GM Volt Forum banner

1 - 20 of 33 Posts

·
Registered
Joined
·
847 Posts
Discussion Starter #1 (Edited)
Analysis of unexpected "Low Traction" (ABS) with the "brakes out" feeling

Some of you have encountered the anti-lock brakes engaging (seeing the "Low Traction" message) in situations where you would not expect. It is scary, as if the brakes are not working, or that the car is accelerating (as it felt to me). A number of forum members have reported this on grooved pavement or going over potholes.

The concern is that the ABS should not be coming on in such situations, and that it could be defective. If the car is indeed accelerating when it should be slowing down, it could easily cause an accident.

Yesterday, it happened to me for the second time. I was going about 15MPH in a parking lot -- one I have traveled on dozens of times without incident, and the road was dry -- and it felt like the brakes went out and the car seemed to speed up; the "Low Traction" light appeared.

Fortunately, I have been recording data from the OBD2 port for most trips since October, and had the data ready to analyze. Going back to the location, I saw that there was a storm drain that was slightly depressed in the road. Clearly, a tire went over it (while I happened to have my foot on the brake, so the ABS knew I wanted to go slower), and the wheel speed changed quickly compared to the other 3 wheels, causing the ABS to kick in.

After looking at the data for a long time, I found the data for one of the wheels. Although all four wheels were slowing down (since I was braking), this one wheel was slowing down much more quickly (looking at a graph, it is quite obvious). The logs seem to show the ABS kicking in within 16ms after the car reported the tire speed.

Although both times it felt like the car sped up when the ABS kicked in (and therefore like I was losing control of the car), the data makes it look like that was my imagination. The data shows that when the ABS kicked in, the car immediately started decelerating more quickly (which is good). There was only 1/10 of a second when the deceleration was less than before the car hit the storm drain, but combined with the immediate high deceleration, the car was stopping more quickly than I had intended. Despite the feeling of acceleration, the car did not speed up at all (assuming, of course, the data and my interpretation of it is accurate).

Brake pedals on cars have "feedback", where the harder you push, the harder it becomes to push them (so you "know" that they are working). My guess is that when the ABS kicks in, the brake pedal loses its feedback (the data shows I pushed harder on the brake right after the ABS kicked in). So when the ABS kicks in, it is easier to push on the brake pedal, making it feel like the brakes are out (and I'm guessing that psychologically it also makes it seem as though the car is going faster).

So from what I can see from the numbers, it looks like the ABS functioned as designed. Perhaps it is a bit too sensitive and should not have engaged, or perhaps it could have been designed to cause less fear in the driver (if I'm right that the brake pedal loses its feedback). But, there is nothing in the data suggesting the car actually went faster (at no point did it accelerate until after I came to a full stop a few seconds later). And perhaps that panic feeling as the brakes seem to go out may help you drive better under the circumstances.

[EDIT: Added graph]
 

·
Registered
Joined
·
3,165 Posts
Interesting. Thanks for sharing.

Is there any change in regen happening when the ABS turns on?
 

·
Premium Member
Joined
·
14,156 Posts
This is really interesting. I've never experienced this in the Volt but I have in another car when going over metal plates. It's very unnerving. Great to know that the Volt is in fact slowing down.

You post seems to refer to a chart which isn't attached. Not sure if that is purposeful or not.
 

·
Banned
Joined
·
289 Posts
I've had this happen once when slowing down, going downhill, and the right front tire went over a manhole cover. My theory is that when you're slowing down normally (not a panic stop), the regen is contributing lets say 60% (guestimate) to the braking force, and the brake pads are doing the other 40%. When the ABS kicks in, the regen braking disengages and it gives the impression of an acceleration because 60% of the braking force just disappeared.

Just my theory.
 

·
Registered
Joined
·
847 Posts
Discussion Starter #6 (Edited)
You post seems to refer to a chart which isn't attached. Not sure if that is purposeful or not.
It was purposeful -- I was too lazy to add it. :) I've gone ahead and added it to the original post. The graph isn't pretty, but functional (it is also easier to read if you view the full image).
 

·
Registered
Joined
·
847 Posts
Discussion Starter #7
Is there any change in regen happening when the ABS turns on?
Unfortunately, I do not know -- I haven't yet figured out what most of the braking-related ID's are for. :(
 

·
Registered
Joined
·
59 Posts
I have had "low traction" come on regularly when I go down a hill near my home at about 20mph with L selected. This happens before I apply any brakes at all.
 

·
Premium Member
Joined
·
14,156 Posts
I've gone ahead and added it to the original post. The graph isn't pretty, but functional (it is also easier to read if you view the full image).
Thanks for taking the time to post it. It's great and easy to see what you're describing.
 

·
Registered
Joined
·
3,165 Posts
I have had "low traction" come on regularly when I go down a hill near my home at about 20mph with L selected. This happens before I apply any brakes at all.
Agreed.. i see it regularly.. but often when I'm in L doing regen.. so I wondered if it detects the need for ABS even when there is no foot on the pedal.
 

·
Moderator
Joined
·
8,680 Posts
Unfortunately, I do not know -- I haven't yet figured out what most of the braking-related ID's are for. :(
If the data that you're logging includes either MG B or Battery current flows, it'd be easy to look for a spike/dip in them that matches the timeline. My past impression was that it cut regen for half a second or do when 'Low Traction' sets - but that may be like the surging forward we all thought it was doing...
 

