# Federal Motor Vehicle Safety Standards Rear Impact Guards; Rear Impact Protection

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A96-682

## Record

- **Collection:** Federal Register
- **Document type:** Rule
- **Published:** January 24, 1996
- **Citation:** 61 FR 2004

## Text

SUMMARY: This final rule establishes two Federal Motor Vehicle Safety
Standards (FMVSS) which will operate together to reduce the number of
injuries and fatalities resulting from the collision of passenger
vehicles with the rear end of heavy trailers and semitrailers. The
first standard (FMVSS No. 223, Rear Impact Guards, or the ``equipment
standard'') specifies performance requirements that rear impact guards
(guards) must meet before they can be installed on new trailers and
semitrailers. It specifies strength requirements, as well as test
procedures that NHTSA will use to determine compliance with the
standard. The guard may be tested for compliance while mounted to a
non-vehicle ``test fixture'' or a complete vehicle. The equipment
standard also requires the guard manufacturer to provide instructions
on the proper installation of the guard. The final rule also specifies
requirements to ensure energy absorption by the guards.
The second standard (FMVSS No. 224, Rear Impact Protection, or the
``vehicle standard'') requires that most new trailers and semitrailers
with a Gross Vehicle Weight Rating of 4,536 kilograms (kg) (10,000
pounds (lbs)) or more be equipped with a rear impact guard meeting the
equipment standard. Requirements for the location of the guard relative
to the rear end of the trailer are also specified in the vehicle
standard. The vehicle standard further requires that the guard be
mounted on the trailer or semitrailer in accordance with the
instructions of the guard manufacturer.

DATES: This rule will become effective on January 26, 1998. Petitions
for reconsideration of this rule must be received no later than March
11, 1996.

ADDRESSES: Petitions for reconsideration should refer to the docket
number and notice number and be submitted in writing to: Docket
Section, National Highway Traffic Safety Administration, Room 5109, 400
Seventh Street, SW, Washington DC 20590. Telephone: (202) 366-5267.

FOR FURTHER INFORMATION CONTACT: Dr. Leon DeLarm, Dr. George
Mouchahoir, or Mr. Sam Daniel, in the Office of Vehicle Safety
Standards (Telephone: 202-366-4919), or Mr. Paul Atelsek, in the Office
of the Chief Counsel (202-366-2992), National Highway Traffic Safety
Administration, 400 Seventh Street, SW, Washington, DC 20590.

SUPPLEMENTARY INFORMATION:
Table of Contents

I. The Safety Problem
II. Existing Regulations
III. Past Proposals
IV. Summary of the 1981 NPRM
V. Summary of 1981 NPRM Comments
VI. Summary of the 1992 SNPRM
VII. Summary of 1992 SNPRM Comments
VIII. Recent Testing by NHTSA
IX. Overview of the Final Rule
X. Summary of Changes From the 1992 SNPRM
XI. Analysis and Response to Comments on the 1992 SNPRM
A. Separate Equipment and Vehicle Standards
B. Standard for Equipment
1. Relationship of Strength, Energy Absorption, and PCI
2. Guard Strength
3. Guard Energy Absorption
4. Vertical Cross-sectional Height of Horizontal Cross-member
5. Shape of the Horizontal Cross-member
6. Guard Attachment
7. Compliance Test Requirements and Procedures
a. Dynamic Versus Static Testing
b. Test Sites
c. Labeling and Certification
C. Standard for Vehicles
1. Configuration Issues
a. Maximum Guard Ground Clearance
b. Guard Width
c. Specification of the Rear Extremity
d. Distance Between the Guard Rear Surface and the Vehicle Rear
Extremity
2. Exclusions
a. Single Unit (Straight Body) Trucks
b. Special Purpose Vehicles
c. Wheels Back Vehicle
D. Costs
E. Benefits
F. Lead Time
G. Miscellaneous Issues
1. Metric System Units
2. Federal Highway Administration Rulemaking on Underride Guards
XII. Rulemaking Analyses and Notices
A. Executive Order 12866 (Federal Regulation) and Regulatory
Policies and Procedures
B. Regulatory Flexibility Act
C. Executive Order 12612 (Federalism)
D. Preemptive Effect and Judicial Review
E. Paperwork Reduction Act
I. The Safety Problem
This rule addresses the problem of rear underride crashes, in which
a passenger car, light truck, or multipurpose vehicle with a Gross
Vehicle Weight Rating (GVWR) of 4,563 kg (10,000 lbs) or less (referred
to collectively in this rule as passenger vehicles) collides with the
rear end of a trailer or semitrailer (trailers and semitrailers are
referred to collectively in this rule as trailers) and the front end of
the passenger vehicle slides under (i.e., underrides) the rear end of
the trailer. Underride occurs to some extent in most collisions in
which a passenger vehicle crashes into the rear end of a large trailer
because most trailer beds are higher than the hoods of passenger
vehicles. In the worst cases, referred to as passenger compartment
intrusion (PCI) or ``excessive underride'' crashes, the passenger
vehicle underrides so far that the rear end of the trailer strikes and
enters its passenger compartment. PCI collisions generally result in
passenger vehicle occupant injuries and fatalities caused by occupant
contact with the rear end of the trailer.
The solution to PCI is upgrading underride guards to make them
stronger, but this introduces another concern. Even if guards succeed
in preventing PCI, overly rigid guards may stop the passenger vehicle
too suddenly, resulting in excessive occupant compartment deceleration
forces and killing or injuring passenger vehicle occupants.
The agency estimates that about 11,551 rear-end crashes with
trucks, trailers, and semitrailers occur annually. These crashes result
in approximately 423 passenger vehicle occupant fatali-ties and about
5,030 non-fatal injuries.
II. Existing Regulations
The initial Federal regulation addressing the issue of heavy
vehicle rear underride was issued in 1953 by the Bureau of Motor
Carriers of the Interstate Commerce Commission (presently the Office of
Motor Carriers of the Federal Highway Administration, DOT). This
regulation (49 CFR 393.86), which is still in effect, requires heavy
trucks, trailers, and semitrailers to be equipped with a rear-end
device designed to help prevent underride. The rule requires that the
ground clearance of the underride guard not exceed 760 mm (30 inches
(in)) when the vehicle is empty. The rule also requires that the device
be located not more than 610 mm (24 in) forward of the rear of the
vehicle and that it extend laterally to within 460 mm (18 in) of each
side. The regulation further requires that the ``[guards] shall be
substantially constructed and firmly attached.''
The Research and Special Programs Administration (RSPA) of DOT has

[[Page 2005]]

specified configuration requirements for guards on tankers that carry
hazardous materials (49 Part 178.345-8). The bottom of the guard must
be at least 100 mm (4 in) below the lower surface of any part of the
rear of the vehicle, and not more than 1,520 mm (60 in) from the ground
when the tanker is empty. The guard must be very strong. It must
deflect 150 mm (6 in) forward when subjected to a 20 m/s\2\ (2 G)
impact while loaded, without contacting the cargo tank. These
requirements are designed primarily to protect the tank and piping, not
the colliding vehicle, in the event of a rear end collision.

