I would like more information on ABS

(it beats "learn me")

While driving the wife's car we encountered a "panic brake" situation where I felt the ABS kick in, but I "felt" like we shouldn't be near the limits of the available grip yet.

So I ask: How does ABS know when to kick in?

(you may skip this next section if you wanted to)

The car in question is a 2006 Audi A4 Wagon, diesel, automatic, FWD running on almost brand new Continental Premium Contact tires, brand new pads and rotors all around (probably 95% pad life left). Tire pressure was about 42PSI, just short of the manufacturers maximum recommended tire pressure. Pavement was dry and warm. We had a roof box installed with about 130lbs of luggage, three people in the car (my wife, myself, and my son who is one year old). The back was also pretty full of luggage, plus some tools.

I stomped on the brake pedal with the reassurance that I could "plant that thing to the floor" and "trust" ABS to do the rest. We were coming down from about 60kph (I think that's around 35 or so MPH). We stopped relatively quick, but I "feel" like the brakes were a bit "slippery". I "feel" like the tires had a bit more to give. I didn't feel like the tires were "giving" yet, and they certainly weren't making any noise in protest (not even a squeak).

So: Is ABS a magical tool that finds the limits of the available traction and then brings you as close as possible to the "ragged edge" of under braking to allow you to make the quickest, safest, stop possible?

or

Is ABS a generic device that, once programmed, is set to a certain traction "coefficient" that ends up "underutilizing" better gripping tires than those fitted by the OEM thus negating any "advantage" that the better gripping tires might give?

OR!

is it something in between?

Thanks for reading...

ABS pretty much..... well, I know how older ABS works by checking individual wheels against each other and also a known table of how fast a wheel can stop (looking for lockup) but my newer cars have a metric ass load of yaw sensors and accelerometers and I'm not sure I know any more.

Interesting. Most of the aircraft I work on have "wheel transducers". If a wheel on the left gear isn't spinning but its "partner wheel" on the right side is, then it backs the brake pressure off of the stopped wheel.

I didn't know if that was the same concept in automotive applications.

Did you consider that when you felt ABS working, it wasn't "doing ABS" on all wheels? For example, you get massive weight transfer and say poor traction on one wheel and ABS reduces the pressure on that line, while you're still getting full brake application (or more, depending on the system) that your foot is asking for on the other three lines. So while you might not have thought one wheel was locking up, perhaps it actually was? Also, modern systems are so fast that the wheel doesn't have to actually LOCK (as in anti-lock) but be slowing down so quickly that it's obviously about to lock without intervention...then the system intervenes. Where that intervention occurs depends on the system tuning, obviously.

ABS is getting pretty complex, so it incorporates stability control, traction control, etc. these days as well. That can also affect what you perceive should happen versus reality.

Bryce

"Stomping" might be the problem, shocking the tires into loosing traction is a bad way to get the most grip out of them. The stomp caused initial lock up then the abs struggled to regain traction. Think of the tire's grip as silly-putty, when you play with silly putty, and you yank it suddenly, it snaps, right? But if you pull slower it stretches and doesn't break. Your tires act like that. Try back to back reenactments of this scenario one time stomping, one time progressively increasing braking input over a second or so to the same pedal pressure, then tell us what happened.

It just needs to see one or more wheels spinning slower than the others. Doesn't mean that that wheel is locked up, just spinning a few MPH slower than the others.

As mentioned, the newer systems (like in your Audi) also use yaw sensors and brake line pressures to determine when to interfere. A tire with a little lower tire pressure can set it off too. The smaller diameter of the under-inflated tire will look like a different road speed to the wheel speed sensor.

Does 42 psi sound a little high to anybody else?

carbon wrote: Does 42 psi sound a little high to anybody else?

Not really. I see pressures that high listed on the inflation sticker pretty often, especially on wagons. It will definitely decrease traction though.

Silly of me in hindsite, but I figured if the ABS was kickin in then ALL the tires were on or about their limit.

I really just felt the pedal start to do its "shimmy shake" and said: "already? It cant be..." (like I've never said that before).

I'm not sure how advanced my stability control is on the Audi, I don't remember seeing a means of "interface" to the dampers. No wires or anything that might indicate an active stability control...