·
Registered
Joined
·
847 Posts
Discussion Starter #13
If the data that you're logging includes either MG B or Battery current flows, it'd be easy to look for a spike/dip in them that matches the timeline. My past impression was that it cut regen for half a second or do when 'Low Traction' sets - but that may be like the surging forward we all thought it was doing...
Good thinking. I do have what appears to be battery flow (slightly positive when charging, large positive numbers during regen, negative numbers when not charging at idle, large negative numbers when accelerating). It shows that regen started a little over a second before the wheel speed was funky, and it continued with regen for about 4.5 seconds afterwards.

The power flow from regen does goes down, but the regen is still working (600ms after the ABS starts, the power flow is at 30% of when the ABS started).

So it looks like the regen is active even when the ABS takes over. It's possible that it may change the force of the regen to the problem wheel (or the other ones), but the regen overall does stay active.
 

·
Registered
Joined
·
284 Posts
The feeling of acceleration is actually a pulsed lack of rapid deceleration. If you've ever applied athe brakes to a vehicle with abs on a sheet of ice that is precisely the feeling you get. None the less the abs system allows you to stop more rapidly than you would have otherwise and more importantly, allows you to steer during a skid to avoid an impending collision. So if you are rapidly running out of stopping distance you can steer around the obstacle. As for regen, it would be automatically disabled, which would once again give you that false feeling of acceleration I suspect the regen is then either pulsed or hydraulic abs becomes more prominent because regen is akin to a rapid downshift ( 2 gears rather than 1). I've had the misfortune of doing a 2 gear downshift on my motorcycle during the early misty part of a rain storm, the part where they tell you to stay off the middle of the lane because that part of the lane has usually collected all the oil, brake fluid and antifreeze leaked from poorly maintained cars, not to mention fuel leaks. Well I downshifted from 6th to 4th and my rear tire starters to come out from behind me. I was traveling at 57 mph at the time and my motorcycle weighs 804 lbs by itself. I calmly released the clutch and maintained the handle bars in the direction of the skid, the bike straightened out and I slowed down to 40 mph as the bike straightened out and was none the worse for wear. I haven't done a double downshift since. Live and learn.
 

·
Registered
Joined
·
5,188 Posts
I am not so convinced that the car's data logger is telling the whole story. I would want an independent reading from an external g-force meter. If you are stepping on the brake and you feel a sudden acceleration in your gut while the car is transitioning from regen to hydraulic braking... I suspect the car actually IS loosing braking force for a few milliseconds during this transition. And if you can FEEL the surge it's probably a pretty significant event.

The car's data logger is showing what the electronics THINK is happening. (Not necessarily what is actually taking place from an actual g-force measurement.)
 

·
Registered
Joined
·
847 Posts
Discussion Starter #16
I am not so convinced that the car's data logger is telling the whole story. I would want an independent reading from an external g-force meter. If you are stepping on the brake and you feel a sudden acceleration in your gut while the car is transitioning from regen to hydraulic braking... I suspect the car actually IS loosing braking force for a few milliseconds during this transition. And if you can FEEL the surge it's probably a pretty significant event.

The car's data logger is showing what the electronics THINK is happening. (Not necessarily what is actually taking place from an actual g-force measurement.)
It would be great to get a g-force measurement. Especially since all the data I have access to is undocumented (and therefore interpreted, sometimes accurately, sometimes not). And, if there was something that was happening for a duration of less than ~100ms, the data might not appear (the important data is updated about every 50-100ms).

However, the data suggests (again, that's all it can do without help from GM!) that the regen is indeed active throughout the ABS incident. So from that perspective, there should be no reduction in deceleration as a result of ABS kicking in.

I'm not sure if a reduction in deceleration would feel the same as a switch to acceleration.
 

·
Registered
Joined
·
5,188 Posts
The fact that you are able to FEEL a sudden surge of acceleration makes me suspicious that you may actually be feeling a sudden surge of acceleration. It needs to be measured. I think there are some cheap iPhone & Android apps that measure g-force. It might be worth trying one of those.
 

·
Registered
Joined
·
1,976 Posts
I get a sense of acceleration when mistakenly in "D" (instead of "L"), I lift my foot off the accellerator. "Whoa, where did that come from? Oh, I'm in D. How did that happen?" Of course, it's not accelerating, but it sure feels like it.

The mind can play funny tricks, as I'm learning right now in detail from listening to Steven Pinker's "How the Mind Works" (on the Volt's 30 GB hard drive).
 

·
Moderator
Joined
·
8,680 Posts
Good thinking. I do have what appears to be battery flow (slightly positive when charging, large positive numbers during regen, negative numbers when not charging at idle, large negative numbers when accelerating). It shows that regen started a little over a second before the wheel speed was funky, and it continued with regen for about 4.5 seconds afterwards.

The power flow from regen does goes down, but the regen is still working (600ms after the ABS starts, the power flow is at 30% of when the ABS started).

So it looks like the regen is active even when the ABS takes over. It's possible that it may change the force of the regen to the problem wheel (or the other ones), but the regen overall does stay active.
That's interesting, I think. It sorta sounds like the car switches from pure regen to mostly friction in response to the low traction (likely because regen is less responsive (need time to collapse fields,) and can't be vectored to different wheels.) I suspect if the 30% hadn't been enough to recover the wheel, it would have gone to pure friction...
 

·
Registered
Joined
·
3,165 Posts
My ObdII/Torque has a acceleramitor in the phone.. I'll do that test to see what it does. I've been meaning to setup a do a MA on a second OBDII device and maybe this will push be over the edge to get to doing it..
 
1 - 20 of 33 Posts
Top