In a message dated 01/22/2000 6:52:11 PM Pacific Standard Time,
mvheim@earthlink.net writes:
> OK, well, the only reason I asked the question is that someone (not me)
> pointed out that sensing the wheel rotation (as you are describing) is
> NOT the same as measuring G-forces (your term) directly. For instance,
> the case that was proposed was if you ran onto a patch of ice while
> braking. The wheel rotation would come to a stop (since the brakes were
> being applied), and the ABS, based on wheel sensors alone, would be
> unable to tell the difference between the car being immobile (already
> stopped) and the car skidding on all four wheels (into the ditch).
Max, actually it's much easier than you think. The ABS designers know how
much traction is available in the best of times (if a car will pull say 0.80G
cornering then this is about what it can be expected to be able to brake to
as tire slip is the limiting factor) So the designer takes a number somewhat
higher than what the car may expected to obtain and uses any number above
that as a skid. In the example given all four wheels on ice and the brakes
applied suddenly and hard. Well, since the friction between the tires and
road surface would allow the tires to come to an almost instantaneous stop
(60 MPH to 0 in two feet) the computer would know that an 80 G stop is
impossible and ABS would come into action. Simple huh?
> The example you are giving is one where there is a traction differential
> (hence a rotational speed differential) between two or more wheels. But
> what if they all lock simultaneously? Admittedly this is a hypothetical
> situation.
>
Perhaps not hypothetical, but see above.
> And you can't state that the ABS doesn't let the wheels "lock up" (come
> to a stop), because if it didn't, you'd never be able to bring the car to
> a complete stop. My understanding is that the ABS stops working at a
> certain lower limit of speed, anyway, below 10 or 20mph.
You are correct, various systems probably have different disengagement
points. On the cars that I am familiar with disengagement is at about 2-3
MPH. BTW most systems won't come into play until a min. speed has been
reached 8-10 MPH say.
>
> Anyway, I don't really have any axe to grind in this argument, and in
> fact I looked up the ABS Education website for a full description, but in
> the interest of fairness I thought I'd point out that you haven't really
> answered the question, which might be paraphrased as "in the absence of
> direct vehicle speed and vector information, can a system based on wheel
> rotation sensors be fooled into inappropriate responses?"
Again different systems work in different manners, but on the cars I drive
and teach on I have never had any type of inappropriate response. This comes
from working, driving, and teaching on these systems since 1987. Frankly the
computer is better at this stuff than I am. Now as I have said before I have
driven other cars and been underwhelmed. As far as direct vehicle speed goes
just how are you going to measure this speed? In lieu of a 5th wheel (which
no stock car has) your choices are, cable off of trans or front axle, or
electrical off diff or a signal from ABS. All of these choices are capable
of being fooled by a locked wheel. An interesting side note. I was
interested in 0-60 times of a new model car that I got. A friend came over
with a small thing that looked like radar detector. He suctioned cupped it
to the windshield and told me to take off. It was a G meter that gave 0-60
times. The most surprising thing about the entire experience is that the
Speedo runs behind the actual speed by about 5-10 MPH during hard
acceleration.
>
> Good luck with the jury thing... you realize you'll have to disqualify
> yourself if it's a class action case involving auto braking systems!
I'm sitting on a criminal case.
Rick Ewald
|