Could also be like you said: one tire sitting on some dirt on the pavement while the rest were giving 100%. ABS kicks in, and in my mind I'm thinking that ALL the tires are begging for mercy.

So that kind of brings me back to one of my questions about "system tuning": If I put on some super sticky tires, do you think the ABS is capable of taking full advantage of them?

OOH!

Dear Grassroots Editors:

Could you do a "stopping distance test" with a couple different "ages" of ABS systems with tires of different "grip" levels and see how the stopping distance is affected?

Respectfully, -Bill

carbon wrote: Does 42 psi sound a little high to anybody else?

Yes, that's rallycross anti de-beading pressure right there.

As a side note, when I was running the computer watching things for autocross, it was very rare to have more than one wheel lock at any time, even under hard straight line braking. I'm not sure I learned anything about driving by recording data, but it was really cool.

The roads here are pretty awful and they only get worse in countries like Albania. "Unimproved" would be the best compliment I think I could give them. I figured the higher than normal, but still under the MFGR's max, would be a good idea for the trip.

Seemed to work

The abs on my E150 will kick in if one wheel loses traction. Notable, I ran over a large socket stopping at a light. That was enough to trip the abs system. Manhole covers in the rain will do it too. I would hope you Audi would be smarter.

Hungary Bill wrote: Silly of me in hindsite, but I figured if the ABS was kickin in then ALL the tires were on or about their limit. I really just felt the pedal start to do its "shimmy shake" and said: "already? It cant be..." (like I've never said that before). I'm not sure how advanced my stability control is on the Audi, I don't remember seeing a means of "interface" to the dampers. No wires or anything that might indicate an active stability control...

I am by no means an expert on modern ABS/Stability/traction control systems, but I do know that vast majority of them work entirely through the braking system, & many can apply pressure to individual brake calipers with out the help of the driver using the brake pedal. Very few have any active connection to the dampers, in fact I'm pretty sure that only the super high end stuff has that ability.

So what I'm getting at is that in addition to what has already been said about the possibly of just one wheel needing some intervention from the ABS, there could also have been some stability control action happening simultaneously.

.

ABS is either a minor irritant, or the work of the devil, depending upon which group of lawyers held the whips over the guys programming it. Most Euros are fairly unobtrusive, and will allow a bit of wheel slip before it really starts to get nuts. GM, on the other hand, will shut all the brakes off by pushing the pedal and your foot up when it senses that one of your tires has hit a pothole and momentarily slowed as it bounced. Once its goddam good and sure that there is no danger of the tires ever slowing down, it may allow you to get some braking done just before you are hit by the crosstown bus.

I pulled the fuse on my truck the third time it terrified me.

Toyman01 wrote: The abs on my E150 will kick in if one wheel loses traction. Notable, I ran over a large socket stopping at a light. That was enough to trip the abs system. Manhole covers in the rain will do it too. I would hope you Audi would be smarter.

Huh? ABS should kick in, for that one wheel. You realize that ABS has multiple channels, right? Current systems have one channel per wheel, so it can reduce pressure on the single wheel that's lost traction without affecting any of the others. Older systems were 3 channel (2 fronts, 1 rear, which works OK) or 2 channel (diagonal, which pretty much sucks). I think there might also have been 1 channel systems for the pickup truck rear-only ABS systems which were just stupid.

The multi-channel thing is why a good race driver, threshold braking, on a dry, straight, flat pieces of pavement with a properly tuned brake bias might be able to beat a good ABS system in stopping distance, but will pretty much always lose out in real world street driving where none of those qualifiers are true. He'd need four brake pedals and four feet to do it properly.

A surprising amount of modern cars and trucks do not have 4 channel abs. especially domestic trucks and vans.

http://jalopnik.com/this-is-how-abs-esc-and-traction-control-work-513807036

The ABS on my Fiesta allows a chirp from the tires when braking hard at an auto cross. At least one anyway.

Pretty sure the van has 3 channel ABS. It feels like the Ford system bleeds pressure off that one brake, which gives the soft pedal feel. The system probably works great in an emergency situation when your foot is buried in the pedal. It pretty much sucks when you are stopping normally and it kicks in because you hit a bump or a socket in the road. Over all the system is too sensitive, I'd rather get a little bit of lockup before it kicked in. The 02 van is much better than the 97 I used to have. It doesn't bother me enough to disable it, and who knows, I might actually need it one day.

