# Federal Motor Vehicle Safety Standards; Hydraulic Brake Systems; Passenger Car Brake Systems

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A95-2324

## Record

- **Collection:** Federal Register
- **Document type:** Rule
- **Published:** February 2, 1995
- **Citation:** 60 FR 6411

## Text

DEPARTMENT OF TRANSPORTATION

National Highway Traffic Safety Administration

49 CFR Part 571

[Docket No. 85-06; Notice 8]
RIN 2127-AA13

Federal Motor Vehicle Safety Standards; Hydraulic Brake Systems;
Passenger Car Brake Systems

AGENCY: National Highway Traffic Safety Administration (NHTSA),
Department of Transportation.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: This rule establishes a new Federal motor vehicle safety
standard, FMVSS No. 135, Passenger Car Brake Systems, and replaces
Standard FMVSS No. 105, Hydraulic Brake Systems, as it applies to
passenger cars. NHTSA's decision to establish the new standard results
from the agency's efforts to harmonize its standards with international
standards. The agency has determined that this new standard will
achieve the goal of international harmonization while remaining
consistent with the statutory mandate to ensure motor vehicle safety.

DATES: Effective Date: The amendments made by this rule are effective
March 6, 1995. As of this date, manufacturers have the option of
complying with either FMVSS No. 105 or with FMVSS No. 135. Compliance
with FMVSS No. 135 becomes mandatory on September 1, 2000.
Petitions for Reconsideration: Any petition for reconsideration of
this rule must be received by NHTSA no later than March 6, 1995.

[[Page 6412]]

ADDRESSES: Petitions for reconsideration should be submitted to:
Administrator, National Highway Traffic Safety Administration, 400
Seventh Street SW., Washington, DC 20590.

FOR FURTHER INFORMATION CONTACT: Ms. Terri Droneburg, Office of Vehicle
Safety Standards, National Highway Traffic Safety Administration, 400
Seventh Street SW., Washington, DC 20590 (202) 366-6617.

SUPPLEMENTARY INFORMATION:

TABLE OF CONTENTS

I. Background
A. Federal Motor Vehicle Safety Standards
B. European Braking Requirements
C. Harmonizing US and European Braking Requirements
D. Antilock Brake Systems
II. Summary of comments on the 1991 SNPRM (Notice 5)
III. NHTSA Decision
A. Overview
B. Application
C. Definitions
D. Equipment Requirements
1. Lining Wear Indicator
2. ABS Disabling Control Switch
3. Vehicle and Reservoir Labeling
4. Brake System Warning Indicator
E. General Test Conditions
1. Ambient Temperature
2. Road Test Surface
3. Instrumentation
F. Road Test Procedures and Performance Requirements
1. Permissible Wheel Lockup
2. Road Test Sequence
3. Pre-Burnish
4. Burnish
5. Adhesion Utilization
a. General
b. Wheel Lock Sequence Test
c. Torque Wheel Test
6. Cold effectiveness
7. High speed effectiveness
8. System failure
a. Stops with Engine Off
b. Antilock Functional Failure
c. Variable Proportioning Functional Failure
d. Hydraulic Circuit Failure
e. Power Assist Unit Inoperative
9. Parking brake requirements
a. Dynamic
b. Static
10. Fade and Recovery
a. Heating Snubs
b. Hot Performance
c. Recovery Performance
G. Miscellaneous Issues
IV. Regulatory analysis
A. Executive Order 12866 and DOT Regulatory Policies and
Procedures
B. Regulatory Flexibility Act
C. Executive Order 12612 (Federalism)
D. Executive Order 12778 (Civil Justice Reform)
E. National Environmental Policy Act

I. Background

A. Federal Motor Vehicle Safety Standards

The National Traffic and Motor Vehicle Safety Act (``the Safety
Act''), recently revised and codified ``without substantive change'' at
49 U.S.C. Chapter 301, authorizes the National Highway Traffic Safety
Administration (NHTSA) to issue Federal motor vehicle safety standards
(FMVSS) to ensure motor vehicle safety. The Safety Act requires that
each FMVSS be objective and practicable so that a manufacturer can
certify that each of its vehicles meets all applicable standards. Each
FMVSS specifies the performance requirements and any necessary test
conditions and procedures that NHTSA uses in its periodic tests of
motor vehicles and motor vehicle equipment. Each tested vehicle must
meet the objective requirements contained within the applicable FMVSS.
Under this self-certification system, the government does not
subjectively approve or disapprove a type of vehicle or a type of
braking system.

B. European Braking Requirements

Unlike the self-certification system used in the United States, the
European community has established a ``type approval'' system in which
the government approves each type of motor vehicle or item of motor
vehicle equipment, based on whether it can meet the safety
requirements. For example, the current United Nations Economic
Commission for Europe (ECE) braking regulation, Regulation 13 (R13) and
its proposed harmonized regulation, R13H, use a calculation method to
determine the adhesion utilization of a vehicle as designed.
Manufacturers submit their calculations (or the input parameters
necessary to make the calculations) to governmental authorities along
with a prototype vehicle, and the governments then approve or
disapprove the vehicle type based on a review of those calculations and
testing of actual vehicles.

C. Harmonizing US and European Braking Regulations

In order to eliminate any unnecessary non-tariff barriers to trade
in accordance with the General Agreement on Tariffs and Trade (GATT),
the United States has participated in discussions held within the
Meeting of Experts on Brakes and Running Gear (GRRF) of the ECE. As a
result of these discussions, NHTSA has issued a series of rulemaking
notices proposing to establish a new FMVSS, FMVSS No. 135, Passenger
Car Brake Systems. Likewise, the GRRF has also developed a proposed new
Regulation 13-H, which would be compatible with FMVSS No. 135.
Throughout the rulemaking, NHTSA has emphasized that any requirements
it adopts must be consistent with the need for safety and the Safety
Act. The agency emphasizes that safety cannot be sacrificed in its
efforts to harmonize the FMVSS with the ECE regulations.
On May 10, 1985, NHTSA published in the Federal Register (50 FR
19744) a notice of proposed rulemaking (NPRM; Docket 85-06, Notice 1)
to establish FMVSS No. 135, which would replace FMVSS No. 105 as it
applies to passenger cars. On January 14, 1987, NHTSA published in the
Federal Register (52 FR 1474) a supplemental notice of proposed
rulemaking (SNPRM; Docket 85-06, Notice 4), to improve and refine the
proposed Standard. On July 3, 1991, NHTSA published in the Federal
Register (56 FR 30528) a second SNPRM (Docket 85-06, Notice 5) as a
result of comments on the SNPRM and vehicle testing by NHTSA.
In these previous notices, NHTSA set out its overall approach to
developing the proposed harmonized standard. The agency stated that the
new standard would differ from the existing one primarily in containing
a revised test procedure based on harmonized international procedures
developed during discussions held between NHTSA and GRRF. NHTSA stated
its belief that the new FMVSS would ensure the same level of safety for
the aspects of performance covered by FMVSS No. 105, while improving
safety by addressing some additional safety issues. The agency proposed
establishing new adhesion utilization requirements that it believes
would ensure stability during braking under all friction conditions.
In this final rule, after considering the public comments on all of
the notices, NHTSA has made several minor revisions to the requirements
proposed in the July 1991 SNPRM. This document explains the changes
incorporated in the final rule and the reasons for the agency's
decision.
D. Antilock Brake Systems

One issue that NHTSA considered during the process of developing a
harmonized standard was what requirements are appropriate for vehicles
equipped with antilock brake systems. While NHTSA was evaluating
comments to the July 1991 SNPRM, Congress enacted the Highway Safety
Act of 1991, which directs NHTSA to publish an advance notice of
proposed rulemaking (ANPRM) to consider the need for additional brake
performance

[[Page 6413]]
standards for passenger cars, including ABS standards. (59 FR 281,
January 4, 1994.) Vehicles included in this evaluation effort are
passenger cars, light trucks, and multi-purpose vehicles (MPV's).
Given that NHTSA is reviewing the need for antilock systems
separately, the agency has decided not to include requirements
addressing ABS performance in this final rule to establish FMVSS No.
135. The previously proposed section on ABS will be reserved until all
the issues in the research program have been evaluated. At that time,
the agency will consider how best to proceed with requirements
applicable to ABS on light vehicles and may initiate a separate
rulemaking for that purpose.

II. Summary of Comments on the July 1991 SNPRM (Notice 5)

Over 30 commenters responded to the July 1991 SNPRM. Commenters
included vehicle manufacturers, brake manufacturers, international
organizations, safety advocacy groups, and individuals. The commenters
addressed a wide range of topics, including adhesion utilization, the
various effectiveness requirements, equipment requirements such as the
failure warning indicators, and test conditions such as the road test
surface, lockup conditions, burnish procedures, and the
instrumentation.
Advocates for Highway and Auto Safety (Advocates) and the Center
for Auto Safety (CAS) generally opposed the supplemental proposal,
believing that the proposed FMVSS No. 135 was less stringent than FMVSS
No. 105 and the previous harmonization proposals. Advocates and CAS
opposed several specific proposals in the 1991 SNPRM, including the
increase in certain stopping distances, eliminating automatic brake
warning indicators, specifying certain aspects of the new adhesion
utilization test, eliminating the pre-burnish test, changing the
burnish testing procedure and the fade and recovery sequence, allowing
momentary wheel lockup, and introducing peak friction coefficient (PFC)
values as a substitute for skid numbers in defining the adequacy of
testing surfaces.
In contrast, the former Motor Vehicle Manufacturers Association
(MVMA),1 General Motors (GM), Ford, Chrysler, and manufacturers
from Europe and Japan have strongly supported harmonized safety
standards in general and a harmonized passenger car brake standard in
particular. For instance, GM stated that the payoff for successfully
harmonizing brake regulations is significant. When the U.S. and
European regulations are commonized, it is most probable that this
uniform set of requirements will be recognized and accepted throughout
all vehicle importing and exporting countries. This will enable
manufacturers to build vehicles with standardized brake systems
acceptable throughout the world, thereby providing significant cost
savings to vehicle buyers. It continued that harmonization of brake
regulations will also represent an important milestone in the ongoing
efforts to commonize motor vehicle safety regulations, and thereby
dismantle one of the most significant non-tariff barriers to
international motor vehicle trade.

\1\ The MVMA became the American Automobile Manufacturers
Association in early 1993. This notice will refer to the group by
its former name, MVMA. The membership of the new group is slightly
different than that of the MVMA, and to refer to the group by its
new name would lead to imprecision in indicating which manufacturers
were represented by its comments.
---------------------------------------------------------------------------

Notwithstanding their general support for harmonization, vehicle
manufacturers expressed concern about what they perceive as the
increased stringency of portions of FMVSS No. 135 in relation to FMVSS
No. 105.

III. NHTSA Decision

A. Overview
After reviewing the comments, NHTSA has decided to establish FMVSS
No. 135, with respect to hydraulic brake systems on passenger cars. The
new standard includes equipment requirements, dynamic road test
requirements, system failure requirements, and parking brake
requirements, as well as test conditions and procedures related to
these requirements. With respect to the equipment requirements, FMVSS
No. 135 includes provisions addressing the brake lining wear indicator,
an ABS disabling switch, reservoir labeling, and a brake system warning
indicator. With respect to the test conditions, FMVSS No. 135 includes
provisions addressing the ambient temperature, the road test surface,
instrumentation, and the initial brake temperature. With respect to the
dynamic road tests, FMVSS No. 135 includes provisions addressing
permissible wheel lockup, the test sequence, burnish, the wheel lock
sequence test, the torque wheel test, the cold effectiveness test, the
high speed effectiveness test, the hot performance test, and the fade
and recovery test. FMVSS No. 135 also includes requirements for a
static parking brake test and several types of system failure tests,
including stops with the engine off, ABS functional failure,
proportional valve functional failure, hydraulic circuit failure, and
power assist failure.
The following discussion follows the order set forth in the
regulatory text for FMVSS No. 135 to facilitate the reader's
understanding of the issues.

B. Application

In each previous proposal, NHTSA proposed that FMVSS No. 135 would
apply to passenger cars. Kelsey-Hayes asked whether this definition
included all purpose vehicles, mini-vans, and light trucks.
NHTSA notes that 49 CFR 571.3 defines passenger car, multipurpose
passenger vehicle, and truck. All purpose vehicles and mini-vans
ordinarily come within the definition of multipurpose passenger
vehicle. At this time, FMVSS No. 135 will apply only to passenger cars
and not to multipurpose passenger vehicles or trucks, although
application to other types of vehicles may be considered at a later
date.