III. Past Proposals

From time to time, NHTSA has assessed the requirements of the
Federal Highway Administration's (FHWA) regulation and considered
whether NHTSA should issue a Federal Motor Vehicle Safety Standard
(FMVSS) requiring heavy vehicles to be equipped with rear underride
protection. The issues of particular concern have been the requirements
for rear end guard ground clearance, guard strength, and the injury and
fatality benefits of such a standard. The most recent of several NHTSA
notices was a Supplemental Notice of Proposed Rulemaking (SNPRM) issued
in 1992 (57 FR 252; January 3, 1992). Prior to the 1992 SNPRM, the
agency issued a Notice of Proposed Rulemaking (NPRM) in 1981 (46 FR
2136; January 8, 1981.) The notices of proposed rulemaking issued by
NHTSA and FHWA prior to the 1981 NPRM are cited and discussed in the
1981 NPRM (Docket 1-11; Notice 8).

IV. Summary of the 1981 NPRM

The 1981 NPRM proposed to adopt a FMVSS for all new trucks and
trailers with a GVWR of 4536 kg (10,000 lbs) or more. This NPRM was
issued after research and computer modeling studies indicated that it
was feasible to manufacture light-weight guards that could prevent
excessive underride and absorb crash energy. Guard energy absorption is
important because overly rigid guards could result in passenger
compartment forces that would increase the risk of occupant injuries
even in the absence of underride.
The 1981 NPRM proposed that heavy trailers, semitrailers, and
single unit (i.e., unarticulated) trucks be equipped with an underride
guard that met certain requirements for strength and configuration. The
NPRM proposed exclusions from this requirement for trailers with
chassis that are low enough to the ground to meet the configuration
requirements for the underride guard (low chassis vehicle), trailers
that have the rear tires set back to within 305 mm (12 in) of the rear
(wheels back vehicle), and trailers that have work-performing equipment
in the lower rear whose function would be impaired by a guard (special
purpose vehicle).
NHTSA tentatively concluded that the proposed standard was superior
to the FHWA regulation in three major ways. First, NHTSA specified
objective requirements for guard strength (FHWA requires that the guard
be ``substantially constructed and firmly attached''). Second, the NPRM
proposed a guard configuration that permitted less ground clearance 560
mm (22 in), less longitudinal distance between the guard and the
trailer rear extremity 305 mm (12 in), and less lateral distance
between the guard and the vehicle side extremities 100 mm (4 in), than
the FHWA regulation. Third, the NPRM specified detailed procedures for
testing the guards as installed on the vehicle for which they were
intended by applying a specific force at certain points on the guard.

V. Summary of 1981 NPRM Comments

The agency received over 100 comments on the NPRM. Many of the
comments were from vehicle manufacturers and operators who believed
their vehicles should be excluded from the requirements because they
were special purpose vehicles. Some commenters objected to the proposed
requirements and suggested alternative means of reducing the injuries
and deaths caused by rear underride crashes. The alternative approach
most often cited involved reducing the incidence of underride crashes
through improved heavy vehicle conspicuity.
The agency agreed that conspicuity was an important issue. The
Fatal Accident Reporting System (FARS, a database containing a census
of all vehicle fatalities in the U.S.) statistics had indicated that
about 65 percent of the fatalities resulting from passenger vehicle
collisions with the rear end of heavy vehicles occurred under non-
daylight conditions. NHTSA conducted a fleet study between 1980 and
1985 of the effectiveness of improved conspicuity. As a result of this
study, the agency determined that conspicuity improvement could reduce
the incidence of the accidents by about 15 percent. Consequently, the
agency published a NPRM on improved heavy vehicle conspicuity in
December 1991, (56 FR 63474) and a final rule on conspicuity
improvement in December 1992 (57 FR 58406).
The agency believes, however, that improved rear impact guards
could mitigate some of the rear impact fatalities and serious injuries
not addressed by the improved conspicuity rule. The rear impact guard
is especially important in cases in which the passenger vehicle
driver's abilities are impaired by alcohol or drowsiness. Accident data
indicate that alcohol is a factor for passenger vehicle drivers in
about 30-40 percent of fatal rear underride accidents.
Commenters on the 1981 NPRM also expressed concern that the
proposed requirements would be a substantial financial burden on some
truck and trailer manufacturers. Several commenters argued that the
agency's cost estimate for rear underride guards was well below the
actual cost of equipping the wide variety of single unit trucks with
compliant guards. As to the trailer manufacturing industry, its members
were said to be predominantly small firms that lack the engineering
capabilities to meet the requirements of the proposed rule. In response
to the comments and statistical data, the agency sought to determine if
it could revise the proposed rule to reduce the financial burden on the
manufacturers.

VI. Summary of the 1992 SNPRM

The 1992 SNPRM contained requirements that are similar to those in
the 1981 NPRM in terms of the guard's strength and configuration.
However, the SNPRM differed substantially from the NPRM in terms of its
impact on the industry. In place of the 1981 proposal of a single
vehicle standard specifying the testing of guards on a completed
vehicle, the SNPRM proposed two standards: (1) An equipment standard
providing for the testing of guards on a test fixture, and (2) a
vehicle standard requiring installation of guards complying with the
equipment standard.
The equipment standard proposed strength requirements and an
objective test for determining compliance with these requirements. The
guard manufacturer would conduct a test involving quasi-static loading
of the guard with the guard mounted on a rigid test fixture rather than
installed on a completed vehicle. Guards certified as passing the test
could then be marketed to vehicle manufacturers for installation in
accordance with the configuration requirements of the vehicle standard.
Testing in this manner would relieve vehicle manufacturers, especially
small ones, of the burden associated with compliance testing.
The other major difference from the NPRM is that the SNPRM proposed
to exclude single unit trucks from the rulemaking. NHTSA added this

[[Page 2006]]
exclusion to those in the NPRM because single unit trucks are far less
likely to be involved in fatal accidents than combination trucks (i.e.,
trailers and semitrailers). FARS and GES accident statistics indicate
that only about 27 percent of the 423 average annual rear end
fatalities and 18 percent of the 5,030 injuries involve single unit
trucks, even though these vehicles represent 72 percent of the
registered heavy vehicles. Thus, single unit trucks are significantly
under-represented in rear end crashes. On the other hand, trailers are
highly over-represented in rear end crashes, as they represent only 28
percent of the registered heavy vehicles, but account for 73 percent of
the occupant fatalities and 82 percent of the injuries. Therefore, the
agency believed that excluding single unit trucks from the proposed
rule would result in a better rule in terms of the ratio of benefits to
costs.