VLV4LF wrote: http://jalopnik.com/this-is-how-abs-esc-and-traction-control-work-513807036

i read the article but did not watch the videos.

the article gives a pretty comprehensive and accurate description of how the systems work. thanks for saving me a lot of typing!

comments to specific posts above:

• @ Carbon: the systems are fast enough now that the only instance i can think of in which stomping might push a tire too deep into slip, thereby requiring longer recovery time on that tire, would be the rear axle of an unloaded truck, where the brakes are sized to be able to stop the vehicle when it's at GVW.

• @ Hungary Bill: the systems are calibrated for the OE tires. some systems have a "maximum possible deceleration for this vehicle" value in teh software, and in this case it is possible that the system will not allow the full decel capability of a really sticky tire to be utilized. the following numbers are for example only, to demonstrate the concept. imagine a car that came with all-seasons now wearing Hoosier A6s. if the OE tire was good for 0.95g decel in perfect conditions, the OE system might have a "max possible decel" programmed at something like 1.1g. the A6s might be good for 1.2g in the same conditions, but the computer thinks 1.1g is the max the vehicle could ever pull, so it considers 1.2g to be all four wheels going into unstable slip simultaneously, and it may go into a pressure regulation routine to maintain vehicle decel at 1.1g.

• @ HappyAndy: The only ABS systems that can apply higher than driver's command are those which are also equipped with a "Panic Brake Assist" feature, which senses the rate at which pressure was initially built by the driver's command, and if that pressure rise rate was above the "Panic Brake" threshold calibration, the ABS pump is instructed to build sufficient pressure to cycle ABS on all channels, until the driver reduces their input to the pedal. AFAIK, "Panic Brake" features ALWAYS reduce pressure when the pedal sensor (or pressure sensor depending on how the system is mechanized) sees driver command being reduced.

• @ Streetwiseguy: there is no ABS system now or ever in the past produced by any US manufacturer which "shuts off ALL brakes..." according to your description. perhaps you didn't like the perceived reduction in deceleration when the system isolated the affected wheel-end(s) to prevent them from going deeper into slip, but your description is, shall we say, in-credible.

• @ Toyman: vans are a lot like trucks. the foundation brakes are sized to stop a GVW vehicle, and the ABS calibrations are developed at LLVW, GVW, and a few loadings in between. sometimes the occasional false activation at LLVW on a bumpy surface is a compromise that was accepted by the vehicle manufacturer in trade for better wheel control at GVW, or on some other surface, etc. we don't do it just to be annoying! ;-)

@ Streetwiseguy: there is no ABS system now or ever in the past produced by any US manufacturer which "shuts off ALL brakes..." according to your description. perhaps you didn't like the perceived reduction in deceleration when the system isolated the affected wheel-end(s) to prevent them from going deeper into slip, but your description is, shall we say, in-credible.

Yes, but:

A surprising amount of modern cars and trucks do not have 4 channel abs. especially domestic trucks and vans.

I've experienced two scenarios where i essentially lost braking because an ABS channel was shared with 2 wheels. Once was going down a steep hill with a space saver spare on a front tire. That wheel locked up, it pulsed the channel that resulted in less braking to ANOTHER wheel, and i had only 2 brakes left to stop the vehicle on the steep hill, which barely worked.

The other scenario was that i was braking while changing lanes into a turn lane to make a left turn at speed (no oncoming traffic). My left front tire ran over a piece of sheet metal (flashing or ducting of some sort), and the tire grabbed it and slid it down the road. So the abs system sees that wheel as fully locked and pulses braking in that circuit. So now i have inadequate braking AND horrendous understeer to get out of a turn into oncoming lanes that i'm already sort of committed to. I slid totally past the intersection and partially into an oncoming lane and had to pull a U-turn and go back to the intersection i missed.

In both cases i essentially 'lost braking' to a dangerous degree because the ABS was not a 4-channel system. Both cases COULD have resulted in accidents if i'd been unlucky. In both cases, my opinion is that it would have been entirely the fault of the ABS system.