C. Definitions

In the 1991 SNPRM (Notice 5), NHTSA proposed definitions for
certain terms, including directly controlled wheel and antilock brake
system.
Bendix and Mercedes Benz requested a clarification of the
definition of an ABS ``directly controlled wheel.'' Bendix recommended
that the definition include a select average or drive shaft sensor
control of an axle, which it believed would provide sufficient accuracy
to control individual wheel slip, thereby avoiding adhesion utilization
testing. GM commented that the definition in the 1991 SNPRM would
prohibit a type of ABS control known as ``select low'' that uses a
single, centrally located sensor on the rear axle to partially control
the systems operation.
Given that NHTSA is considering whether to equip vehicles with ABS
in a separate rulemaking, the agency has decided that it is not
necessary at this time to define ``directly controlled wheel.''
Accordingly, this term is not included in the definition section of the
regulatory text. The agency may revisit this issue if the agency
decides to propose requirements for antilock brakes on passenger cars.
The agency has included a new definition for ``antilock brake system.''
The GRRF and Fiat requested that the definition of initial brake
temperature be based on the temperature of the hottest service brake
rather than the average of both brakes on an axle, claiming that there
should be little difference in the ``cold'' temperature across each
axle.

[[Page 6414]]

After reviewing the comments, NHTSA has determined that there is no
reason to modify the proposed initial brake temperatures. Commenters
provided no convincing data or arguments to support their requested
changes to initial brake temperatures that have been proposed in the
NPRM and the two SNPRMs.

D. Equipment Requirements

1. Lining Wear Indicator
In the 1991 SNPRM (Notice 5), NHTSA proposed that the harmonized
standard include requirements to warn the driver about excessive brake
wear. Specifically, this warning could be done either by a device that
warns a driver that lining replacement is necessary or by a device that
provides a visual means of checking brake lining wear from outside the
vehicle. The agency believed that this proposal would reduce the
likelihood that cars would be driven with excessively worn brake
linings.
Advocates recommended that all cars have an in-cab visual or
audible alarm, stating that an outside visual check would be
ineffective, therefore resulting in many owners being unaware of brake
lining deterioration. Advocates further stated that the increasing
intervals between maintenance checks required of newer cars means that
repair personnel would not have an opportunity to discover brake lining
wear before it reaches dangerous levels. Honda commented that, for drum
brakes, inspection holes on drums may be insufficient to spot the areas
of worst brake wear, and recommended allowing removal of the brake
drum.
After reviewing the comments, NHTSA continues to believe that the
proposed requirements for warning drivers about excessive brake wear
are appropriate. Section S5.1.2 of FMVSS No. 135 requires a
manufacturer to warn of worn brake linings in one of two ways: (1) An
acoustic or optical device warning the driver at his or her driving
position, or (2) a visual means of checking brake lining wear from the
outside or underside of the vehicle, using tools or equipment normally
supplied with the vehicle. The agency notes that FMVSS No. 105 does not
require an in-cab warning indicator. Based on this fact, the agency
disagrees with Advocates about the need to mandate an in-cab visual or
audible alarm.
NHTSA has decided not to adopt Honda's request to allow the removal
of the drum brake to identify the wear status. The agency believes that
it has provided appropriate ways to determine excessive brake wear. The
agency is concerned that adopting Honda's request might be detrimental
to safety.
VW, Fiat, Mercedes Benz, GRRF, and Toyota requested that the agency
permit the use of the International Organization for Standardization
(ISO) brake symbol, a circle with two arcs outside the circle on
opposite sides, for the brake wear indicator in lieu of the proposed
words. The commenters stated that symbols are more appropriate for a
harmonized standard.
NHTSA has decided to permit use of the ISO symbol as a supplement
to the words ``brake wear.'' Nevertheless, the agency believes that it
would be inappropriate to allow only the ISO symbol as an alternative
to the required words. The agency believes that the symbol's meaning
would be unclear or ambiguous to a driver, since in this country they
are not generally understood to represent the concept of brake wear.
2. ABS Disabling Control Switch
In the 1991 SNPRM (Notice 5), NHTSA proposed (S5.3.2) to prohibit,
for vehicles equipped with ABS, a manual control that would fully or
partially disable the ABS. Previous notices did not address an
automatic disabling switch. The subject was discussed within GRRF,
however, and it was decided that R13H would not allow a disabling
switch.
JAMA, and Toyota requested a change in the regulatory text to
permit ABS disabling switches for off-road vehicles. The commenters
stated this is necessary because ABS tends to lengthen stopping
distances in rough, gravelly, or muddy terrain. MVMA, Chrysler and Ford
opposed permitting a manual ABS disabling switch, but wanted the agency
to allow an intelligent or automatic switch (i.e., one not controlled
by the vehicle occupants) to accommodate off-road conditions.
NHTSA has decided not to permit either a manual or an automatic ABS
disabling switch. The agency notes that no commenter requested any kind
of ABS-disabling switch for passenger cars, which are the subject of
this rulemaking. Moreover, Mercedes, MVMA, Ford, and Chrysler stated
that passenger cars should not have an ABS disabling switch. While
those commenters favoring an ABS disabling switch focused on its use
for off-road vehicles, FMVSS No. 135 applies only to passenger cars as
defined in Sec. 571.3(b). These definitions preclude including MPV's as
passenger cars. The agency therefore believes that there is no reason
to permit an ABS-disabling switch under the new standard.
3. Vehicle and Reservoir Labeling
In the 1991 SNPRM (Notice 5), NHTSA proposed requirements for the
reservoir label in S5.4.3 and the warning indicators in S5.5.5. The
agency tentatively concluded that it would be inappropriate to allow
use of ISO symbols with respect to these devices, except that such
symbols could be used in addition to the required labeling to enhance
clarity. The agency noted that this was consistent with FMVSS No. 101,
Controls and Displays and past agency decisions made in response to
petitions for inconsequential noncompliance based on the use of ISO
symbols in place of words or symbols required by FMVSS No. 101.2
The agency has denied these petitions in cases where it believed that
the symbol's meaning would not be readily apparent to drivers.

\2\NHTSA notes that FMVSS No. 101 allows the use of some ISO
symbols, but not the ones at issue.
---------------------------------------------------------------------------

VW, Fiat, Mercedes Benz, and Toyota commented that the agency
should permit use of the ISO brake symbol in FMVSS No. 135 in lieu of
the words ``brake,'' ``park,'' or ``parking brake,'' and in lieu of the
words ``ABS'' or ``anti-lock'' for ABS failure. GRRF stated that
symbols are more appropriate for international use than words in any
single language.
Notice 5 and this final rule (Section S5.5.5(a)) allow the use of
ISO symbols in addition to the required labeling for the purpose of
clarity. However, the agency has decided not to allow the ISO symbol
alone to be used as a substitute for the required words. NHTSA believes
that the ISO symbol can be ambiguous to some drivers since the ISO
symbol, is not universally understood to represent brakes. The agency
notes that the commenters did not provide any data showing that the ISO
brake failure warning indicator is clearly understood by drivers in
countries in which it is currently in use. Moreover, the meaning of the
symbol is not readily apparent from its appearance, in contrast to some
symbols, such as the one for horns, whose meaning is understandable on
its face.
Fiat and the GRRF requested that S5.4.3 be amended to allow the ISO
brake fluid symbol to be used on the brake reservoir instead of DOT
fluid designations.
NHTSA has decided not to allow the ISO symbol instead of the DOT
brake fluid designations (e.g., DOT 3, DOT 4, and DOT 5). The purpose
of this requirement is to inform drivers about what kind of brake fluid
to add to their vehicles and to avoid use of an improper fluid. The
agency notes that

[[Page 6415]]
the ISO has no rating equivalent to DOT 5 fluid and does not
differentiate between DOT 3 and DOT 4 fluids. Even though the agency
has decided not to allow use of the ISO symbol, a manufacturer may use
the ISO symbol as a supplement to the required textual words.
4. Brake System Warning Indicators
In the SNPRMs (Notices 4 and 5), NHTSA proposed to require (S5.5.2)
brake system malfunction indicators to be activated by either an
automatic brake indicator check function or a manual check function.
While FMVSS No. 105 currently requires brake indicator lamps to be
activated automatically when the vehicle is started, in Europe the
check function often requires manual action, such as pressing a button
or applying the parking brake.
Advocates and CAS opposed the use of a manual check function to
check brake system integrity in lieu of an automatic check function.
Advocates argued that the existing requirement for all operating
systems to be automatically monitored for the driver when turning the
ignition key has been ``one of the great advances in American
automobile regulation'' and disagrees that the need for safety will be
met by this approach.
After reviewing the available information, NHTSA has decided to
permit the manual check function in the final rule, as an alternative
to the automatic check function. The agency believes that requiring an
automatic check function is not necessary to ensure safety. Moreover,
the agency has granted several petitions for inconsequential
noncompliance from manufacturers that did not provide an automatic
check function. These decisions to grant the petitions are consistent
with the agency's current belief that allowing use of a manual brake
warning indicator, which is consistent with international
harmonization, will not have any corresponding detriment to safety.
BMW recommended that NHTSA modify S5.5.3 which specifies the
duration during which an indicator is activated. BMW claimed that some
ABS warning indicators can only be detected after a certain minimum
wheel speed is achieved. Accordingly, it requested that the antilock
failure indicator only be required to activate when a road speed of 10
km/h is achieved.
While NHTSA agrees with BMW that the wheel must be rotating to
properly check a wheel sensor, the agency believes that it is important
for the check function to be able to be performed while the vehicle is
stationary. Given the current state of technology, NHTSA believes that
the ABS malfunction warning system can be designed to remember if there
had been an ABS sensor failure the last time the vehicle's speed was
over the threshold, even after the ignition has been turned off.
Accordingly, BMW's request is denied.
VW recommended decreasing the minimum lettering height for the
brake warning indicator letters to 2 mm (5/64-inch), claiming that the
proposed 3.2 mm (1/8-inch) height is larger than necessary.
NHTSA has decided to retain the minimum letter height, based on its
concern that some drivers, especially elderly drivers, would not be
able to distinguish letters under 3.2 mm. The agency further notes that
the 1/8'' dimension is the same as the dimension currently specified in
FMVSS No 105.
Kelsey-Hayes commented that, if a separate indicator is used for
ABS failure, rear-only ABS equipped vehicles should use a failure
indicator specifying ``Rear Anti-lock.''
NHTSA believes that it would be inappropriate to require the words
``Rear Anti-Lock'' to distinguish a rear wheel ABS from a four wheel
ABS. The indicator's purpose is to inform the operator that there is a
malfunction with the vehicle's ABS. The driver should be aware, through
the owner's manual and/or information provided at the time of the
vehicle's purchase, whether it is equipped with a four-wheel or rear-
only ABS. However, even though the agency will not require this
information, adding the word ``rear'' to the ABS failure warning is not
prohibited under the standard.
Kelsey-Hayes stated that both red service brake failure warning
indicators ``Brake'' and yellow ``ABS'' malfunction indicators should
be activated simultaneously in the case of a service brake failure in
cars equipped with separate lights.
NHTSA disagrees with Kelsey-Hayes' recommendation for simultaneous
activation of both lights in case of a service brake failure, unless
the service brake failure is one that also disables or impairs the
operation of the ABS. The two lights signal different types of
failures, with different consequences. There can be failures that
affect both systems, in which case both indicators would activate.
However, automatically activating the ABS indicator in case of any
service brake failure would be misleading, and therefore inappropriate.