VII. Summary of 1992 SNPRM Comments

The agency received approximately 2,250 individual comments on the
SNPRM. Industry-related comments were generally supportive of the
proposal, while consumer interest organizations, local and State
governments, and private citizens were generally critical.
Representing the industry were comments from automobile and truck
manufacturers, trade associations, manufacturers of trailers and
semitrailers, and manufacturers of specialized usage heavy duty
vehicles. Most of these commenters supported Federal rulemaking in this
area. The trade associations and manufacturers of trucks and trailers
were generally in agreement with the proposed requirements.
Manufacturers and operators of specialized vehicles suggested that the
proposed rule be modified to better define the types of vehicles that
would be excluded from the standard.
The vast majority of the critical comments were post cards or
letters with multiple signatures from private citizens. These post
cards and letters, as well as more detailed submittals from consumer
interest organizations, expressed concern that the agency's proposal
had three deficiencies. First, the commenters recommended that the
rulemaking apply to single unit trucks as well as trailers and
semitrailers. Second, the commenters recommended that the proposed
maximum ground clearance, 560 mm (22 in), be reduced to a 405 to 455 mm
(16 to 18 in) range. Third, these commenters expressed the opinion that
the agency should mandate ``energy absorbing'' rear impact guards for
heavy vehicles, i.e., guards with hydraulic pistons or shock absorbers
designed to deflect or deform in a controlled manner upon impact and
thereby lessen the deceleration experienced by passenger vehicles
colliding with them. Several consumer interest organizations and
private citizens also suggested that the proposed minimum guard
strength requirements were insufficient.
The consumer interest organizations and some private citizens also
expressed concern that the proposed equipment standard for the rear
impact guard did not require guards to be tested while mounted on a
vehicle. As a result, guards complying with the proposed strength
requirements could be installed on vehicles in such a location or in a
manner that the guard/vehicle combination would be ineffective. In
addition, some of these commenters stated that the crash tests NHTSA
relied on in formulating the SNPRM were inadequate because they were
not conducted under representative conditions of guard height, car
bumper height, and car speed. Specifically, they stated that car bumper
height would be depressed if the driver were braking to avoid colliding
with the trailer, thus increasing the likelihood that the car hood
would underride a 560 mm (22 in) high guard without engaging any
substantial body structure. The consumer interest organizations also
questioned the validity of the FARS accident data that NHTSA used to
determine the benefits of the SNPRM, contending that the agency had
underestimated the benefits of the rule. The state and local
governments that commented expressed concerns similar to those raised
by private citizens and consumer interest organizations.
A summary of comments has been prepared and is available for
inspection in Docket No. 1-11. Significant SNPRM issues raised by the
commenters and NHTSA's response to the comments are discussed below. In
response to the comments, the final rule includes several modifications
to the rule proposed in the SNPRM, including clarified definitions,
improved compliance test procedures, and a minimum guard energy
absorption requirement.

VIII. Recent Testing by NHTSA

In responding to comments to the SNPRM and a congressional request
for a report on several heavy truck underride issues, NHTSA conducted a
research project on Heavy Truck Rear Underride Protection at the
Vehicle Research and Testing Center (VRTC) between September 1992 and
June 1993 to evaluate the effectiveness of an underride guard meeting
the requirements of the SNPRM. A copy of the test report (VRTC-82-0267)
was placed in the public docket (No. 01-11-N09-54. See also Publication
No. DOT-HS-808-081).
For the purposes of the evaluation, NHTSA took the conservative
approach of modifying the most common conventional guard design and
developed a rear impact guard that was only slightly (10 percent)
stronger than the minimum requirements of the SNPRM when tested at the
vertical supports, which is the most significant location along the
width of the guard's horizontal member. NHTSA arrived at this
``minimally complying'' design through an iterative process of
fabrication and testing in accordance with the proposed compliance test
procedures.
These minimally compliant guards were then evaluated in two series
of full scale crash tests. The guards provided the proposed maximum
ground clearance 560 mm (22 in). For the initial series of crash tests,
the guards were mounted to a test fixture simulating the geometry of
the rear end of heavy trailers. The guards were mounted on a late model
production trailer for the other series. A total of seven crash tests
were conducted with the minimally complying guard design. The tests
were conducted at an impact speed of 48 kph (30 miles per hour (mph))
with late model compact and subcompact cars with mass between 1135 and
1590 kg (or weight between 2500-3200 lbs). In each category, vehicles
were selected which had low hood profiles, and were therefore most
likely to underride the 560 mm (22 in) guard height.
Four of the seven crash tests resulted in no PCI when the minimally
compliant guard was mounted flush with the rear extremity of the
trailer and simulated trailer. See Tables 6, 8, and 10 of the VRTC test
report. The hood of one passenger car was driven through the windshield
during one of these tests (Corsica 1, VRTC test report, page 26). The
magnitude of the passenger compartment intrusion by the hood was
marginal, however, and the test dummies were not contacted by the hood
during the collision. Two cases of PCI were caused by guard system
failure, one in a simulated trailer test and one in a production
trailer test (respectively, Saturn 1 in Table 8 and Corsica (trailer)
in Table 10 of VRTC test report). The guard system failure in the
simulated trailer test was due to attachment hardware failure. The
failure in the production trailer test was the

[[Page 2007]]
result of trailer structural failure at the guard attachment locations.
In each case, the guard attachment hardware and the trailer structure
were upgraded with simple, inexpensive materials for subsequent tests.
Retests with the modified hardware and trailer frame showed adequate
guard system performance.
All these crash tests included Hybrid III test dummies positioned
in the driver and outboard front passenger seating locations for each
crash test. The procedures used for frontal barrier crash test
preparation under FMVSS No. 208, Occupant Crash Protection, were
followed with respect to dummy positioning, restraint usage, and dummy
instrumentation. Dummy instrumentation indicated very low potential for
serious or fatal injury in all seven of the crash tests with the
minimally compliant guard, even those in which there was PCI.
The VRTC research project also performed a crash test using a very
strong, i.e., ``rigid,'' guard, to compare the amount of underride and
deceleration forces generated with those generated by the minimally
compliant guard. The 48 kph (30 mph) impact generated a peak force of
about 415 kN (93,000 lbs) and the guard sustained an insignificant
amount of permanent deformation. Although underride in this crash test
was minimal, occupant compartment forces generated during the crash
were significant, with on-board dummy readings indicating a potential
for serious driver chest injuries (dummy chest acceleration was 61 G,
slightly higher than the 60 G permitted in FMVSS No. 208, Occupant
Crash Protection). A similar crash test with the minimally compliant
guard was conducted with the same make and model passenger vehicle. The
minimally compliant guard, with a force resistance capability of about
200 kN (45,000 lbs), allowed more underride than the rigid guard and
marginal PCI. However, at 48 kph (30 mph), the minimally compliant
guard test generated occupant compartment forces low enough that they
posed essentially no potential for life-threatening occupant injuries.
This test further demonstrated the adequacy of the proposed guard
ground clearance requirement of 560 millimeters (mm) (22 in).