Vigo wrote: I've experienced two scenarios where i essentially lost braking because an ABS channel was shared with 2 wheels. Once was going down a steep hill with a space saver spare on a front tire. That wheel locked up, it pulsed the channel that resulted in less braking to ANOTHER wheel, and i had only 2 brakes left to stop the vehicle on the steep hill, which barely worked.

Fronts are ALWAYS controlled separately. There is no such thing as a two-channel ABS on a passenger vehicle or light truck. a front/rear system can be either 3- or 4-channel. i believe all diagonal split systems are 4-channel. No matter whether it's a front/rear or a diagonal split foundation brake system, the outputs of the MC go into the ABS HCU. Inside the ABS HCU, each channel has it's own isolation and dump valves, with dedicated fluid paths to each wheel end brake.

What you experienced was a stack-up of (1) reduced grip on one front corner due to the space-saver, combined with (2) more front weight and (3) less rear weight due to heading down a steep hill. your situation reduced your rear vertical force, which limited how much braking the rears could do. your space-saver limited how much grip you had on that front corner. End result is reduced braking, for sure, but can't be blamed on the ABS.

Vigo wrote: The other scenario was that i was braking while changing lanes into a turn lane to make a left turn at speed (no oncoming traffic). My left front tire ran over a piece of sheet metal (flashing or ducting of some sort), and the tire grabbed it and slid it down the road. So the abs system sees that wheel as fully locked and pulses braking in that circuit. So now i have inadequate braking AND horrendous understeer to get out of a turn into oncoming lanes that i'm already sort of committed to. I slid totally past the intersection and partially into an oncoming lane and had to pull a U-turn and go back to the intersection i missed. In both cases i essentially 'lost braking' to a dangerous degree because the ABS was not a 4-channel system. Both cases COULD have resulted in accidents if i'd been unlucky. In both cases, my opinion is that it would have been entirely the fault of the ABS system.

on the second one, and on your conclusion, i'd have to say your opinion is berkeleyed. how is it entirely the fault of the ABS that there was a piece of sheetmetal laying in the road, that you ran over it, that it stuck to your tire, and that you were going too fast to recover? if the sheetmetal truly stuck to your tread and was between your tire and the road, then you're dealing with the friction coefficient of sheetmetal to road, not tire to sheetmetal and not tire to road.

and let's think about the effect of reduced brake force on a single front tire. reducing the longitudinal (braking) demand on a tire increases that tire's lateral force capability. So, by dumping pressure to a front tire in a turning maneuver, the front lateral force capability goes up! So in the scenario as you describe it, the ABS actually gave you more front turning grip. perhaps not enough to allow you to negotiate the radius you desired, on the surface you found yourself, at the speed with which you entered the maneuver, or with the amount of steering you dialed in.

remember that "A lat equals V squared over R", which means the amount of lat accel you're requesting is equal to the square of the speed you're traveling, divided by the radius of the path.

also remember that "A max equals mu peak", which means that the maximum acceleration (in any direction) that a tire can generate, is equal to the peak of the mu-slip curve of that tire on that surface.

so, if you're going in too fast for that radius on that surface, you saturate the tires and your vehicle seeks a larger radius, ie you go off the outside.

sorry man, but the ABS is not to blame in either of your scenarios.

AngryCorvair wrote: - @ Streetwiseguy: there is no ABS system now or ever in the past produced by any US manufacturer which "shuts off ALL brakes..." according to your description. perhaps you didn't like the perceived reduction in deceleration when the system isolated the affected wheel-end(s) to prevent them from going deeper into slip, but your description is, shall we say, in-credible.

We have had this discussion before. Again, I invite you to Saskatoon, where I will be happy to demonstrate in almost any General Motors vehicle built a decade or so ago.

We can then drink beer.

If the tire has never breaks grip, it never has to regain traction, in the time that the stopped wheel is struggling to regain grip, the tire that has traction is slowing the vehicle, building temperature (and thusly increasing grip), and loading (increasing grip). I dont deny that some of the new abs systems are impressive, but they dont trump physics. Like I said in my original post, dont take my word for it, test different methods for yourself, see what works best. Another thing that is commonly a factor in long stopping distances with abs is that the driver senses abs activation and decreases pedal pressure, releasing pressure on the corners that have grip, increasing stopping distances.