E. General Test Conditions

1. Ambient Temperature
In S6.1.1 of the 1991 SNPRM, NHTSA proposed that for all tests
specified in S7, the ambient temperature be between 0 deg.C (32 deg.F)
and 40 deg.C (104 deg.F).
Bendix commented that NHTSA should permit the low adhesion tests to
be conducted at temperatures less than 32 deg.F because the ambient
temperature provision requires testers either to wet the test surface
or artificially make ice.
NHTSA notes that the issue of low temperature testing is moot since
Bendix's comment was made with respect to the ABS performance test in
proposed S7.3, which the agency has decided not to adopt in today's
final rule. Even if this test had been adopted, NHTSA notes that it
would be unnecessary to use ice to represent a low PFC. The agency
further notes that no other commenter suggested the need to use ice for
any test.
2. Road Test Surface
In the 1991 SNPRM, NHTSA proposed that the primary stopping
distance tests be performed on a test surface with a PFC of 0.9. This
road test surface specification differed from FMVSS No. 105, the NPRM,
and the 1987 SNPRM, all of which specified a skid number of 81 to
define the road test surface. In response to comments to Notice 4,
NHTSA decided to propose a PFC for the test surface. The agency noted
that PFC is a more relevant surface adhesion measurement for the non-
locked wheel tests required by FMVSS No. 135, since the maximum
deceleration attained in a non-locked wheel stop is directly related to
PFC, but not skid number.
Fiat, Toyota, and GRRF stated that ECE R13 specifies that the test
surface should be ``a road surface affording good adhesion.'' VW
requested that the standard provide the option of specifying either a
skid number or a PFC.
NHTSA, after reviewing its test data and other available
information, continues to believe that a PFC of 0.9 is an appropriate,
objective value for the test surface. ECE R13's specification that the
road surface should afford ``good adhesion'' is unreasonably subjective
and therefore inappropriate for an FMVSS. Such an imprecise test
condition would lead to unreasonable variability, thereby causing test
results that varied based on the road surface and not the vehicle's
actual braking ability. Similarly, it would be inappropriate to allow
the optional use of skid numbers, which would result in unnecessary
variability, since the same

[[Page 6416]]
vehicle might have different test results based on which method was
used to define the test surface. As explained in the 1991 SNPRM (Notice
5), PFC is more relevant than skid number for the non-locked wheel
tests, since the maximum deceleration that can be attained in a non-
locked wheel stop is directly related to PFC, which represents the
maximum friction available.
GM and MVMA requested that the agency adopt a dry road PFC of 1.0,
since compared with a PFC of 0.9, they believe 1.0 more closely
parallels a skid number of 81 specified in FMVSS No. 105. Ford
requested that the test surface be specified at 0.95 PFC. GM stated
that not raising the PFC to 1.0 would require manufacturers to
compensate for the loss of adhesion by equipping vehicles with higher
rolling resistance tires, which would adversely affect the fuel economy
of GM's car fleet by 1.2 mpg. GM further commented that compared with
FMVSS No. 105, a cold effectiveness stopping distance of 70 m on a PFC
of 0.9 would significantly increase the requirement's stringency.
Based on industry-government cooperative testing to evaluate the
effect of fluctuations of PFC on vehicle stopping performance, NHTSA
has determined that a PFC of 0.9 reasonably represents stopping on a
dry surface and will not be a significant source of variability in the
stopping3 distance tests. While this testing focused on heavy
vehicle stopping performance, the agency believes that the test
findings are applicable to passenger cars subject to FMVSS No. 135,
since the tests addressed the road surface coefficients of friction.
Testing indicates that the expected minor variability of a high
coefficient of friction surface appears to have a negligible impact on
vehicle stopping distance performance. Variation of the average
stopping distances for the six different surfaces was small, with the
deviation from the average being only 5 feet. Accordingly, the agency
believes that any variability in the stopping performance on a high
coefficient of friction surface is more likely due to variation in the
vehicle's performance rather than test surface variability.

\3\``MVMA/NHTSA/SAE Round Robin Brake Test,'' Transportation
Research Center of Ohio, Report No. 091194, August 26, 1991.
---------------------------------------------------------------------------

NHTSA has decided that a test road surface specification of PFC 1.0
would result in practicability problems for the agency. It would have
to conduct compliance testing on a surface with a PFC higher than 1.0.
Such a surface is difficult to find. The agency also notes that GM
conducted an extensive survey of actual road surfaces, which indicated
that a PFC of 0.9 is fairly typical.
As explained in detail in NHTSA's decision to require heavy
vehicles to be equipped with antilock brake systems, using PFC values
to express test surfaces is appropriate even though these values may
indicate some fluctuation. Given this fluctuation, the agency has
considered whether the fluctuation significantly affects the
requirement's objectivity. In an earlier rulemaking about FMVSS No.
208, Occupant Crash Protection, the agency explained that since some
variability in any test procedure is inherent, the agency need only be
concerned about preventing ``unreasonable'' or ``excessive''
variability to avoid causing manufacturers to ``overdesign'' vehicles
to exceed the minimum levels of protection specified by the Federal
safety standards. (49 FR 20465, May 14, 1984; 49 FR 28962, July 17,
1984.) With respect to the tests in FMVSS No. 135, variability of the
PFC value of the test surface will have a negligible impact on a
vehicle's ability to comply with the requirements.
Ford stated that it would be impossible to build a track to exactly
a PFC of 0.9, given PFC variability, test tire variability, and
changing track surfaces due to aging and weathering.
In evaluating the requirement's practicability, NHTSA has
considered possible difficulties with respect to building and
maintaining test surfaces with a PFC of 0.9 for the high coefficient
stopping tests. (Those interested in building and maintaining a test
surface should refer to NHTSA's ``Manual for the Construction and
Maintenance of Skid Surfaces,'' (DOT HS 800 814.) Variations in PFC for
high coefficient of friction surfaces do not affect stopping distance
test results appreciably. After reviewing the comments and available
information, NHTSA has concluded that specified test surfaces can be
achieved and maintained. As explained above, recent ``Round Robin''
testing related to research about heavy vehicle braking by the agency
and others on several test tracks indicates that the test surface
specification does not raise practicability or objectivity concerns.
MVMA, GM, and Ford recommended use of a correction factor for
stopping distance to account for testing on surfaces with PFCs that
differed from those prescribed in the standard. They stated that a
manufacturer is fortunate if the tests they conduct are actually
carried out on surfaces with the precise PFC as specified in the
harmonized standard.
NHTSA believes that it would be inappropriate to specify a stopping
distance correction factor, as requested by the comments. The agency
notes that the same variables that will apply to manufacturer testing
in accordance with FMVSS No. 135 also applied to their testing under
FMVSS No. 105, and no correction factor was established or needed at
the time. NHTSA further notes that a manufacturer may test its vehicles
on whatever surface it likes, and may make any corrections it chooses.
The FMVSS specifies requirements with which manufacturers must certify
that their vehicles comply on a given surface under specified test
conditions. Moreover, the agency will follow the procedures specified
in the FMVSS for purposes of compliance testing. If a manufacturer is
confident that its testing on a different surface will yield results
comparable to agency test results under FMVSS No. 135 (by applying a
correction factor), it need not exactly follow every agency
specification.
Advocates opposed the proposal to replace skid numbers with PFC. It
claimed that PFC numbers cannot be correlated to skid numbers because
they do not describe the same event. Advocates further commented that
most state highway authorities use skid numbers to evaluate a roadway's
skid resistance, and that NHTSA would make it impossible for data
comparison by encouraging different authorities to use different
measurement standards. In contrast, Fiat, Ford, ITT-Teves, GRRF, OICA,
Mercedes, and MVMA stated that using PFC rather than skid numbers will
lead to more repeatable road surface adhesion measurements and that PFC
directly correlates to vehicle stopping distance.
PFC and skid number can both be measured simultaneously during
traction tests. However, the two road surface specifications are used
for different purposes. Highway officials use skid numbers to determine
when to resurface a road, not to determine test vehicle performance in
stopping tests. The agency notes that because FMVSS No. 135 evaluates a
vehicle's capability during braking to use the available friction
capability at the interface between the tire and road, PFC is the more
appropriate measure for that purpose. It is not necessary to establish
a correlation between the two numbers, for any given surface.
While ITT-Teves, MVMA, and Ford agreed with the proposed use of the
ASTM test tire and test procedure, the GRRF, VW, Mercedes Benz, Fiat,
and OICA, stated that the ASTM test methods for determining PFC are not

[[Page 6417]]
familiar in Europe. They requested NHTSA to consider other methods of
determining adhesion or PFC, but suggested no specific test method or
procedure.
NHTSA is aware that the ASTM trailer and test method are not widely
used outside of the United States. However, any method of determining
PFC specified in the standard must be objective and repeatable. Those
commenters that requested consideration of other methods did not
provide any evidence that there are other standardized methods in
existence that are as objective, repeatable, and universally accepted
as the ASTM method that has been specified.
NHTSA also notes that the concerns expressed by several European
entities about compliance need not adversely affect them, since the
agency does not insist that any manufacturer use a specific test method
or procedure. Rather, the individual manufacturer must determine
whether to test exactly to the specifications of FMVSS No. 135 or to
use its own methods of determining that its braking systems will meet
the requirements of the standard. NHTSA, as stated earlier, will use
the procedures established in FMVSS No. 135 in its own testing. The
agency has decided to specify the ASTM test procedure for all of its
compliance tests. The agency emphasizes that GRRF's suggested wording
(i.e., ``a surface affording good adhesion'') would be inappropriate
for a Federal safety standard since it is not objective. The two
specifications are not in conflict with each other, however. Because
NHTSA's goal is to define ``good adhesion'' objectively, the agency has
decided to specify a surface measured with a standard test method to a
specific adhesion level.
Honda recommended that the test condition state ``PFC shall be
situated between the slip ratio of 10 to 30 percent and the friction
coefficient of the road surface.'' It stated that this slip ratio was
appropriate because most roads are within this range. It stated that
slip ratios can vary even if PFC value remains constant.
NHTSA believes that slip ratios are not appropriate for defining a
pavement surface to be used for stopping distance tests, because the
minimum stopping distance is obtained at the maximum traction value,
which is defined directly by the PFC. The agency believes that it is
most important to provide a surface with the available traction defined
so that all vehicles have an equal chance for achieving the shortest
stop, regardless of the optimum vehicle slip ratio for each vehicle.
For a given PFC, the vehicle slip ratio at which maximum traction is
achieved varies depending on the vehicle characteristics. Accordingly,
slip ratio cannot be used to define a test surface, because it is
vehicle- dependent.
3. Instrumentation
In the 1991 SNPRM (Notice 5), NHTSA specified in S6.4, the
instrumentation to measure brake temperature, brake line pressure, and
brake torque.
The GRRF, Ford, Fiat, and VW recommended that NHTSA allow
alternative methods to measure brake temperature. Ford stated that plug
type thermocouples develop problems as brake pad wear occurs and that
use of rubbing-type thermocouples would reduce cost and time.
NHTSA notes that a standard must include a specific method to
ensure objectivity, so that the requirements are the same for all
vehicles. In addition, a specific method ensures uniformity and thus
facilitates compliance testing. The specification of plug-type
thermocouples is the same as specified in Society of Automotive
Engineers' (SAE) Recommended Practices and is identical to that
specified in FMVSS No. 105, FMVSS No. 121, and FMVSS No. 122. The
agency is not aware of any problems resulting from use of this
procedure. NHTSA further notes that while the agency will use plug type
thermocouples specified in S6.4.1 for its own testing, a manufacturer
may use whatever type of brake temperature measuring device it prefers
for its own testing. Nevertheless, NHTSA does not recommend using
rubbing-type thermocouples in FMVSS No. 135, based on agency testing
that indicates that the two types of thermocouples give different
readings for brake temperature.
Bendix recommended that NHTSA specify whether brake linings can be
heated up to an initial brake temperature (IBT) before the static
parking brake test and that a procedure be specified. The procedure
would be important for vehicles with parking systems not utilizing the
service friction elements.
NHTSA notes that IBT as defined in S4, and S6.5.6, describes the
procedure for establishing IBT, and S7.12.2(a) sets the maximum IBT (no
minimum) for the parking brake test regardless of the type of friction
elements. The non-service brake friction materials should not be heated
because under normal driving circumstances they are never used (heated
up) until the parking brake is applied after the vehicle stops. This is
not necessarily the case with service brake friction materials.
Therefore, it would be unrealistic to describe a heating procedure.
However, the agency has decided to revise section S7.12.2(a) as
follows to clarify the requirements on IBT for both service and non-
service parking brake friction materials. Specifically, the revised
language makes clear that IBT applies to both service and parking brake
friction materials.
``7.12.2(a) IBT.
(1) Parking brake systems utilizing service brake friction
materials shall be tested with the IBT 100 deg.C
(212 deg.F) and shall have no additional burnishing or artificial
heating prior to the start of the parking brake test.
(2) Parking brake systems utilizing non-service brake friction
materials shall be tested with the friction materials at ambient
temperature at the start of the test. The friction materials shall have
no additional burnishing or artificial heating prior to or during the
parking brake test.''

F. Road Test Procedures and Performance Requirements

1. Permissible Wheel Lockup
In the 1991 SNPRM (Notice 5), NHTSA proposed to allow wheel lockup
of 0.1 seconds or less of any wheel during several road tests. This
differed from earlier proposals that prohibited any type of lockup. The
agency concluded that, due to pavement irregularities, it would be
extremely difficult for a test driver to achieve maximum deceleration
without causing momentary lockup of one or more wheels. The agency
believed that the brief lockup time permitted would not result in
vehicle instability, especially considering that, even ABS controlled
brakes occasionally undergo nominal, self-correcting lockup conditions
for very short periods of time.
Advocates and CAS opposed permitting any lockup, stating that it
may result in vehicle instability. Advocates believed that allowing
momentary lockup would result in the sale of more rear-biased vehicles
that are susceptible to skidding. Bendix recommended a revised wheel
lock criteria to increase the permitted lockup time, stating that it
would take longer than 0.1 seconds for a driver to detect and react to
wheel lock up. It believed that this would lead to less aggressive
driver performance in testing to FMVSS No. 135 specifications, as
drivers tried to avoid any type of lockup.
NHTSA has decided to permit a minimal amount of wheel lock up to
facilitate vehicle testing. The agency believes that it will not be
detrimental to safety as alleged by Advocates.