IX. Overview of the Final Rule

This rule establishes two Federal Motor Vehicle Safety Standards.
The two standards are being announced in this single notice because
they are complementary and because their substantive requirements both
derive from a single standard proposed in an earlier NPRM (Docket No.
1-11, notice 8). The first standard will be referred to as the
``equipment standard'' because it sets forth requirements that a rear
impact guard must meet as an item of motor vehicle equipment. The
second standard will be referred to as the ``vehicle standard'' because
it requires a new trailer or semitrailer to be equipped with a guard
that meets the equipment standard.
The equipment standard specifies the procedures that the agency
will use when testing a guard. The guard is first mounted to a rigid
test fixture or a secured trailer, in accordance with the installation
instructions which the guard manufacturer is required to provide. The
standard describes how to select three test locations across the width
of the guard. At these three locations, the testing procedure provides
that force be slowly applied until the guard has been deflected by 125
mm (5 in). The standard specifies procedures for determining whether
the tested guard has met the minimum requirements for strength and
energy absorption. Guards that can pass the strength and energy
absorption tests may be certified and labeled as complying with the
equipment standard and sold to vehicle manufacturers if accompanied by
the necessary attachment hardware and mounting instructions.
The guard mounting instructions are a crucial interface between the
equipment standard and the vehicle standard. NHTSA has modified the
equipment standard proposed in the SNPRM to require the guard
manufacturer's instructions to include (1) a description of the types
of structures to which attachment must be made, and (2) the manner in
which attachment must be made, in order for the guard to perform in its
designed fashion.
The vehicle standard requires that most new trailers and
semitrailers be equipped with a rear impact guard certified to the
equipment standard. The vehicle manufacturer can manufacture and
certify the guards according to the equipment standard, or simply
purchase and install certified guards from a guard manufacturer. The
vehicle standard requires that the guards extend laterally to within
100 mm (4 in) of the sides of the trailer, that the guard have a ground
clearance of no more than 560 mm (22 in), and that the guard be placed
as close to the rear of the vehicle as possible. To ensure that the
guard will perform properly, the vehicle standard further requires that
the guard be mounted on the trailer or semitrailer in accordance with
installation instructions provided by the guard manufacturer.
The vehicle standard lists and defines certain types of vehicles
that are excluded from the requirement to have rear impact guards.
Single unit (unarticulated) trucks, truck tractors, pole trailers, low
chassis vehicles, special purpose vehicles, and wheels back vehicles do
not have to have rear impact guards.

X. Summary of Changes From the 1992 SNPRM

The greatest change from the SNPRM is the addition to the equipment
standard of a requirement for energy absorption. The SNPRM would have
permitted fairly rigid guards because it did not require the guard to
yield in response to force. Rigid guards may stop the passenger vehicle
too quickly, causing occupant deaths and injuries from sudden
deceleration. To ensure that the guards will yield, this rule adds a
requirement that the guards absorb a certain amount of energy during
the strength test. The new requirement does not necessitate the use of
any additional new test equipment or the following of any additional
test procedures. It does require more frequent measurements of the load
during the strength test, and a few extra calculations after the test.
The test procedures in the equipment standard have been modified to
allow velocity-sensitive rear impact guards. Velocity-sensitive guards
would have failed the quasi-static strength test procedure proposed in
the SNPRM because these guards are designed to provide resistance that
is proportional to the displacement rate, and the test procedure
displaces the guard very slowly. The final rule provides for modifying
the guards to deactivate the energy absorbing components prior to the
strength test. Because velocity sensitive guards typically have
excellent energy absorption characteristics and because quasi-static
testing does not test their energy absorbing capabilities, velocity-
sensitive guards do not have to be tested for energy absorption. The
only type of velocity-sensitive guards that the agency is aware of use
hydraulic fluid properties to deform in a controlled manner. Therefore,
these ``hydraulic guards'' are the only ones excluded from the energy
absorption test.
The final rule requires greater specificity in statements regarding
trailer structure in the installation instructions provided by the
guard manufacturer. The SNPRM said only that the instructions had to
specify the types of vehicles for which the guard was intended, state
the necessity for

[[Page 2008]]
attaching the guard to the vehicle chassis, and explain how the
attachment hardware was to be used. The regulatory text of the final
rule makes it clear that the installation instructions must specify all
aspects of the trailer that are necessary to the proper functioning of
the guard. The test procedure has been modified to indirectly test the
adequacy of the attachment.
NHTSA has changed some of the guard configuration requirements in
the vehicle standard. The SNPRM proposed to require that the horizontal
member of the guard extend to within 100 mm (4 in) of the side
extremities of the vehicle and to within 305 mm (12 in) of the rear
extremities. These requirements have been modified to allow rounded
guard ends. The final rule allows an extra six inches in these
dimensions only for the portion of a guard that is curved. Using
rounded guard ends will diminish the hooking potential of the guards
when the trailer is turning sharply. Guard ends that are rounded upward
and attached to the vehicle may add strength to the horizontal member
near the side extremity of the vehicle.
To account for high, overhanging rear protrusions on trailers,
NHTSA changed the definition of the vertical zone to be considered when
determining the trailer's ``rear extremity.'' Determination of the
``rear extremity'' is important because the location of the guard is
based on the location of the rear extremity. The SNPRM defined ``rear
extremity'' as the rearmost point above 560 mm (22 in) from the ground.
Since high overhangs pose no risk to colliding passenger vehicles,
NHTSA has set a maximum height of 1905 mm (75 in) from the ground on
the zone. Higher protrusions will not be considered as the rear
extremity.
Another change in the configuration requirements is that the final
rule requires the guard to be mounted as close to the rear extremity as
practical within the 305 mm (12 in) zone forward of the rear extremity.
The SNPRM did not regulate where in the zone the guard had to be
mounted.

