Technical paper on ABS and Tor*en

["Scott Mo." <scottmo@teleport.com>]

Well,

Here is an excerpt from a 16 page article I came across which discusses
ABS braking and Tor*en. It was written in 1988 by a gentleman from Audi
AG Ingolstadt. It "may" shed some light as to why Audi chose to use the
Torsen, then again it may only bring more darkness....you make the call

I can provide more details on this article in private email if anyone is
interested.

Here we go.........

"In 1980, when the Audi quattro appeared as the first of a new
generation of road vehicles, there was also considerable controversy
about the braking behavior of vehicles of this type.

With four-wheel drive, the difference between maximum acceleration
ability and braking ability is naturally reduced, but this is something
that has now been accepted, and four-wheel drive has since
earned the same recognition as four wheel brakes.

Braking behavior and stability can be greatly improved by intelligent
brake force regulating systems, drive train measurements, ABS, and
peripheral electronic regulating systems.

This paper is intended to show some of the problems, solutions and
possibilities involved.

INTRODUCTION

After many years of practical experience, drivers have come to  expect
that maximum braking ability is always twice as high as maximum
acceleration.

Since the introduction of permanent four wheel drive far high
performance road cars, such as the Audi quattro, this situation is no
longer quite so simple.

Considering a single road wheel, it can be said that drive torque and
braking torque are proportional.

But with four-wheel drive, unlike two wheel drive, the drive torque and
braking torque ere both multiplied by the same factor.

A lack of understanding for such fundamental facts of physics on the
part of some drivers gave rise to certain doubts amongst owners and
experts alike when the permanent four wheel drive system was first
introduced. Particularly in winter conditions, in-experienced Quattro
drivers had the occasional sobering experience where their car came off
the road. If they were lucky, the extra traction of four-wheel drive
then at least enabled them to get back from "off-road" to "on-road"

Cases of this kind showed that the braking ability of four-wheel drive
cars really needs to be a lot better than is actually physically
possible.

So experience has shown that, providing appropriate stability can be
maintained, braking performance with four-wheel drive must be  given
much higher priority than it has in the past with two-wheel drive. 

BRAKE FORCE DISTRIBUTION

In order to achieve optimum brake force distribution it would be
"necessary to have an ideal vehicle concept which gave a weight
distribution Of 50 : 50 for all rates of deceleration.

Not surprisingly. such an ideal concept is not feasible

Some modern car designs differ considerably from this ideal case.
As a practical example:. Fig. 1 show the brake force distribution Of an
Audi 5000  with front-wheel drive.


The ideal brake force distribution with the theoretically possible
worst-case weight distribution, unladen, is 78:22 at a braking ratio of
z  =0.8.

The actual brake forces on the vehicle, in spite of the use of a
load-sensitive brake force regulator on the rear wheels, are no more
than a satisfactory approximation.

As a means of evaluating the efficiency of a brake system, We refer to
the relative amount of lost braking effective (delta v).
This factor is the difference the integral values of ideal and actual
brake force distribution. (NOTE: There is a nice graph describing this
concept)

The Audi 5000 Quattro is an example which shows how optimization of
braking performance has been achieved, firstly due to the weight
distribution inherent in the design and secondly from the  actual
braking forces on the vehicle.

The partially laden and fully laden conditions gives corresponding
improvement, which can be seen in fig. 1, 2 (sorry no figures shown
here)

BRAKE FORCE REGULATING SYSTEM, AUDI 5000 QUATTRO

(NOTE: The rear brake pressure regulation system is described in this
section (3 pages), it has been left out for brevity.)

THE INFLUENCE OF MANUAL OR AUTOMATIC DIFFERENTIAL LOCKS ON BRAKING
BEHAVIOUR

In the early stages of four-wheel drive euphoria, those involved looked
far all possible ways of showing the advantages of this drive layout.

Locking the central differential between the front and rear wheels was
found to be an effective means of improving braking ability.

This, in effect, does give the ideal brake force distribution.

But the gain in braking effect also meant a loss of cornering force at
all four wheels under lock-up conditions.

For a normal driver, especially in a panic stop where stability is
important, this kind of "improvement" is of doubtful benefit.

Furthermore, comprehensive testing has shown that the potential
improvement in the stopping distance is less than 5 meters.

Extensive tests with viscous couplings (working an the Ferguson
principle) have shown that the rear wheels are forced to lock-up because
of the response characteristics of the coupling, which is unsuitable for
normal application.

This does not apply to other concepts, which avoid this problem with
additional measures (such as a free wheel or ABS, etc.).

For Audi it is important that a particular drive system still allows
virtually free choice of brake force distribution for a specific
vehicle.

In the course of development of permanent four-wheel drive, the
philosophy behind the use of the differential locks has changed. Whereas
the differential lock was initially regarded as an effective means of
bringing the actual brake force distribution strictly in line with the
ideal distribution, its use is now only recommended for improving
traction under extremely bad conditions at low speed. The center and
rear differential locks are only necessary to free the car if it should
get stuck.

A provision for switching off this traction aid above a certain,
relatively low speed is also desirable.

THE TORSEN DIFFERENTIAL AS AN INTER-AXLE DIFFERENTIAL FOR IMPROVED
TRACTION AND BRAKING BEHAVIOR

The Torsen (torque sensing) differential is well suited for the
philosophy out-lined above.

In simple terms, this is a classic, mechanical differential with worm
gearing instead of the conventional bevel-type differential gears.

Irrespective of the torque split inherent in the car's design (5O : 50
for the Audi quattro), this type of differential is capable of varying
the distribution of torque as required, approaching the ideal torque
distribution up to a limit of 78 : 22 % and vice versa.

The "intelligence" of the Torsen differential in the context of braking
behavior lies in the way its function is reversed on overrun (closed
throttle)

The two driven axles are almost entirely de-coupled, which enables free
brake force distribution, unrestricted   ABS operation, and significant
improvements in stability.

Series of tests on ice surfaces with friction values of uf <    0.1, and
at low ambient temperatures, have confirmed that the overrun torque with
a Torsen differential is equivalent to that of a conventional free
differential."

(NOTE: Description of the typical ABS brake systems follows this, (4 or
5 pages) with some  discussions regarding the braking problems that
occur with the differentials locked.)

(Jump to the Conclusion...)


CONCLUSIONS, FUTURE DEVELOPMENT TRENDS

The headlong developments over the past few years have shown that
four-wheel drive and ABS can be combined to form a well-balanced whole.

The safety benefits of four-wheel drive are matched in equal measure by
the efficiency of an electronically regulated braking system

Both today and in future, stability and steer-ability will have priority
over all other factors.

That is why the ABS on earlier Audi cars was switched off automatically
when the driver enabled the manual differential locks. At first sight
this may have appeared questionable, but bearing in mind Audi's
philosophy it was in fact logical, in order to retain the order of

priorities:

(  stability
( steer-ability 
( shorter stopping distances

in all possible conditions.

>From current development activities in the field of ABS, we can identify
the following main tendencies:

1. Efforts to improve the control performance of present systems by
a)  improving reference information by means of additional data such as
longitudinal/later acceleration, etc.
b)  and taking the engine overrun torque into account by intervention in
the power train, and active engine control.

2.  The development of more inexpensive fully and partially integrated
systems while still maintaining the current safety levels.

END OF ARTICLE

HTH

-- 
Scott Mo.
1989 200TQ
1988 5000TQ
1984 4000S
1966 VW Beetle
http://www.teleport.com/~scottmo