[[Page 6418]]
Allowing momentary wheel lockup during compliance testing will not
affect a vehicle's real world ability to lock or not lock its wheels.
Rather, this provision merely acknowledges that momentary lockup may
inadvertently occur during compliance testing due to road surface
irregularities, as test drivers attempt to achieve the shortest stops
possible. Therefore, this provision ensures that entire test runs are
not invalidated due to such an occasional occurrence.
NHTSA also notes that while Advocates claimed that the proposal to
permit momentary lockup during stops represents ``a significant
modification of the current FMVSS No. 105 test procedure'' whose real-
world safety implications are unknown, FMVSS No. 105 in fact generally
permits lockup of one wheel during stopping distance tests. The
provision being adopted today thus represents a more stringent test
condition, not a less stringent one.
In response to Bendix's comment, the momentary lockup is not a
situation that a driver is supposed to detect and respond to; it is
simply an allowance for a minor, inadvertent occurrence during testing.
Therefore, Bendix's request to permit a longer lockup period is not
necessary or appropriate.
Honda and Ford recommended that S7.2.1(f) be changed to define
wheel lock as an angular velocity of zero, rather than the current
definition of 10 percent of vehicle speed. They reasoned that it would
be difficult to read the definite value with a 10 percent margin,
because speed recorded on the data sheet changes gradually and the data
also includes vehicles vibration.
The wording proposed for S7.2.1(f) was not intended to redefine
wheel lockup as 10 percent of vehicle speed (90 percent wheel slip).
Rather, it was intended to provide a practical criterion for making a
determination that wheel lockup (100 percent wheel slip) exists, given
the limitations of current instrumentation and recording devices. The
proposal was based on the agency's experience at the Vehicle Research &
Test Center (VRTC). Much of the vehicle testing that NHTSA has relied
on to formulate FMVSS No. 135 was conducted at VRTC. This testing
indicated that, with the instrumentation used by VRTC, it would be
difficult to accurately measure zero angular velocity, due to spurious
``signal noise''. Thus, it would be extremely difficult to ascertain
when a wheel reached an angular velocity of zero.
The comments expressed by Ford and Honda indicate that they have
experienced similar problems with ``signal noise'' due to vibration and
``drift'' of the signal when reading the vehicle speed trace, which
make it more difficult to relate the wheel rotational speed measurement
to that variable than to read its absolute value. The difference
between the agency's experience and that of Ford and Honda is probably
due to differences in the instrumentation packages used.
After further reviewing this issue, NHTSA has decided to remove the
proposed S7.2.1(f) entirely, because it was probably biased toward a
particular type of instrumentation, and the agency does not want to
impose unnecessary restrictions on what instrumentation is used to test
for compliance with the standard. In order to clarify the meaning of
wheel lockup, a definition stating that wheel lockup means 100 percent
wheel slip has been added to S4. This definition is the same as has
recently been added to both FMVSS No. 105, Hydraulic Brake Systems, and
FMVSS No. 121, Air Brake Systems.
As a practical matter, NHTSA notes that there is essentially no
difference between the method proposed in Notice 5 and that recommended
by Ford and Honda. Once a wheel reaches 90 percent slip, complete
lockup will be essentially instantaneous. As clarified in this final
rule, there is no question of what is meant by wheel lockup. How that
is measured is left to individual testing organizations, as is true for
other aspects of standard.
2. Road Test Sequence
In the 1991 SNPRM (Notice 5), NHTSA proposed the following road
test sequence: Burnish and wheel lock sequence at gross vehicle weight
rating (GVWR); wheel lock sequence, ABS performance, and the torque
wheel test at lightly loaded vehicle weight (LLVW); the torque wheel,
cold effectiveness, high speed effectiveness, stops with engine off at
GVWR; cold effectiveness, high speed effectiveness, failed ABS, failed
proportional valve, and hydraulic circuit failure at LLVW; and
hydraulic circuit failure, failed ABS, failed proportional valve, power
brake unit failure, the static and dynamic parking brake tests, heating
snubs, hot performance, brake cooling, recovery performance, and final
inspection at GVWR.
JAMA and GRRF supported the proposed road test sequence, even
though R13H does not specify a test sequence. GM recommended modifying
the test sequence by eliminating two of the four ballast changes (i.e.,
reduce the times needed to switch between lightly loaded and fully
loaded). It also recommended not including the full ABS test and the
dynamic parking brake test.
As explained below, NHTSA has decided not to include the full ABS
test and the dynamic parking brake test. Nevertheless, the agency
believes that it would be inappropriate to change the test sequence for
the sake of reducing the test preparation effort. The agency emphasizes
that the test sequence being adopted specifies that the GVW and LLVW
wheel lock sequence tests be conducted first, since their results
determine whether the torque wheel test needs to be conducted. The
agency further notes that the test sequence being adopted permits
removal of the torque wheels as soon as that test is completed. This is
important since the torque wheels might get wet or otherwise adversely
affected if they were not removed. Based on these considerations, the
agency has determined that it would be inappropriate to switch the test
sequence, which would result in fewer ballast changes.
3. Pre-Burnish
FMVSS No. 105 specifies a pre-burnish requirement to evaluate
brakes in the brand new condition. In the initial NPRM (Notice 1),
NHTSA proposed a similar requirement for the harmonized standard.
However, in the 1987 SNPRM (Notice 4), the agency explained that it no
longer believed a pre-burnish test was necessary for safety, given the
relatively short period of time that the vehicle's brakes remain in the
pre-burnished condition.
In comments to both SNPRMs, Advocates and CAS strongly opposed
deleting this test. They stated that it takes hundreds of miles of use
before brakes are properly burnished, especially for vehicles used in
rural areas, in which long distances may be traveled with few brake
applications. Advocates stated that certain brakes, most particularly
disc-type brakes, are highly resistant to burnishing. That organization
argued that the agency acknowledged this high mileage need for proper
burnishing in the 1985 NPRM, but attempted to rationalize this
concession in the first SNPRM. It also argued that stopping distance
performance may be considerably greater before burnish than afterwards.
Advocates stated that deleting a pre-burnish test would allow
manufacturers to produce and sell cars whose pre- burnish, on-the-road
braking capability is unknown. It stated that it does not believe this
is in the best interests of traffic safety, and that it does not
believe the agency can allow cars to be sold and used that have no
regulatory control

[[Page 6419]]
over their stopping distances before burnishing takes place.
NHTSA is not persuaded by the comments from CAS and Advocates
regarding the need for a pre-burnish test, and has decided to not
include this test in the final rule. The arguments by CAS and Advocates
are essentially the same as those made in response to the 1987 SNPRM
(Notice 4). These comments were already addressed in the preamble to
the 1991 SNPRM (Notice 5, 56 FR 30533).
Advocates has made an unsupported statement that disc brakes are
highly resistant to burnishing. No test data or other evidence was
supplied to support this allegation. Regardless, the pertinent question
is not how long or how many miles it takes to burnish brakes in use,
but whether there is a big enough difference in performance before and
after the 200-stop burnish specified in the standard to present a
safety problem. If some types of brakes do take a long time to become
fully burnished, then they would not be fully burnished after the 200-
stop burnish sequence specified in the standard, so they would have to
meet the cold effectiveness stopping requirements in a partially-
burnished state. If that were the case, their eventual, fully-burnished
performance would be even better than that required by the standard.
Advocates also argued that stopping distances before burnish may be
considerably longer than after burnish. This statement was also
unsupported by any test data. Agency testing conducted during the
development of this standard (Harmonization of Braking Regulations--
Report No. 1, Evaluation of First Proposed Test Procedure for Passenger
Cars, Volume 1, May, 1983, DOT HS 806-452) showed that in some cases
stopping distances were somewhat shorter after burnish, and in other
cases stopping distances were shorter in the unburnished state.
However, the overall conclusion was that the burnish had a small effect
on stopping distances. Also, this research was done using the burnish
procedure specified in FMVSS No. 105, which is more severe than that
specified in FMVSS No. 135, and would therefore have a greater effect
on braking performance.
4. Burnish
Burnish procedures serve as a conditioning to permit the braking
system to achieve its full capability. In the 1987 SNPRM (Notice 4),
NHTSA proposed specifying 200 burnish stops. The agency stated that the
burnish procedures would stabilize brake performance and reduce vehicle
and test variability. In the 1991 SNPRM (Notice 5), the agency proposed
almost the same requirements as the earlier SNPRM. The only substantive
change from the earlier notice entailed specifying that the pedal force
would be adjusted as necessary to maintain the specified constant
deceleration rate.
Kelsey-Hayes and Honda recommended that the burnish procedures be
made consistent with the ones in FMVSS No. 105, with respect to the
number of burnishes, the test speed, and the deceleration rate.
Specifically, both commenters recommended that the test speed be 65 km/
h (40.4 mph) and the deceleration rate to be 3.5 m/s (11.5 fps). While
these conditions enabled Kelsey-Hayes to conduct the FMVSS No. 105
burnish on a secluded public road, the proposed burnish requirements
for FMVSS No. 135 would have to be conducted at a commercial test
facility, which may not be readily available. Honda stated that the
cost of the proposed FMVSS No. 135 burnish test was more than the cost
of the FMVSS No. 105 burnish, even though the brake temperatures at the
end of the respective burnish procedures are the same. JAMA and Toyota
recommended that the test speed be reduced from 80 km/h to 70 km/h
because the brake temperature would increase too much under the
proposed burnish speed.
NHTSA has decided to adopt the burnish procedure as proposed in the
1987 and 1991 SNPRMs. As explained in those notices, the agency
purposely changed the burnish procedure from the one in FMVSS No. 105
to provide a more realistic burnish. NHTSA believes that the new
burnish procedure will more closely match real world situations,
including the actual type of burnish most drivers will achieve in the
course of normal driving. The burnish procedure in the harmonized
standard will better reflect the real world capabilities of the brakes
in a passenger car. The new burnish procedure itself will not affect
the time or mileage needed to burnish brakes for the average driver.
NHTSA believes that the burnish procedures adopted by today's final
rule represent an efficient burnish procedure that is consistent with
R13 and the ECE harmonized version of R13H.
NHTSA is not able to determine the meaning of JAMA's comment that
the temperature ``would increase too much'' under the specified burnish
procedure. As previously stated, the agency believes that the specified
burnish is more representative of actual driving experience. Therefore,
any temperature increase during burnish would also be experienced on
the road.
Advocates and CAS stated that the burnish procedure proposed for
FMVSS No. 135 would not ensure that cars are tested with properly
burnished brakes. They stated that decreasing the deceleration rate,
lowering the initial brake temperature, and introducing a variable
pedal force would extend the time and mileage needed to complete a full
burnish. Advocates further believed the proposed burnish procedure
would not evaluate how well the brake system reacts to higher
temperatures, along with the resulting potential for fade during the
initial burnishing.
NHTSA believes that Advocates and CAS misunderstand a fundamental
principle of brake burnish procedures: a less severe burnish results in
a more severe test. The burnish procedure has no bearing whatsoever on
how long it will take a vehicle to achieve full performance in actual
use. More specifically, the agency notes that the changes proposed in
the 1987 SNPRM (Notice 4) about the burnish procedure (e.g., lower
initial brake temperature, lower deceleration rate) would be more
similar to typical driving than those in FMVSS No. 105. Moreover, NHTSA
believes that most vehicles will not be driven for long periods of time
in a significantly less burnished condition than that obtained from the
burnish procedures being adopted.
Advocates also said that it does not agree with NHTSA'S claim that
drivers rarely exceed a deceleration rate of 3.0 m/s(2) except in
emergencies. Advocates claimed that typical stop-and-go braking
deceleration rates, especially in congested urban expressway traffic
with high speed differentials, can exceed this rate. NHTSA acknowledges
that deceleration rates can exceed 3.0 m/s(2), but burnish is meant to
simulate typical use, not these unusual circumstances.
MVMA, Ford, Chrysler, and GM requested a modification of initial
brake temperature from 4 CAS appears to misunderstand the agency's
regulatory framework, since a vehicle either passes or fails a
requirement in a FMVSS; there is no provision for a marginal pass. For
instance, a vehicle that ``marginally passes'' FMVSS No. 105 still
complies with the standard. Therefore, the agency believes CAS's
argument is not relevant to the regulatory framework set forth by
statute and incorporated in the Federal motor vehicle safety standards.
The agency further notes that the Hunter test apparatus is a simplified
version of the road transducer pad that the NHTSA in light of comments
by the industry considered prior to selecting torque wheels as the most
acceptable method of measuring adhesion utilization. Therefore, the
agency believes that it would be inappropriate to require this method
of evaluating compliance.