XI. Analysis and Response to Comments on the 1992 SNPRM

A. Separate Equipment and Vehicle Standards

Companies such as Waltco and industry groups such as the National
Truck Equipment Association supported the separate equipment and
vehicle standards as a method to prevent undue testing burdens.
One of the concerns raised by consumer interest organizations is
that allowing the guards to be tested on a ``non-vehicle'' rigid test
fixture posed a problem if it is done in the expectation that the
guards would necessarily perform in a similar manner once they are
installed on vehicles. The Institute for Injury Reduction (IIR)
commented that neither the equipment standard nor the vehicle standard
specifies or regulates the interface between the guard and the vehicle.
Therefore, IIR was concerned that there are no ``real-world'' tests
performed on the guards as installed on the vehicle and suggested that
it is unclear whether a failure of such a test would represent
noncompliance by the guard manufacturer, the vehicle manufacturer,
both, or neither.
NHTSA agrees that an underlying assumption of this regulatory
scheme is that the guards would perform in the real world in a manner
similar to the way they do in the tests. This assumption is supported
by the results of the VRTC research project, which show that the
maximum force measured in quasi-static tests is similar to the maximum
force generated in dynamic crash tests. Moreover, this regulatory
scheme has worked well for tires, which also have separate equipment
(49 CFR 571.109) and vehicle (Sec. 571.110) standards.
NHTSA disagrees with IIR's argument that separate guard and vehicle
standards leave the guard/vehicle interface unregulated. The vehicle
standard specifies that the guard be attached in accordance with the
guard manufacturer's installation instructions, the same instructions
used to attach the guard to the test fixture during agency compliance
testing under the equipment standard.
When writing installation instructions, the guard manufacturer must
take into account the possibility of inadequate trailer structure to
support the guard. Depending on the guard design, the guard
manufacturer may want to specify in the instructions that the guard
cannot be attached to certain structures (e.g., floorboards) and that
it must be attached to other surfaces, for example, frame rails with a
horizontal surface and specified wall thickness of a certain material
(e.g., hardened steel). The guard manufacturer may have to specify
local reinforcement if the trailer chassis is inadequate to pass the
compliance test with the chassis surface mounted on the rigid test
fixture.
The installation instructions must be appropriate to the trailer
design, so that the vehicle manufacturer knows which guard to purchase
and does not have to deviate from the instructions to install the
guard. To help assure this, the regulatory text has been modified to
make it clear that the guard manufacturer must either list appropriate
trailers or specify in the installation instructions all attributes
that make a trailer suitable for the proper installation and
functioning of the guard. These include the types of trailer
structures, design types with dimensions, materials thickness and tire
track widths that are appropriate as an installation location.
NHTSA will install the guards during compliance testing based on
these instructions. Therefore, it is essential that the attachment site
and attachment method be adequately specified. This is especially
important to avoid failure of the attachment itself during the test.
In a VRTC test of the minimally complying guard mounted on a
typical trailer, the trailer frame rails worked with the guard by
bending/deforming to absorb the colliding vehicle's crash energy.
However, the attachment site on the frame rails had to be strengthened
with an inexpensive local reinforcement.
IIR's argument that failure during compliance testing would leave
the identity of the non-complying party in doubt is incorrect. The only
testing procedures in NHTSA's rule are the compliance tests in the
equipment standard. Therefore, the only party that can be responsible
for a testing failure is the guard manufacturer. Noncompliance by the
vehicle manufacturer may be established by inspecting the vehicle and
observing improperly installed guards, such as during an FHWA heavy
truck inspection. If the vehicle manufacturer manufactures the guard
which it uses, as NHTSA believes will usually be the case, there will
be no ambiguity as to the party responsible for testing failure or
improper installation.

B. Standard for Equipment

1. Relationship of Strength, Energy Absorption, and PCI
In specifying performance standards for rear impact guards, the
agency must balance various performance attributes. The vast majority
of the commenters, including virtually all of the consumer safety
groups, asserted that underride guards should be strong, yet energy
absorbing. NHTSA agrees that these are both desirable properties in an
underride guard, but emphasizes that an increase in strength may result
in a decrease in the capability of the guard to absorb energy, and vice
versa. An impact guard strong enough to restrain a large car travelling
at high speeds would impart high deceleration forces to a small car
crashing into it at the

