% brake pressure in that turn. They worked perfectly after they were up to temperature and at lower pressure.
Have you braked in a corner at 120mph? It's kind of an edge case for most drivers not on the autobahn or a racetrack with high speed corners. Common sense doesn't have anything to do with it, clearly, because I just gave you another anecdote that shows the ABS can and did come on in a similar situation, and should have (in mine). But I'm just some guy on the internet, just like you and the OP.

OP- you might try braking a bit before entering bends at that speed, to bring the tire temp up a bit and see if that helps.

 

[stickman]

I've felt the issue that OP described a handful of times. Always at high speed, in a slight turn, slam the brakes and keep turn radius. The car feels like it is dancing through pulsation of brakes and possibly powering the front and rear motors in an attempt to keep you going on a path that if feels you should be on.
.

heh. Why would you "slam" the brakes on in a high speed turn? If you did that and the ABS/VSC engaged that's totally un-surprising.

 

[Dangerous Fish]

What the heck does 50% brake pressure even mean? I assume you mean a ~0.5g stop (half of maximum deceleration rate) because 50% of the maximum brake pressure should be more than enough to engage ABS. There's a lot of margin in a brake system!

I agree. Difficult for even a professional driver to say what 50% brake pressure means. Pedal half down?
But 50% braking, if that's what the OP thinks he was doing, is never going to be enough to get the ABS going in the conditions he described. That's why I think it's a problem with the software.

Or maybe there was oil on the road? Or maybe one of his tyres was deflated? We can only go on what he says, but what I'm saying is that in good conditions with everything working as it should, you can brake through a corner and there will be enough grip from a tyre such as the Pilot Sport to prevent lock up.

I used the wrong term, I meant tire slip. But you're wrong. You don't need bad tires, and you certainly don't need to "stand on" the brakes to activate ABS, you just need to exceed available traction enough that the speed sensors detect slip. I had Pilot Sport 4S, with less than 1000 street miles on them, and I used a lot less than 50% brake pressure in that turn. They worked perfectly after they were up to temperature and at lower pressure.
Have you braked in a corner at 120mph? It's kind of an edge case for most drivers not on the autobahn or a racetrack with high speed corners. Common sense doesn't have anything to do with it, clearly, because I just gave you another anecdote that shows the ABS can and did come on in a similar situation, and should have (in mine). But I'm just some guy on the internet, just like you and the OP.

OP- you might try braking a bit before entering bends at that speed, to bring the tire temp up a bit and see if that helps.

I'm afraid you are still using the wrong terms. ABS works when the sensors detect lock-up, not slip. An ABS system releases brake pressure when lock-up is detected. You are confusing grip with traction and ABS with traction control. And yes, I have braked many times at 120mph before, into and during corners in road going cars with road tyres, in cars with and without ABS. Can't say I can remember experiencing what the OP described.

The Model 3 is a road cars with relatively soft suspension, so the weight transfer into a bend isn't going be dramatic. There's still a lot of grip available. If you're on a track and entering a bend too fast with too much steering lock on and you panic brake, then maybe you might just about activate the ABS. But not in the conditions the OP described. I've raced and tracked a lot of road-going and race cars over the years so I'm not making this stuff up.

Anyway, I'm going to leave you lot to it and hope someone somewhere does some objective testing to see if they can reproduce the problem, as all we're doing in this thread is arguing about something we have very little data on. I've offered some opinions on what I think is happening but of course everyone is free to dismiss them if they so wish.

 

[stickman]

Yeah the tires slip when the wheels lock up, or when you exceed lateral grip. Traction is traction you have a limited amount that can be used by acceleration deceleration and/or cornering. I know how ABS works. ABS can engage when the tire is slipping but you don’t even feel the wheels lock. But yes let’s agree to disagree and hope Tesla does some testing.