\4\A heavily front biased vehicle will skid but remain stable
heading forward, since the front wheels will lock first. In
contrast, a rear biased vehicle will spin out, since the rear wheels
will lock first and those wheels would tend to lead.
Advocates stated that the real-world effects of the adhesion
utilization test are uncertain and that NHTSA has not demonstrated a
connection between real-world situations and the wheel lock sequence
results. Advocates further commented that there is more to braking
stability than front-axle bias and that plow-out skids will result in
lane departures and stopping distances that are too long for safety
purposes, even for vehicles with front axle bias and ABS.
Advocates further stated that

Real-world crash results for cars tested under the two-part Adhesion
Utilization protocol may not be favorable for significant numbers of
production cars. The truncation of the testing protocol that has
accompanied the proposed two-stage system of the current SNPRM
comprising the Wheel-lock Sequence and Torque Wheel (especially due
to adoption of the 90% efficiency rationale) creates a ``window'' of
allowable production variability that can permit a significant, but
unquantifiable, percentage of assembly-line vehicles to be rear-
brake biased. Under certain operating conditions, especially those
uncontrolled by the reduced performance specifications of the
current proposed rule, such as the elimination of a low-coefficient
surface test, many cars may experience serious instability under
severe braking. The plain fact is that even if both parts of the
two-stage test as proposed are used for a given car model, this
still will not ensure that all cars will have appropriate front-
brake bias and does not foreswear the potential for an unknown
number of production units to be susceptible of serious spin-out
crashes in panic braking situations. Despite advocating the two-
stage test in this SNPRM, the agency itself obviously still harbors
doubts over its adequacy to detect cars with rear-brake bias.

Advocates has expressed two concerns. Their first concern is that,
by having a simple wheel lock sequence test, manufacturers would
produce cars that have too much front axle bias in their brake systems,
because such a vehicle would always pass the wheel lock sequence test.
The extreme example of this would be a car with no brakes at all on the
rear wheels. Such a vehicle would always be dynamically stable, but if
braked to the point of wheel lockup would provide no ability to steer.
This concern by Advocates ignores the adhesion utilization requirement
is only one of many requirements in the standard, and therefore is not
the sole factor in determining brake system design. If a manufacturer
were to produce a car with too much front bias, it would compromise the
vehicle's ability to satisfy other requirements of the standard, such
as service brake stopping distances, partial failure, failed power
assist, and parking brake requirements.
Advocates' second concern is that, because of the 10% allowance for
test variability, a vehicle could pass the torque wheel test and still
be rear-biased, and therefore ``susceptible of serious spin-out
crashes.'' While it is theoretically possible for a vehicle to be
slightly rear-biased and still pass the torque wheel test, NHTSA
believes this

[[Page 6421]]
possibility is extremely remote. If a manufacturer were to design a
vehicle to exhibit slight rear bias, production and test variability
would create too great a risk that the vehicle would not comply with
either the wheel lock sequence test or the torque wheel test. Rather,
the 10% allowance is meant to allow cars to be designed with brake
balance that is still front-biased, but closer to ideal than could be
achieved if the manufacturer had to worry about a failure of the torque
wheel test due to test variability. Also, for a vehicle to exhibit a
tendency to spin out, it must experience a condition where the rear
wheels are locked and the front wheels are not. Any vehicle falling in
the 10% ``window'' would be so close to ideally balanced that the point
of wheel lockup would be essentially simultaneous for both axles, and a
condition of rear axle lockup without front axle lockup would be almost
impossible to maintain.
b. Wheel Lock Sequence Test. NHTSA explained its tentative
determination in the SNPRM (Notice 5) that the wheel lock sequence test
would identify those vehicles that are heavily front biased. Such
vehicles have good stability characteristics because their front brakes
always lock first during braking, regardless of test surface.
Accordingly, a heavily front biased vehicle would not need to be
subject to the torque wheel test, since it would be considered to have
inherently good stability characteristics. Under the proposal, a
vehicle would need to meet the wheel lock sequence test requirements on
all test surfaces that would result in a braking ratio of between 0.15
and 0.80, inclusive, at each of two vehicle loading conditions: GVWR
and LLVW.\5\ The wheel lock sequence test would require a brake
application at a linear, increasing rate such that lockup of the first
axle is achieved between 0.5 and 1.0 second.

\5\This is defined in Section S4 as the unloaded vehicle weight
plus the weight of a mass of 180 kg, including driver and
instrumentation.
---------------------------------------------------------------------------

GRRF agreed to the proposed wheel lock sequence test and planned to
add it to R13 and R13H. Ford and Chrysler stated that there were
insufficient data to establish whether the wheel lock sequence test
could be consistently repeated. Ford believed that there is potential
for discrepancies between manufacturer testing and NHTSA testing.
NHTSA believes that Ford and Chrysler are incorrect in their
assessment of the wheel lock sequence test. The agency notes that the
available test data indicate that the wheel lock sequence test is
objective and can be consistently repeated.\6\ As explained above, the
wheel lock sequence test is the first part of the adhesion utilization
test procedure, and evaluates whether there is sufficient front axle
bias to ensure stability in a lock up situation. If a car has
insufficient front axle bias to consistently meet the wheel lock
sequence test, it does not automatically fail to comply with FMVSS No.
135. Rather, it would be tested under the torque wheel method. If the
vehicle passes the torque wheel test, the wheel lock sequence test
results are irrelevant.

\6\``Harmonization of Braking Regulations, Report Number 7,
Testing to Evaluate Wheel Lock Sequence and Torque Transducer
Procedures,'' DOT HS 807611, February 1990.
---------------------------------------------------------------------------

NHTSA expects that 90 to 95 percent of cars will pass the wheel
lock sequence test, meaning only 5 to 10 percent of the cars will have
to be tested with the torque wheel method. This will reduce potential
testing expenses by a greater amount than the agency could have
foreseen at the time it published the 1991 SNPRM.\7\

\7\When the 1991 SNPRM was published, the percentage of cars
that may have been required to be torque wheel tested was already
small, given that the agency expected that 95 percent of all cars
would pass the wheel lock sequence test. Thus, only five percent of
all cars were expected to be torque wheel tested. As a result of the
increased use of antilock brake systems that do not need to be
torque wheel tested, the agency anticipates that in model year 1999,
the number of cars that might need torque wheel testing will be less
than one percent.
---------------------------------------------------------------------------

Ford requested that the agency specify a braking ratio of 0.15 to
0.70 instead of the proposed ratio of 0.15 to 0.80. It believed that
this change would help avoid degradation and flat spotting of tires,
since under its recommended ratios only wet surfaces would be required.
NHTSA has determined that it would be inappropriate to lower the
upper limit in the braking ratios. If Ford's recommendation were
adopted, there would be no assurance of stability on typical dry road
surfaces. Therefore, the agency has decided to require the wheel lock
sequence test be performed at any ratio between 0.15 to 0.80.
More generally, NHTSA has considered whether the range of possible
test surfaces for the wheel lock sequence test raises practicability
concerns. The agency notes that a manufacturer will not need to test a
vehicle on every possible surface but could instead make predictions
based on testing at several points and brake design characteristics.
Moreover, instead of using the wheel lock sequence test to screen out
vehicles, a manufacturer could conduct only the torque wheel tests,
which do not involve a wide range of test surfaces, if a manufacturer
doubted that its vehicle could pass the wheel lock sequence test on all
applicable test surfaces. Given the availability of the torque wheel
test, NHTSA believes that there are no practicability concerns
presented by the wide range of test surfaces in the wheel lock sequence
test.
Bendix requested that NHTSA clarify whether the definition of wheel
lock in S7.2.1(f) is applicable to all testing situations or just those
in S7.2. After reviewing this comment, NHTSA has modified the
description of wheel lock in S7.2.1(f) to clarify that it only applies
for purposes of the adhesion utilization test.
MVMA and Ford noted that the proposed wheel lock sequence test
permits wheel lockups of ``less than 0.1 second;'' however, the balance
of the SNPRM permits lockup ``for not longer than 0.1 second.'' The
agency has decided to standardize this factor so all references to
wheel lockup will read -'' 0.1 second.''
MVMA, Chrysler, Ford, Toyota, and the Japanese Automobile
Manufacturers Association (JAMA) commented that it would be difficult
to comply with the proposed test condition for lockup to be achieved
between 0.5 and 1.0 seconds after initial brake application. Several
commenters suggested an upper limit of 1.5 seconds, which they believed
would still preclude spike stops. Ford suggested that the requirement
specify no maximum time, provided the vehicle's speed was greater than
15 kilometers per second (km/s) at the time lock up occurred.
After reviewing the available information including agency testing,
NHTSA has determined that it is appropriate to raise the ceiling to 1.5
seconds. The agency has decided not to remove the ceiling altogether,
given the need to have a specification that is independent of the
actual pedal force rate since the pedal force rate required to achieve
lock up within a specified time will vary among vehicles.
Suzuki, Toyota, and JAMA recommended that S7.2.3(c)(3) be amended
to allow braking force to be terminated 0.1 seconds after the first
axle locks or when the front axle locks. Suzuki stated that there is no
need to require continued braking beyond the first axle lock, since the
test is designed to determine which axle locks first. Toyota and JAMA
stated that if the rear axle locks first, then the pedal must be
immediately released to prevent accidents.
After reviewing the comments, NHTSA has decided to modify
S7.2.3(c)(3) to state the following: ``The pedal is released when the
second axle

[[Page 6422]]
locks, or when the pedal force reaches 1000 N (225 lbs), or 0.1 seconds
after first axle lockup, whichever occurs first.'' This modification of
the language should avoid the problems cited by the commenters.
BMW requested that the wheel lock sequence test be run at speeds of
50 km/h, claiming that the conditions proposed in the 1991 SNPRM demand
a higher initial speed and brake pedal application rate than the OICA
proposal. NHTSA believes that the proposed test speed of 65 km is
appropriate for safety and consistent with ECE R13H. BMW neither raised
a safety concern nor provided any documentation to support its request
to lower the test speed. Accordingly, the test speed for the wheel lock
sequence test is adopted as proposed.
Ford, Chrysler, and MVMA requested deleting the speed channel
filtering test condition or clarifying it so that it applies only to
analog instrumentation methods. They stated that a low pass filter,
having a low cut-off frequency is applicable to analog data recording
but not digital data recording.
NHTSA has decided to clarify S7.2.3(g) and (h) so that it refers
only to analog instrumentation. These sections address the automatic
recording of data and speed channel filtration and are unnecessary for
digital data recording.
In the 1991 SNPRM (Notice 5), NHTSA proposed a modified wheel lock
sequence test for a vehicle equipped with an antilock brake system on
one or both axles. Under this proposal, an ABS equipped vehicle would
have to be capable of stopping on a surface with a transition from a
high PFC to a low PFC without wheel lockup exceeding 0.1 seconds, after
decelerating in a hard braking from 100 km/g to a stop. The agency
believed that this would test the ABS's ability to compensate for
changes in surface quality and conditions encountered in everyday
driving. The agency requested comment about the need to adopt other
aspects of Annex 13 addressing braking efficiency and split coefficient
of friction surfaces, as more advanced ABS are sold in the United
States.
MVMA and Ford requested that vehicles with axles not directly
controlled by ABS be allowed to be certified as complying with the
wheel lock sequence test. They incorrectly stated that while the 1991
SNPRM only applied the wheel lock sequence test to non-ABS vehicles, a
vehicle with rear wheel only ABS should also be permitted to
demonstrate brake balance by the wheel lock sequence test. They stated
that the use of the wheel lock sequence test is unrelated to whether
the vehicle is equipped with ABS and should be allowed for either
design as an alternative to the torque wheel test.
After reviewing the comments, NHTSA has decided that only vehicles
without any ABS should be required to run the wheel lock sequence test.
The agency notes that differentiating between all-wheel and rear-wheel
ABS as it relates to brake balance is not appropriate since in either
case rear wheel lockup will not occur if the ABS is operational.
c. Torque Wheel Test. Under the 1991 SNPRM (Notice 5), a vehicle
that failed any single test run of the wheel lock sequence test would
be subjected to the torque wheel\8\ test to directly measure braking
forces under a wide range of deceleration conditions and provide data
needed to generate detailed adhesion utilization calculations. Under
the proposal, to pass the torque wheel test, a vehicle would need to
demonstrate that the plots of its adhesion utilization performance fell
within a specified range. Section S7.4.3 sets forth the test conditions
for the torque wheel procedure, including initial brake temperature,
test speed, pedal force, cooling, number of test runs, test surface,
and the data to be recorded.