[[Page 2009]]
same speed. Conversely, an impact guard that is optimized to restrain a
small car without excessive deceleration forces might fail (i.e.,
deform so much that it allows PCI) if a large car crashes into it, or
if a small car crashes into it at higher speeds.
Energy absorption must also be balanced against PCI prevention.
Energy absorption may be maximized by allowing the guard to yield for a
greater distance before bringing the passenger car to a stop. However,
the more the guard yields, the farther the colliding vehicle travels
and the greater the likelihood of PCI. This rulemaking has focussed on
balancing the need for PCI-prevention against minimizing crash
injuries. FARS data show a strong correlation between PCI and
fatalities or serious injuries. Preventing PCI demands a guard that is
strong enough to prevent the passenger vehicle from advancing very far
after contact with the guard.
Compounding the difficulty of balancing the guard's performance
attributes is the wide range of colliding passenger vehicle weight,
speed, and size. The combination of weight and speed determines the
level of kinetic energy to which the guard will be subjected. Passenger
vehicle weight generally correlates with the hood height and length,
which determines how far the vehicle can proceed after contact with the
guard before PCI occurs. Fortunately, these factors offset one another
for large cars (i.e., the greater weight promotes greater amounts of
underride, while the higher hood profile results in better guard
engagement and the longer hood allows for more underride before
experiencing PCI).
Small pickups and vans have relatively high profiles but a
relatively short distance from the front of the vehicle to the occupant
compartment. A guard would have to yield only slightly, or have high
strength to prevent minivans and some pickups (which typically have a
mass more than 1810 kg or weigh more than 4,000 lbs and have short
hoods) from experiencing PCI. Because the passenger compartment is so
close to the front of a heavy standard van, no underride guard is
likely to be very effective in preventing PCI for these vehicles.
Nevertheless, some reduction in fatalities and non-fatal injuries can
be expected due to the initial energy absorption of the guard.
Fortunately, vans have only been involved in 0.5 percent of all
underride fatalities from 1982 to 1992. Pickups have been involved in
about 18 percent of the fatalities during this period.
It should be recognized, therefore, that impact guards cannot be
optimized for all situations. The requirements in this rule should
reduce the incidence of PCI, fatalities, and injuries for all passenger
vehicles, but some more than others. A minimally compliant guard should
protect all passenger vehicles from PCI and excessive deceleration
forces up to some speed in the 40 kph (25 mph) to 56 kph (35 mph)
range, although that speed will vary on a sliding scale depending on
the vehicle weight and front end profile. For example, NHTSA
analytically estimates that mid and full size cars and light trucks and
vans with a mass greater than 1590 kg (3,500 lbs) will experience PCI
at approximately 43 kph (27 mph), while mini-compacts of less than 1135
kg (2,500 lbs) will be able to collide with the required guard at about
61 kph (38 mph) without PCI. This estimate is obtained by equating the
energy absorbed by a 48 kph (30 mph) collision of a 1590 kg (3,500 lb)
vehicle rigid barrier crash to the energy absorbed by a different
weight vehicle). For example, for a 907 kg (2000 lb) vehicle, the
calculated impact speed without PCI is: (square root of (1,590 kg/907
kg)) x 48 kph=63.5 kph, or (square root of (3,500 lb/2,000 lb)) x 30
mph=39.7 mph.
2. Guard Strength
Several consumer interest organizations and private citizens
criticized the 1992 SNPRM's proposed guard strength requirements. These
commenters' objections are either that guards meeting the requirements
would be too weak to prevent underride or that they would be so strong
that the passenger vehicle would be subjected to excessive deceleration
forces. As explained above, the issues of strength and energy
absorption are closely related. However, issues relating primarily to
energy absorption will be addressed in the next section.
The SNPRM, which was premised upon underride protection being
provided by a horizontal member, proposed to require that the
horizontal member resist a force of 50 kilonewtons (kN) (11,240 lbs)
applied at the center (site P2) and near the outboard ends (sites P1),
and a force of 100 kN (22,480 lbs) at an intermediate position (sites
P3), in separate quasi-static strength tests. For these tests, guard
resistance at the specified force level would have to occur at less
than or equal to a 125 mm (5 in) displacement of the guard's horizontal
member.
Several commenters stated that overly ``rigid'' or non-yielding
guards would be permitted by the proposed rule. They expressed concern
that those guards would be too stiff, citing the results of full-scale,
heavy truck rear underride crash tests conducted in the late 1970's and
early 1980's by the Texas Transportation Institute (TTI), Dynamic
Sciences, Inc., and the Insurance Institute for Highway Safety (IIHS).
These crash tests indicated occupant compartment forces generated in
collisions with rigid guards at impact speeds above 48 kph (30 mph)
could produce potentially fatal driver and front passenger head and
chest injuries.
Advocates for Highway and Auto Safety (Advocates) stated that the
proposed guard would not perform as well as the agency expects, and
would be excessively deformed or fail in impacts not much above 40 to
48 kph (25 to 30 mph). Advocates further stated that NHTSA directed its
contracted researcher in 1982 to reduce the impact speed of a dynamic
crash test on a Chevrolet Impala from 48 kph (30 mph) to 40 kph (25
mph), specifically to ensure that excessive underride did not occur.
The actual speed of the tested 1,840 kg (4,060 lb) Chevrolet Impala was
38.5 kph (23.9 mph). Advocates contends that the agency admitted in a
memorandum from Mr. Tomassoni (who worked for NHTSA at the time) that
the test would have resulted in PCI at 48 kph (30 mph). IIHS also
included these criticisms in its comment.
Some commenters recommended that NHTSA require specific levels of
strength higher than those proposed in the SNPRM. Advocates attached a
1991 technical paper by Mr. G. Rechnitzer, of Monash University in
Australia, which reviewed European truck underride data. The example
with the widest application, the Economic Commission for Europe's (ECE)
Regulation No. 58 for heavy truck rear underride guards, currently
requires a guard force resistance of 100 kN (22,480 lbs) at the point
on the guard corresponding with this rule's P3 test point, 50 kN
(11,240 lbs) at the center, and up to 25 kN (5,620 lbs) at the outboard
test position corresponding with this rule's P1 position. Mr.
Rechnitzer recommended that the rear impact guard strength requirements
be upgraded to 150 kN (33,370 lbs) at the P3 location and 100 kN
(22,480 lbs) at the center and P1 locations. Mr. Byron Bloch, of Auto
Safety Design, suggested an even stronger guard. He thought the rule
should require that the guard resist 222 kN (50,000 lbs) at the P3 test
location, where the SNPRM requires that the guard resist a force of 100
kN (22,480 lbs).
The VRTC tests indicate that the strength of the 1992 SNPRM guard
is adequate for preventing underride with