\8\Torque wheels are strain gauge instrumented devices that fit
between the brake rotor or drum and the wheel assembly, and which
directly measure the reaction torque that is developed by the
friction between the tire and road surface during braking.
---------------------------------------------------------------------------

NHTSA tentatively concluded that the torque wheel test represented
an objective and repeatable method for gathering data for the
construction of adhesion utilization curves. The agency noted that the
torque wheel procedure requires more expensive test equipment and more
time to administer than the wheel lock sequence test.
After reviewing the available information, NHTSA has decided to
modify the section on torque wheel testing in S7.4 to exclude from
testing any car equipped with ABS. The agency has determined that
adhesion utilization testing is only relevant for brake balance in the
event of lock up, which will either not occur, or occur for negligible
amounts of time, on wheels controlled by ABS. Assuming the ABS is
operating, this is true for vehicles in which all wheels are directly
controlled by ABS, or on rear wheel-only ABS vehicles. In rear wheel-
only ABS vehicles, the front wheels would always lock before the rear
wheels, which would not lock at all, or lock for negligible amounts of
time. Accordingly, the number of cars that will have to undergo
adhesion utilization testing will drop to a small percentage of the
overall fleet as ABS becomes more prevalent over the next few years.\9\

\9\The agency estimates that by model year 1999, when FMVSS No.
135 will come into full force, approximately 85-90 percent of
passenger cars will be ABS-equipped.
---------------------------------------------------------------------------

GM, Ford, MVMA, and Chrysler requested that S7.4.3 be changed to
require stops from 50 km/h at both GVWR and LLVW, in addition to the
proposal for stops from 100 km/h. They stated that the additional test
runs would increase the database's statistical accuracy and provide
stopping data at the speed at which the wheel lock sequence test is
conducted. They state that specifying an additional test speed will
reduce the standard error in the estimate by 30 percent. In addition,
GM stated that by specifying two test speeds, a manufacturer would no
longer be able to design speed sensitive brake systems specifically
designed to handle stops from 100 km/h. Similarly, Ford commented that
alternating between the test speeds would avoid speed conditioning of
the brakes.
After reviewing the comments and other available information, NHTSA
has decided to modify S7.4.3 to require five stops from 100 km/h, and
five stops from 50 km/h, at each of the test weights, LLVW and GVW, for
a total of 20 stops. The agency agrees with the commenters that stops
from both speeds will prevent speed conditioning and ensure that
manufacturers design brakes that will be effective over a wide range of
initial speeds. NHTSA has decided to increase the maximum pedal force
rate to 200 N/second (45.0 lbs./sec.) for the stops from 50 km/h in
order to achieve sufficient deceleration levels.
Ford stated that the paired torque and force values generated for
S7.4.4 may not be uniformly distributed when plotted against each
other, a situation that may affect the overall outcome. Ford stated
that data point distribution will not be uniform if the pedal force and
the vehicle deceleration are not changing linearly. It recommended
using a linear regression analysis after dividing the input force into
several increments and averaging all data points within the respective
increments to yield a single average value for that increment.
NHTSA has determined that the modification recommended by Ford is
not necessary. The agency believes that there will be no ``constant
pedal force'' increments at all, if the rates of pedal force
application are held within the limits prescribed in S7.4.3(c). The
agency notes that in evaluating this phenomenon in the context of worst
case scenarios, VRTC determined that

[[Page 6423]]
there was no significant change in the results.\10\

\10\``Harmonization of Braking Regulations, Report Number 7,
Testing to Evaluate Wheel Lock Sequence and Torque Transducer
Procedures,'' DOT HS 807611, February 1990.
Ford and MVMA commented that the test condition in S7.4.3(i), which
specifies 20 to 25 snubs from 50 km/h at each of the two loading
conditions, is excessive. They state that one or two stops from each
loading condition would be sufficient for determining variable
proportioning valve (VPV) performance. Alternatively, Ford and MVMA
stated that the digital data obtained for each of the torque wheel test
stops would provide another source of data for determining variable
proportioning valve performance. They requested that if the agency
decides to require 20 to 25 snubs, then the snubs be performed at the
end of the test sequence to avoid any non-repeatable conditioning of
the brake lining.
NHTSA has determined that 20 to 25 snubs to determine the variable
proportioning valve performance may be unnecessary, but that the
suggested 1 to 2 stops would be inadequate to cover the entire range of
brake pressures. The agency has decided to modify S7.4.3(i) to specify
15 snubs. The agency believes that this test procedure will be
sufficient to appropriately evaluate variable proportioning valve
performance without introducing unnecessary conditioning of brake
linings. The agency notes that these extra snubs are only needed when
the vehicle is equipped with a variable proportioning valve. With fixed
proportioning, the test is a static test, which will have no effect on
conditioning of the brake linings.
Ford stated that the linear regression data should only include
torque data collected when the vehicle deceleration is within the range
of 0.15g to 0.80g rather than when torque output values are > 34 N/
minute.
NHTSA agrees with Ford's comment and has modified S7.4.4(b) to
reflect this change. The agency believes that it would be inappropriate
to use data compiled outside the required performance range of the
torque wheel test, since such data may not be relevant to the actual
performance requirements.
GRRF, GM, Ford, the MVMA, Suzuki, JAMA, Toyota, Honda, and OICA
commented that the upper limit line in Figure 2 in S7.4.4(h)
(represented in S7.4.5.1 by the equation z = 0.1 + 0.7 (k-0.2)) is
unnecessary and should be eliminated. Ford and GM stated that the line
is unnecessary because, even though the wheel lock sequence test has no
check for excessive front bias, the cold effectiveness test does.
Suzuki, JAMA, Toyota, and OICA stated that the adhesion utilization
requirement in S7.4.5.2 for a rear axle is more stringent than the
requirement than S7.4.5.1, making S7.4.5.1 redundant.
NHTSA agrees with the commenters that a vehicle that is so front-
biased that it would not satisfy the efficiency requirement proposed in
Notice 5 would in all probability not be able to meet the cold
effectiveness and/or other stopping performance requirements in the
standard. Therefore, the efficiency requirement proposed in S7.4.5.1 of
Notice 5 is essentially redundant. Accordingly, the agency has decided
not to include the upper line in Figure 2. In addition to deleting the
area of Figure 2 defined by the equation z = 0.1 + 0.7 (k-0.2), NHTSA
is modifying S7.4.5 by deleting the text of S7.4.5 and S7.4.5.1, and
renumbering S7.4.5.2 as S7.4.5.
Chrysler recommended using deep dish wheels and changing tires on
the torque wheels, claiming that use of torque wheels will deform
normal road wheels by pushing them further out than their normal
position. Ford and MVMA requested that the agency modify the
requirement to permit use of a separate set of tires in the torque
wheel test, based on its concern that lockup situations in other tests
under FMVSS No. 135 could flatten or wear spots on tires.
NHTSA has decided to permit manufacturers to use a separate set of
tires for the torque wheel test, even though the agency believes that
it is unlikely that the tires will be worn down prior to the adhesion
utilization test which comes at the beginning of FMVSS No. 135's test
sequence. The agency notes that new tires will not alter the adhesion
utilization curve for the vehicle. The agency agrees with Chrysler that
manufacturers using deep dish rims can avoid tire demounting and thus
simplify testing, if they can use such rims with tires already mounted.
Based on these considerations, the agency has modified S7.4.2(d) to
permit optional use of a separate set of tires for the torque wheel
test.
Suzuki commented that for purposes of the torque wheel test, the
definition of LLVW should be changed to unloaded weight plus 200 kg,
rather than the present 180 kg. It stated that 180 kg may be
insufficient to cover the total weight of the driver and required
instrumentation.
NHTSA believes that most instrumentation packages fall within the
180 kg specified in the Standard. Moreover, the agency is not aware of
any instrumentation packages that exceed the weight allowed for LLVW
testing. Based on these considerations, the agency has decided not to
change S7.4.2.
Hunter, a manufacturer of a brake balance tester, stated that its
device can provide results similar to a road transducer pad. It further
stated that its device can be used without the need to modify the
vehicle.
NHTSA is aware of Hunter's brake balance tester, which is a
simplified version of the road transducer pad. While the Hunter device
can provide a rough measure of adhesion utilization, NHTSA believes
that the methods of measuring adhesion utilization adopted by the
agency are superior to the Hunter device, since the torque wheels
evaluate adhesion utilization more precisely. The agency notes that the
automotive industry and foreign governments interested in harmonization
have stated that the proposed methods of measuring AU are appropriate.
In the 1991 SNPRM, the agency stated that assuming one torque wheel
equipment package will service the needs for five years of typical
yearly production runs of 30,000 to 100,000 vehicles, the torque wheel
would result in a unit cost increase of $0.15 to $0.50 per vehicle.
Kelsey-Hayes stated that NHTSA underestimated the expense of torque
wheel equipment. It stated that the agency's discussion of the economic
burden associated with the cost of one set of torque wheels over a test
run is misleading and incomplete, since numerous sets of torque wheel
instrumentation will be required.
NHTSA believes that its estimates in the 1991 SNPRM were reasonably
accurate, with the following minor modifications. The agency expects
that the cost for a set of four torque wheels (including adapters to
accommodate varying wheel mounting bolt patterns) to be approximately
$40,000 and $15,000 for the on-board digital data acquisition system
that will record the testing results. The equipment should last five
production years, which correlates to an annual expense of $11,000 per
year. This figure is further reduced when amortized on a per vehicle
basis. The agency estimates that direct labor costs for each test to be
approximately $50 (including costs for instrumentation technicians, and
drivers). The agency estimates that the marginal cost increase per car
attributed to the torque wheel test will be between $0.10 and $0.16,
depending on the size of the vehicle's production run and the number of
vehicles in the run that the manufacturer wants to test, since the
manufacturer need not test every vehicle in a vehicle run. The agency

[[Page 6424]]
further notes that less than 1.0 percent of vehicles will actually have
to undergo the test by model year 1999, given that most vehicles will
be equipped with antilock systems and even most of those non-ABS
equipped vehicles will pass the wheel lock sequence test. Based on the
above considerations, NHTSA has concluded that the expense and time
required to administer the torque wheel test will not pose an
unreasonable burden on manufacturers.
The agency notes that torque wheels have been in use at least for
the last 50 years for evaluating vehicle characteristics other than
adhesion utilization. Most of the major vehicle manufacturers already
have torque wheels and use them extensively. Therefore, the cost of
torque wheels for FMVSS No. 135 needs to be amortized over more than
just its use in evaluating adhesion utilization.
No costs associated with the test surface are expected for torque
wheel testing because a high coefficient of friction test surface is
already required for testing under the existing standard. No costs are
expected for the wheel lock sequence test because, if enough surfaces
are not already available to potential users, they could use the torque
wheel test, given that it would be cheaper to use than constructing and
maintaining new test surfaces. In other words, costs associated with
the wheel lock sequence test might be so high that manufacturers would
go directly to the torque wheel test to incur lesser costs.
6. Cold Effectiveness
The cold effectiveness test evaluates the ability of a vehicle's
brake system to bring a vehicle to a quick and controlled stop in an
emergency situation. In the 1991 SNPRM, NHTSA proposed the same cold
effectiveness test as proposed in the 1987 SNPRM, with some minor
modifications. Specifically, the agency proposed that vehicles would
have to stop within 70 m in both the fully loaded and lightly loaded
conditions. Based on testing and information supplied by the
commenters, the agency believed that this stopping distance requirement
for a cold effectiveness test is equivalent in stringency to the
current requirement in FMVSS No. 105. The agency continues to believe
that the requirements for the cold effectiveness test are of equivalent
stringency, as explained below.
Like the other effectiveness tests, the proposed stopping distance
requirements for the cold effectiveness test was expressed in the form
of an equation. Specifically, this equation provides that stopping
distance must be less than or equal to 0.10V + 0.0060V, where V refers
to velocity in km/h. The first part of the equation, the 0.10V term,
accounts for brake system reaction time of 0.36 second. The second part
of the equation, 0.0060V, represents an assumed mean fully developed
deceleration rate. The specified performance criterion is not the
deceleration rate or the system reaction time, but the stopping
distance.
Commenters disagreed about the stringency of the proposed stopping
distance tests. While GRRF agreed to the proposed 70 m requirement in
the interest of harmonization, GM, Ford, MVMA, Advocates, and the CAS
disagreed with the proposed stopping distances. GM stated that the
reduction in maximum allowable pedal force increased stringency by 27
percent. It further stated that of nine cars it tested, three failed to
meet the proposed 70 m and an additional four failed to meet the 70 m
within 10 percent compliance margin. Based on this information, GM
argued that a significant number of its vehicles would fail the
proposed cold effectiveness test, even though they would comply with
FMVSS No. 105. Ford and MVMA stated that the stopping distance was
appropriate if the PFC were raised to 1.0.
In contrast, Advocates and CAS commented that the proposed stopping
distances were not sufficiently stringent. Advocates stated that the
stopping distance should be reduced from 70 m in order to force more
original equipment manufacturers to include ABS and brake power assist
units as standard equipment. CAS objected to increasing the reaction
time component in the stopping distance formula.
After reviewing the available information, NHTSA has determined
that requiring a passenger car to come to a complete stop within 70 m
(230 feet) from 100 km/h (62.1 mph) provides an appropriate level of
braking performance. The agency has decided to require the cold
effectiveness test to be conducted at both LLVW and GVWR, with the
pedal force being between 65 and 500 N (14.6 to 112.4 lbs).
As it has emphasized in earlier notices, NHTSA notes that it is
inappropriate to look only at the raw numbers in FMVSS No. 105 and
FMVSS No. 135 and state that one standard is more or less stringent
than the other. Agency tests conducted on identical vehicles to the
performance requirements in FMVSS No. 105 and FMVSS No. 135 indicate
that the average margin of compliance for the cold effectiveness tests
at GVWR in the two standards were almost identical (11.5 percent for
FMVSS No. 135, and 11.9 percent for FMVSS No. 105). Therefore, NHTSA
does not agree with GM's assertions that FMVSS No. 135 is more
stringent than FMVSS No. 105.
NHTSA notes that the stopping distances specified in FMVSS No. 135
are slightly longer than the distances specified in FMVSS No. 105.
Nevertheless, the agency is confident that the two FMVSSs provide a
comparable level of safety, for the following reasons. First, the new
burnish procedure in FMVSS No. 135, which is closer to real world
practice, is not as severe as that in FMVSS No. 105. As a result, the
longer stopping distances in the new standard are mostly attributable
to the less severe, but more realistic, burnish procedures, not to an
inherent weakening of brake efficiency requirements. Second, the
maximum allowable pedal force has been reduced from 150 lbs in FMVSS
No. 105 to 112.4 lbs in FMVSS No. 135. Along with lengthening the
stopping distances slightly, the lower pedal force will more closely
reflect the pedal forces likely to be applied by real world drivers, as
opposed to those on a test track.
NHTSA notes that CAS incorrectly assumes that increasing the brake
reaction time component in the stopping distance equation, by itself,
decreases the test's stringency. Brake reaction time is merely part of
a formula by which stopping distances are gauged, but it is the
stopping distance, and not the formula, which determines the stringency
of the rule. To illustrate, in the 1991 SNPRM, the agency increased the
reaction time component of the cold effectiveness test equation from
0.07V to 0.10V. However, the stopping distance remained at 70 m. To
compensate for this change in the system reaction time, the
deceleration term was modified slightly. Accordingly, a vehicle must
still stop in 70 m, so there is no actual increase or decrease in
stringency from the first SNPRM.
NHTSA believes that Advocates' concern about the installation of
power assist units is moot. According to Ward's Automotive Reports
(December 30, 1993 and April 18, 1994 Reports), all current U.S. cars
and import cars are equipped with power brakes. Moreover, antilock
brake systems are quickly becoming a feature available on many cars. As
stated above, by MY 1999 the agency expects 85 to 90 percent of all new
cars to be ABS-equipped. The market is responding directly to consumer
preference, and therefore Advocates' goal of having more vehicles
equipped with ABS is being achieved without a more stringent stopping
distance requirement.

[[Page 6425]]

NHTSA disagrees with GM's comment that the cold effectiveness
stopping distance requirements are 27 percent more stringent due to
lower allowable pedal force, because cold effectiveness stops are
usually not pedal force limited. In other words, despite the maximum
allowable pedal force of 150 lbs in FMVSS No. 105, vehicles rarely
needed to be braked with such a pedal force to pass the stopping
distance requirement. In fact, pedal forces rarely exceeded the 112.4
lbs (500 N) permitted in FMVSS No. 135. Therefore, the agency does not
believe that the lower maximum pedal force allowed in the new standard
will result in increasing the stringency of the cold effectiveness
requirements in comparison with FMVSS No. 105.
Toyota commented that the minimum initial brake temperature should
be raised from 50 deg.C to 65 deg.C, but did not give any reasons for
the request.
Based on testing conducted at VRTC, NHTSA believes that the present
minimum initial brake temperature, which was proposed in the NPRM and
the two SNPRMs, represents an appropriate temperature at which to begin
the cold effectiveness test runs, and has no information indicating it
should be changed. Therefore, the agency is retaining the initial brake
temperature requirement as proposed.
7. High Speed Effectiveness
In the 1991 SNPRM (Notice 5), NHTSA proposed a high speed
effectiveness test because cars are sometimes driven at higher speeds
than provided for in the cold effectiveness test that is conducted at
100 km/h (62.1 mph). The agency proposed that under the high speed
effectiveness test for vehicles capable of a maximum speed over 125 km/
h, a vehicle would be tested at a speed representing 80 percent of its
maximum speed, with a maximum limit of 160 km/h (99.4 mph). The upper
speed limit was specified due to facility limitations and safety
concerns during testing. The agency proposed that the high speed test
would only be conducted for vehicles with a maximum speed greater than
125 km/h. The agency proposed a new equation to reflect the change in
system reaction time from 0.07V to 0.10V. The agency stated that while
the SNPRM proposal is more stringent than the latest GRRF proposal, the
agency's test data indicated that all test cars would be able to meet
the proposed requirement.
The GRRF generally accepted the high speed effectiveness formula,
and the maximum test speed limit. Nevertheless, it requested that NHTSA
delete the lower speed limit proposed in the 1991 SNPRM, since R13 does
not specify a lower limit. GRRF further stated that the cold
effectiveness test and high speed effectiveness tests are qualitatively
different because the former is run with the engine in neutral, while
the latter is run with the engine in gear.
NHTSA is pleased that the GRRF has agreed to incorporate the
proposed high speed test in R13H. Nevertheless, the agency believes
that it is necessary to include the lower limit test speed.
Accordingly, NHTSA has decided not to conduct the high speed test for
vehicles with a maximum speed under 125 km/h, since it would be
illogical and would provide no safety benefits to conduct a high speed
test at a lower speed than the speed required by the cold effectiveness
test. The agency notes that 80 percent of the lowest maximum speed for
the high speed effectiveness test is 100 km/h. The agency does not
believe that running a high speed test at a speed lower than 100 km/h,
the cold effectiveness test speed, is worthwhile, regardless of engine
drive position.
Ford commented that the test should be run only at GVWR, but gave
no reason for deleting the LLVW run.
NHTSA has decided that it is consistent with the interests of motor
vehicle safety to test at both GVWR and LLVW since vehicles are used at
both weights. Similarly, it is in the interest of international
harmonization to test at both load conditions, since R13 does so.
Accordingly, in FMVSS No. 135's high speed effectiveness test, a
vehicle will be tested at both LLVW and GVWR. The test will be
conducted at a pedal force between 65 and 500 N (14.6 to 112.4 lbs).
JAMA and Toyota recommended specifying only four runs at high
speeds instead of the six proposed in the 1991 SNPRM.
NHTSA previously addressed this issue in the 1987 SNPRM in which
the agency proposed increasing the number of test runs from four to
six. In that notice, NHTSA explained that such a change would minimize
driver effects and decrease test variability, because the prescribed
performance would have to be achieved on only one stop in the six runs.
Even though reducing the number of runs to four might nominally
decrease the expense of the test, such a change could increase the
test's stringency.
8. System Failure
In previous notices, NHTSA proposed stopping distance requirements
for situations involving the engine being off, antilock functional
failure, variable proportioning valve failure, hydraulic circuit
failure, and the power assist unit being inoperative. Aside from the
engine off requirement, FMVSS No. 105 includes similar requirements
which are crucial if part of the service brake system or engine should
fail or become inoperative. These requirements ensure that the
vehicle's brake system will still be able to bring the vehicle to a
controlled stop within a reasonable distance.
a. Stops with engine off.--In the NPRM and two SNPRMs, NHTSA
proposed requirements to address stops with the engine off. The agency
explained that the proposed requirement was reasonable since engine
stalling is a relatively common occurrence, even though FMVSS No. 105
does not include a comparable requirement. The proposal to require
vehicles to stop within 73 m after engine failure was slightly less
stringent than the 1987 SNPRM's proposed requirement for stops within
70 m. The agency stated that the proposal was consistent with the
latest proposal by GRRF and thus will promote harmonization.
Advocates and CAS were concerned that the longer permissible
stopping distance of 73 m in the engine failure condition would
increase crashes. The GRRF recommended that the vehicle be able to stop
after engine failure within 70 m rather than the proposed 73 m. The
GRRF stated that the requirements of R13 and R13H should be easily met,
provided that there is an adequate reservoir in the braking system and
a non-return valve is fitted to the brakes. This equipment should
ensure that the brakes can operate even without the engine running.
NHTSA has decided to adopt the engine failure test with a stopping
distance of 70 m. Throughout the rulemaking, the agency has attempted
to make the engine failure stopping distance consistent with GRRF and
consistent with the stopping distance requirement in the cold
effectiveness test. In the 1991 SNPRM, the agency stated that its
proposal was consistent with the GRRF. This was true when the stopping
distance was 73 m for both the cold effectiveness and engine off tests.
Since the cold effectiveness stopping distance is now 70 m, the agency
is adopting a stopping distance of 70 m for the engine off test. The
engine off test will be performed at GVWR, with six stops from 100 km/
h, using a pedal force between 65 N and 500 N.
b. Antilock functional failure.--In the two SNPRMs, NHTSA proposed
separating the antilock and variable proportioning valve failure
requirements into different sections to

[[Page 6426]]
reflect the differing failure modes. In the 1991 SNPRM, the agency
proposed slightly different stopping distances to reflect the increase
in system reaction time and higher decelerations on the cold
effectiveness test, while maintaining the same percentages as in the
1987 SNPRM.
For Antilock functional failure, NHTSA proposed a stopping distance
of 85 m from a test speed of 100 km/h. The proposed requirement would
apply only to functional failures of the ABS system and not to
structural failures that are covered by the hydraulic circuit failure
requirements. The proposed stopping distance maintains the philosophy
that antilock functional failure performance should be 80 percent of
the cold effectiveness performance requirement, and is consistent with
the requirements adopted for Regulation R13H.
Without explaining what it perceived to be inconsistent, Fiat
requested that the agency make the antilock failure requirements in
FMVSS No. 135 consistent with R13H. Advocates and CAS requested that
NHTSA adopt a stopping distance of 80 meters as proposed in the NPRM.
They commented that the SNPRM's proposed stopping distance of 85
meters, while lower than the distance proposed in the 1987 SNPRM, still
exceeded the NPRM by 5 meters.
NHTSA has decided to adopt the 85 meter stopping distance
requirement for antilock functional failure, as proposed. The agency
believes Fiat's comment must have been based on a mistaken impression
that the requirement in Regulation 13H was some other value. In fact,
the two requirements are harmonized.
The observations of CAS and Advocates that the performance
requirement has changed by 5 meters since the NPRM (Notice 1) is
correct. Due to various changes in the equations, which have been
explained in the two SNPRMs, the proposed requirement went from 80
meters to 86 meters, and then back to 85 meters. Nevertheless, the 80
percent of cold effectiveness performance concept has been maintained
throughout this rulemaking. The value being adopted is in agreement
with that philosophy, is harmonized with the proposed Regulation 13H,
and is considerably more stringent than the corresponding requirement
in FMVSS No. 105. CAS and Advocates have provided no justification for
returning to an 80 meter value.
Ford, ITT-TEVES, GM, BMW, Chrysler, the GRRF, and MVMA requested
that the agency clarify the definition of an ABS ``functional failure
simulation'' to indicate that only the ABS system is covered by this
requirement. GM and Chrysler stated that the ABS failure test should
not be misunderstood to include failures affecting other aspects of the
service brake system. They explained that although ABS have previously
been added on to the service brake system, increasingly ABS is
completely integrated into the service brake system.
Based on the comments, NHTSA believes that it is necessary to
clarify the meaning of the phrase ``any single functional failure in
any such system.'' Since this requirement applies to antilock systems,
only a failure in an antilock system is covered by this requirement.
Nevertheless, if a functional failure of the ABS also affects or
degrades the service brake system, no artificial means are entailed to
keep the service brake system intact when that failure is introduced.
In such a situation, the vehicle with the failed ABS and failed service
brake system resulting from the single failure, will then be subject to
both the ABS failure and partial system failure tests. As the
commenters state, manufacturers are increasingly building integrated
brake systems rather than installing add-on antilock systems. The
agency believes that this requirement is appropriate since it will
prohibit any single ABS failure from degrading the service brake
systems beyond the performance requirements of the ABS failure test. To
ensure clarity, NHTSA has decided to add the following provision to
S7.8.2(g)(1): ``Disconnect the functional power source, or any other
electrical connector that would create a functional failure.''
Ford recommended deleting the ABS functional failure test at LLVW,
stating it was the same as the LLVW cold effectiveness test, if the ABS
functional failure is limited to a non-actuation failure mode. In the
cold effectiveness test, ABS is active and therefore may actuate during
the test. For the ABS functional failure test, the ABS is not working.
If the ABS is of an add-on type design rather than an integrated
system, and if the cold effectiveness test is conducted at a brake
force level that does not result in activation of the ABS, then it is
true that the tests would be redundant. However, in many cases one or
both of those conditions are not met, so the tests would be different.
Therefore, it would be inappropriate to delete the test as requested by
Ford.
Bendix stated that with respect to S7.8.2(g)(2)11, the
electrical function failure induced should be one that makes the system
inoperative in order to activate the warning indicator. Kelsey-Hayes
requested that the agency clarify the meaning in S7.8.2(g)(2) about the
continuing operation of the system.