[[Page 2010]]
PCI in a collision with an impact speed of up to 48 kph (30 mph) for
vehicles with a mass of about 1,450 kg (3,200 lbs). PCI resistance
would be expected at higher impact speeds for lighter vehicles and
lower impact speeds for heavier vehicles. The test data also indicate
that rear impact guards having somewhat more strength than the proposed
level of strength could resist PCI at higher impact speeds without
generating life-threatening passenger compartment force levels.
Although stronger guard strengths may be desirable, the agency cannot
quantify the increased benefits that might be obtained without further
testing.
Based on the VRTC tests, the agency believes that the guard
strength requirements proposed in the 1992 SNPRM are of sufficient
magnitude to prevent PCI for most late model passenger vehicles at
impact speeds of about 45 kph (28 mph). This rule has an additional
requirement that guards yield enough to maintain survivable levels of
occupant compartment deceleration when impacted by passenger vehicles.
Therefore, the agency has decided to retain the strength requirements
of the SNPRM in the final rule.
The IIHS advocated a specific guard design, which it said was
preferable for strength purposes. That organization believes a diagonal
strut from the horizontal member of the guard to the trailer chassis
could augment guard strength without a large increase in guard weight.
NHTSA agrees with the IIHS that this type of design is quite efficient
with respect to weight and strength, though not necessarily with
respect to energy absorption. However, the agency does not believe that
it is necessary or desirable to mandate a specific design, since
similar crash performance may be achieved with other designs.
3. Guard Energy Absorption
Although all non-rigid guards absorb some of the kinetic energy of
the striking vehicle, there was considerable concern that the SNPRM did
not require energy absorbing guards. The consumer interest
organizations and about 2,200 private citizens urged NHTSA to mandate
``energy absorbing'' guards. By deforming, rear impact guard structures
absorb some of the kinetic energy of the striking vehicle. The more
energy the guard absorbs, the less energy must be absorbed by
deformation of the striking vehicle before it stops. Commenters were
concerned that the SNPRM would have permitted rigid guard designs that
would impart high levels of crash forces to the striking vehicle's
occupants.
As used by the consumer interest groups, the term ``energy
absorbing guards'' generally refers to guards whose vertical support
members are designed to pivot about their attachment braces at the
vehicle chassis. These guards absorb energy by means such as
cylindrical, telescoping hydraulic or plastic struts, which are also
attached to the guard's horizontal member and the vehicle chassis. When
impacted, these energy absorbing units respond by compressing without
substantial deformation until the units have reached their maximum
deflection, or ``bottomed out.'' On the other hand, the primary energy
absorbing mechanism of a fixed guard, such as the design used in the
VRTC tests, is the flexing and bending of the guard's vertical
supports. Keeping this in mind, the agency uses the term ``energy
absorbing guards'' below in the same sense as used by the commenters,
as a shorthand way of referring to guard designs with special energy
absorbing design features.
Advocates recommended that guards be required to be energy
absorbing so that 64 kph (40 mph) impacts of small cars with the rear
of heavy vehicles are survivable through the combined energy absorption
of the car and the guard. The National Association of Independent
Insurers (NAII) suggested that the proposed rule be modified to require
a more flexible, energy absorbing guard. Citizens for Reliable and Safe
Highways (CRASH) stated that the agency fails to acknowledge the need
for and potential benefits from improved, slightly more expensive,
energy absorbing guards that are in use in Europe.
To ensure that the guard will provide the combination of strength
and energy absorption necessary to prevent underride with PCI at a
specified impact speed, as recommended by Advocates, a full-scale
dynamic compliance test including a passenger vehicle would be
necessary. VRTC conducted full-scale crash tests with guards that were
also tested in accordance with the SNPRM compliance procedures. These
tests demonstrated that the proposed quasi-static compliance test is
adequate for determining guard strength. The peak forces generated by
the guard in the quasi-static compliance tests and the full-scale crash
tests were approximately the same. Guard strength or peak force
capability is the primary factor in underride prevention. Guard energy
absorption characteristics determine the guard's ability to maintain
impact forces at survivable levels in the striking vehicle, as well as
the guard's resistance to structural failure.
The agency has decided to retain the quasi-static compliance test
for guard strength due to the greater complexity and cost of a dynamic
compliance test procedure. Although the guard's ability to resist PCI
at a specific impact speed will not be tested directly, the VRTC tests
show that dynamic guard performance can be accurately estimated from
the quasi-static compliance test results. Therefore, it is not
necessary to conduct expensive full-scale dynamic tests to attain most
of the benefits of dynamic testing.
Advocates also stated that British researchers assess the potential
fatality reduction effectiveness of stronger, energy absorbing guards
at 25 to 35 percent. This is about twice the current guard
effectiveness in Europe, according to the document cited by Advocates,
an opinion paper by P.F. Gloyns, et al., of Vehicle Safety Consultants,
Ltd., entitled ``Legislative Implications of Accident Experience in the
UK of Rear Under-Run Guards.'' The Gloyns paper does not quantify the
increase in guard strength or the magnitude of guard energy absorption
required to achieve the estimated increase in guard effectiveness. The
agency acknowledges that various combinations of guard strength and
energy absorption capability could increase the effectiveness of rear
impact guards. However, without more quantitative information, NHTSA
cannot address the guard effectiveness claims of Gloynes, et al.
It may be that energy absorbing rear underride guards, which were
referred to by CRASH and which are currently in use on one to two
percent of vehicles in Europe, are superior to a moderate strength,
fixed guard meeting the minimum performance requirements specified in
the rulemaking proposal. The agency notes that these European guards,
or guards with similar energy absorbing characteristics and design
features, would not be prohibited by NHTSA's proposed rule and will no
doubt be considered by the industry as a possible means of compliance,
just as they were in Europe.
The agency has tested one guard, the Quinton-Hazel rear impact
guard, which utilized pivoting vertical support members along with
telescoping hydraulic struts and coil springs. The guard demonstrated
excellent overall performance in a crash test conducted in 1979 by the
Texas Transportation Institute. The striking crash test vehicle was a
1,810 kg (4,000 lb) Chevrolet and the impact speed was 56 kph (35 mph).
The collision did not result in PCI, and all measured occupant
responses indicated that the potential for driver

[[Page 2011]]
and front passenger serious injuries was low. It is estimated that
similar guards would weigh about 1.33 to 3 times more and cost 3 times
more than a fixed, moderate strength guard designed to meet the
requirements of the SNPRM. In other words, it would cost $300-$350 and
have a mass of 136 kg (300 lbs) to 181 kg (400 lbs). Further, hydraulic
energy absorbing guards would be considerably more complex than fixed
guards that comply minimally with this rulemaking, and would require
periodic maintenance. It is NHTSA's understanding that there are
currently no guards in production in this country or in Europe that
utilize hydraulic or plastic energy absorbing, telescoping units. A
letter from one of the former manufacturers, Quinton-Hazel, indicates
that the market probably rejected them as too costly.
Nevertheless, in response to the comments recommending energy
absorbing guards, the agency has added a performance requirement for
guard energy absorption to the rule. The requirement does not include
design specifications such as pivoting vertical supports or telescoping
energy absorbing units. The agency is requiring that each guard absorb
a minimum amount of energy based on the forces and displacements
specified in the 1992 SNPRM. The same quasi-static compliance test
procedure proposed for strength testing will be used to determine
compliance with this new specification. The test for guard energy
absorption will be conducted only at the P3 location used for guard
strength testing. The minimum magnitude of guard energy absorption will
be 5,650 joules (4,170 foot-pounds), which is based on the force
required to comply with the strength test at the P3 test location and
the maximum displacement allowed for the guard to generate the force
(125 mm, or 5 in). The energy absorption test will require that the
guard's horizontal member undergo 125 mm (5 in) of displacement while
the force generated by the guard is recorded at least ten times per 25
mm. The magnitude of guard energy absorption at the P3 location is
sufficient to absorb about 12 percent of the total kinetic energy of a
48 kph (30 mph) centric collision with a 1,135 kg (2500 lb) vehicle.
This magnitude of guard energy absorption capability is also similar to
the amount recommended in several British research papers provided by
Advocates.
Several commenters, including consumer interest organizations and
trailer manufacturers, stated that the proposed rule would permit
overly ``rigid'' or non-yielding guards that would absorb little or no
crash energy. The commenters expressed concern that those guards would
be too stiff and would result in fatal driver and front vehicle
passenger head and chest injuries.
The agency has drafted the energy absorption requirement to address
these concerns. NHTSA recognizes the potential trade-off between
designs of underride guards that minimize occupant injury criteria
responses and those that provide the most protection from PCI. The
agency also recognizes that an increase in the level of rigidity from
the minimally compliant guard used in the VRTC tests is desirable, but
this should not be at the expense of energy absorption. On the other
hand, the agency does not want to restrict or dictate guard design by
specifying the rigidity of the guard. Therefore, to discourage overly
rigid guards, this rule requires that a minimum amount of the energy be
absorbed during the energy absorption test from permanent yielding, or
plastic deformation, of the guard. After the guard has reached the full
125 mm (5 in) of deformation, the load is reduced and any elastic
``rebound'' of the guard is measured until the load is zero. The
elastic component of the energy that is returned by the guard is not
included in the calculation of total energy absorbed by the guard. This
method gives guard designers flexibility to select guard material
properties and frame member spatial configuration.
Some commenters observed that the test procedures proposed in the
SNPRM precluded the use of hydraulic energy absorbing guards. Mr. John
Tomassoni stated that the 125 mm (5 in) displacement maximum allowed in
the strength test would allow only passive structures such as steel
struts designed to bend on impact. This is because active energy
absorbing struts that are hydraulic (analogous to a vehicle shock
absorber) are velocity sensitive. With the slow application of force
during the quasi-static test, the hydraulic fluid units would develop
almost no resistance. He recommended adding a ``bottoming'' provision
to allow static testing after hydraulic systems have reached full
stroke.
NHTSA agrees that quasi-static test procedures are inappropriate
for hydraulic guards, or any other type of velocity sensitive guard
(although NHTSA is unaware of any non-hydraulic guards that are
velocity sensitive). A dynamic test would be required to assess their
energy-absorbing capabilities by supplying the sudden onset of force
their energy absorbing units require to generate resistance. Because
the agency does not want to discourage the use of these advanced guard
designs by requiring expensive dynamic tests, and because these guards
typically have excellent energy absorbing capabilities, the final rule
excludes these guards from the energy absorption requirements.
There are also problems with subjecting velocity sensitive guards
to the strength requirement. However, complete exclusion of those
guards from the performance requirements would be inappropriate.
Accordingly, the agency has modified the test procedures to allow
velocity sensitive guards to be tested for compliance with the strength
requirement. The agency is concerned that, if the hydraulic energy
absorbing units do not operate properly, the guard will not generate
significant resistance and energy absorption. NHTSA wants to assure
that the guard has enough residual strength, even without the energy
absorbing units, to meet the same strength requirements as other
guards. Therefore, velocity sensitive energy absorbing guards will be
tested by slowly compressing the energy absorbing units to the full
extent of their designed travel or 610 mm (24 in), whichever occurs
first. This will allow the frame of the guard itself to generate
resistance, rather than having the piston simply compress the hydraulic
shock absorbers.
4. Vertical Cross-sectional Height of Horizontal Cross-member
The SNPRM proposed a minimum vertical cross sectional height of 100
mm (4 in) across the entire width of the guard's horizontal cross-
member. Advocates stated in its comment that the guard must be at least
205 mm (8 in), and preferably 305 mm (12 in), high to better manage the
loading impact forces and assure full engagement of the vehicle front
end. In contrast, the Truck Trailer Manufacturers Association (TTMA)
suggested reducing the requirement, urging that the guard be only 50 mm
(2 in) high because that is all that is required for adequate strength.
It asserted that requiring greater vertical cross section height just
adds unnecessary weight and cost to the guards.
NHTSA agrees with Advocates' position that a higher vertical cross
section has the potential to better distribute the impact forces, but
this does not mean that the proposed 100 mm (4 in) height is
insufficient. The 100 mm (4 in) height would be inferior if it sheared
or ``cut'' through the front of the striking vehicle, thus allowing
forward vehicle motion without much energy absorption due to the low
magnitude of