\11\This section requires a determination of whether an ABS
electrical functional failure activates the brake system warning
indicator.
---------------------------------------------------------------------------

An electrical functional failure that makes the ABS inoperative is
required by S5.5.1(b) to activate the warning indicator. S7.8.2(g)(2)
is the test to determine compliance with S5.5.1(b). In response to
Kelsey-Hayes, the agency notes that an unplugged ABS module should
activate the antilock system warning indicator. The agency has decided
to clarify paragraph S7.8.3 by adding the words ``service brake''
before the word ``system.''
c. Variable brake proportioning functional failure.--In the 1991
SNPRM (Notice 5) NHTSA proposed a stopping distance of 110 meters from
a test speed of 100 km/h to evaluate variable proportioning valve
failure. This was slightly shorter than the distance of 112 meters
proposed in the 1987 SNPRM. In both notices, the proposal was based on
the mean fully developed deceleration rate of 60 percent of that
required for the cold effectiveness test. In the 1991 SNPRM, the agency
revised the proposal to better define how a variable proportioning
valve failure is simulated and to clarify that a warning to the driver
of valve failure is only required where there is an electrical
functional failure in the variable proportioning valve.
Fiat commented that the variable proportioning valve functional
failure test is not necessary given that neither EEC directive 75-524
nor R13 and R13H test for this type of failure, despite years of
experience.
NHTSA believes that the lack of documented variable proportioning
valve passenger car failures in the U.S. is not a sufficient reason
against specifying this requirement. The agency notes that there have
been considerable problems with variable proportioning valves on
trucks, the vehicle type most typically equipped with variable
proportioning valves, both in the U.S. and in Europe. Fiat produced no
data to support its assertion that the test is unnecessary for
passenger cars. NHTSA notes that a corresponding requirement is
included in the proposed Regulation 13H.
ITT-TEVES recommended a stopping distance of 168 m for the variable
proportioning valve failure test. It reasoned that vehicles would not
be able to meet the 110 m stopping distance because of wheel lock
caused by a dynamic load transfer from the rear to the front of the
vehicle during braking.

[[Page 6427]]

NHTSA disagrees with ITT-TEVES recommendation to dramatically
increase the stopping distance requirement for the variable
proportioning valve test. The agency believes that it would be
inconsistent with motor vehicle safety to allow a vehicle that is so
greatly influenced by an operational variable proportioning valve that
when the valve fails the brakes lock up and the vehicle needs 168
meters to stop. The agency further notes that the problem discussed by
the commenter, which might affect trucks in rare cases, is even less
likely to affect passenger cars.
The GRRF stated that the 60% cold effectiveness requirement is more
stringent than the European specification in Regulation 13.
Nevertheless, the GRRF stated that it could accept the proposed
performance requirement for variable proportioning valve functional
failure for purposes of Regulation 13H, provided that its concerns set
forth below with respect to S7.9.2(g)(1) are met.
Chrysler, Ford, MVMA, and the GRRF commented that when a variable
proportioning valve is disconnected or fails for any reason, it reverts
to a default position, functioning at the lowest pressure possible in
its proportioning range. Therefore, they state that S7.9.2(g)(1) should
be changed to reflect this default condition. They believe that to
require the proportioning valve to be operated in any specified
position in its operating range would require equipment that is not
found on current vehicles.
NHTSA agrees with the commenters that S7.9.2(g)(1) should be
revised to allow the variable proportioning valve to return to its
normal, default, position, when disconnected, since this will more
accurately test the vehicle's real world braking ability. Accordingly,
the agency has decided not to require the variable proportioning valve
to be held in any position in its operating range, thus allowing it to
revert to its uncontrolled condition.
NHTSA notes that the stopping distances for variable proportioning
valve functional failure are shorter than those of FMVSS No. 105 (while
the stopping distances for structural failure are longer). The agency
has determined that the stopping distances which are more stringent for
functional failures are appropriate, since functional failures are more
likely to occur.
d. Hydraulic circuit failure. In the 1991 SNPRM (Notice 5), NHTSA
proposed a stopping distance of 168 m (551 feet) from a test speed of
100 km/h. This proposal is identical to that included in the proposed
Regulation 13H. It maintains the same deceleration term as in the 1987
SNPRM (Notice 4), but reflects the proposed reaction time changes in
the equation for the cold effectiveness performance requirement.
Advocates stated that increasing the stopping distance in the
hydraulic circuit failure test by 42 feet from the NPRM (Notice 1)
decreased the Standard's stringency compared to the initial proposal.
It further stated that the 1991 SNPRM (Notice 5) also was less
stringent than the 1987 SNPRM (Notice 4). There were no other comments
regarding the stringency of this requirement.
Based on testing and other available information, NHTSA has decided
to adopt the proposed stopping distance of 168 meters (551 feet) from a
test speed of 100 km/h for both the hydraulic circuit failure tests.
The agency has decided to adopt the stopping distance formula
(0.10V+0.0158V2), as proposed in the 1991 SNPRM. As explained in
previous notices, it is not possible to compare the stringency of FMVSS
No. 105 and FMVSS No. 135 directly when discussing hydraulic circuit
failure requirements. This is primarily because there is a significant
difference in allowable pedal force during the test. FMVSS No. 105
limits pedal force to 150 lbs, whereas the maximum pedal force in FMVSS
No. 135 is 500 N (112.4 lbs). Although as a general matter, the
stopping distance of a vehicle improves as greater pedal force is
applied, it is not possible to quantify a precise relationship between
stopping distance and pedal force. The relationship between these
factors is non-linear; it varies among vehicle models, and depends upon
various parts of the vehicle, including tires and brake system
components. It is broadly true, however, that as pedal force increases,
stopping distance decreases.
In response to Advocates' comment regarding the changes between the
1985 NPRM (Notice 1) and the 1991 SNPRM (Notice 5), the rationale for
those changes was set forth in the two SNPRMs.
Bendix requested that S7.10.3(f) be clarified so that the induced
failure for testing would be limited to the normal braking circuits,
but not as part of the ABS that is not part of the normal braking
circuit.
NHTSA notes that it is not clear exactly what Bendix means by
``normal braking circuits.'' Section S7.10.3(f) states that the failure
is to be induced in the service brake system. The failure could be
anywhere in that system, including any part of an ABS that is common to
the service brake system. Any part of the ABS that is not common to the
service brake system would be subject to testing to the failed ABS
requirements, not the hydraulic circuit failure requirements. The
agency believes the test condition is clear as stated, and further
clarification is unnecessary. Therefore, S7.10.3(f) is adopted as
proposed.
e. Power assist unit inoperative. In the 1991 SNPRM, NHTSA proposed
a stopping distance of 168 m (551 feet) from a test speed of 100 km/h.
This proposal is identical to that included in the proposed Regulation
13H. It maintains the same deceleration term as in the 1987 SNPRM, but
reflects the proposed reaction time changes in the equation for the
cold effectiveness performance requirement.
Advocates opposed the proposed stopping distance of 168 m for stops
with an inoperative power assist, stating that it compared unfavorably
with the 165 m proposed in the 1987 SNPRM and the 155 m proposed in the
NPRM. In contrast, Ford and GM stated that the agency had proposed a
significant increase in stringency from FMVSS No. 105. These commenters
recommended a stopping distance of 177 meters (580 ft), stating that
such a distance would be equivalent to R13, and would still be more
stringent than the 456 foot stopping distance in FMVSS No. 105 because
of the decreased maximum pedal force.
After reviewing the comments, NHTSA has decided to adopt the
proposed stopping distance of 168 meters (551 feet) from a test speed
of 100 km/h for stops when the power assist is inoperative. The agency
has decided to adopt the stopping distance formula,
(0.10V+0.0158V2), as proposed in the 1991 SNPRM.
As explained in the section on hydraulic circuit failure, it is not
possible to compare the stringency of FMVSS No. 105 and FMVSS No. 135
directly when discussing power assist failure requirements, primarily
because there is a significant difference in allowable pedal force
during the test. None of the commenters who asked for a more or less
stringent stopping distance value provided justification for their
requests.
9. Parking Brake Requirements
a. Dynamic test. In the NPRM and 1987 SNPRM, NHTSA proposed a
dynamic parking brake test that it believed was consistent with the
GRRF decisions. The dynamic test was intended to ensure that the driver
could use the parking brake to stop a moving vehicle during emergency
situations. In the 1991 SNPRM, NHTSA proposed requiring that vehicles
utilizing the

[[Page 6428]]
service brake's friction linings for the parking brake be tested at a
speed of 80 km/h and that vehicles utilizing separate friction linings
for the parking brake be tested at 60 km/h. The agency decided that it
was not necessary to include a stopping distance requirement, as was
proposed in the 1987 SNPRM.
Volkswagen, Mercedes Benz, GM, Suzuki, MVMA, Chrysler, Ford, and
OICA objected to the proposed dynamic parking brake test. These
commenters stated that the agency had not identified any safety need
for a dynamic parking brake test and that FMVSS No. 105 has no such
test. These commenters stated that such a test is neither needed nor
appropriate since the primary purpose of the parking brake is to
statically hold a vehicle on a gradient and not to provide deceleration
capabilities for a moving vehicle. They state that it is potentially
dangerous for drivers to apply parking brakes in a dynamic situation
because it is difficult to modulate the application force. Moreover,
such applications could lead to uncontrollable rear wheel lock up and
loss of vehicle control.
Volkswagen, Mercedes Benz, GM, Suzuki, MVMA, Chrysler, Ford, and
OICA stated that the dynamic parking test was adopted in ECE R13 prior
to the almost universal use of dual split service brake systems. Such
brake systems provide extra braking reserves in the event of a partial
failure because an independent part of the split system remains intact
and unaffected by the failure in the other part of the system.
According to the commenters, ECE is no longer working on revising its
dynamic test, and is even discussing eliminating it.
Mercedes commented that a dynamic test penalizes parking brake
designs that are highly self energizing (i.e., that require a
relatively low control force but are highly effective in static
situations) because their static-efficient design makes them more
susceptible to fading. It stated that deleting the dynamic test would
improve the design of parking brakes by permitting the optimization of
their static holding performance.
In contrast, Advocates and CAS supported including a dynamic
parking brake test, although they opposed the agency's decision not to
propose stopping distance requirements in the 1991 SNPRM. Advocates
stated that the important function of a dynamic standard for parking
brake performance is the ability to control manufacture of parking
brake systems either with or without separate friction that will
reasonably stop a car from controlling test speeds when there is a
complete failure of service brakes. That organization stated that
without a specific stopping distance requirement, the agency was
essentially conceding its attempt to strengthen .105 in order to ensure
adequate dynamic performance of the parking brakes when all service
brakes fail.
CAS commented that NHTSA's defect files contradict GM's comment
that current brake system designs ``obviate the safety need'' for
emergency brakes and performance standards. It believed that in many
instances drivers have had to use the emergency brake as a last resort
to stop the car.
After reviewing the available information, NHTSA has determined
that a dynamic parking brake test would provide no significant safety
benefits. This decision is based on the fact that FMVSS No. 105 does
not include a dynamic parking brake test and on the current state of
braking technology. As the manufacturers correctly stated, the ECE
requirement pre-dated the widespread use of split service brake
systems, which are now standard on all passenger cars. Therefore, the
justification for using the parking brake in an emergency situation is
no longer relevant. The agency further notes that the partial failure
requirements are sufficient in dynamic emergency situations.
Advocates and CAS argued that these requirements are needed to
address the situation of ``complete failure'' of a service brake
system. The agency has no evidence that complete brake failure
(simultaneous failure of both circuits of a split brake system) occurs
with any significant frequency. Moreover, because the parking brake is

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A95-2324. Public record. Not legal advice.