[[Page 2012]]
force generated by the guard. A guard should cause the vehicle to
absorb energy by crushing, rather than shearing through, frontal
vehicle structural components. Shearing through did not occur in the
agency's testing with a 100 mm (4 in) high guard horizontal member.
None of the crash tests conducted pursuant to this rulemaking resulted
in significant shearing of the passenger vehicle's frontal structure
(above the 560 mm (22 in) high guard). The crash tests show that the
100 mm (4 in) profile of the guard horizontal member resulted in
adequate engagement of the car's front end and is harmonized with the
guard specified in ECE Regulation 58. Moreover, a 205 mm (8 in) high
profile may require heavier and more expensive guards. Finally, the
agency notes that 100 mm (4 in) is only a minimum height, so guard
manufacturers are free to manufacture the guards that Advocates
recommends. Accordingly, the agency concludes that a higher vertical
cross sectional height requirement is unnecessary.
NHTSA also disagrees with TTMA's position that a 50 mm (2 in)
vertical cross sectional height would be appropriate. The TTMA did not
provide any data to support its assertion that the strength should be
adequate. Even if the 50 mm (2 in) height were sufficient for strength
purposes, it would have a greater tendency to shear into the front of
the passenger vehicle instead of crushing it. This would result in a
reduction of energy absorption by the guard and an increase of the
striking vehicle damage in low speed crashes of 16 to 24 kph (10 to 15
mph). Accordingly, the agency has decided to retain the 100 mm (4 in)
cross sectional vertical height requirement in the final rule.
5. Shape of the Horizontal Cross-member
Some commenters stated that NHTSA should require the guards to have
blunted or rounded ends. The Florida Department of Transportation,
based on visual evaluations of the installed guard, stated that the
requirement that the guard extend to within 100 mm (4 in) of the side
of the vehicle would make it a dangerous ``hook'' for adjacent
vehicles, especially during sharp turns of the trailer. It suggested
requiring a ``U'' shaped guard, similar to one used by some carriers
which is attached at either end to the underside or rear of the
vehicle. It thought that the ends on these guards could be located
further inboard. The TTMA had a similar suggestion, proposing that
NHTSA allow (but not require) guards with rounded corners, to lessen
the hooking potential when the sliding tandem is positioned forward.
The TTMA suggested that the rule be modified to allow such guards to
begin curving at a point 255 mm (10 in) inboard of the edges of the
vehicle, while retaining the 100 mm (4 in) requirement for straight
guards.
NHTSA agrees that there is some potential for hooking the guard on
the fenders and wheel wells of adjacent passenger vehicles when the
rear end of the trailer swings out laterally during a sharp turn. This
phenomenon would be accentuated when the rear wheels on a sliding
tandem are positioned forward. The rear wheels are generally positioned
forward to give the trailer greater maneuverability, so it is likely
that trailers in this configuration will be making sharp turns.
On the other hand, rounded or U-shaped guards would be more
expensive to manufacture and would weigh more. Moreover, rounded
corners offer very limited potential added value on roadways where
sharp turns are infrequent, such as on the interstate highways, which
are heavily traveled by trailers. Therefore, while the agency wants to
allow guards with rounded ends for operations where they are desired,
NHTSA does not think it is necessary or even appropriate to require
them.
The commenters referred to rounded guard ends that curve upward,
but a rounded end that curves forward could also be useful. It would
serve the purpose of making hooking less likely because the guard end
would sweep through a smaller arc and present a less pointed profile to
adjacent passenger vehicles. Moreover, forward-curving guards could
slightly enhance guard effectiveness if a passenger vehicle strikes the
trailer in the rear corner at an angle. However, forward-curving guard
ends might interfere with the rear wheels if a sliding tandem were
moved to the rearmost position.
NHTSA notes that the SNPRM would not prohibit guards with rounded
ends, but its configuration requirements would have restricted their
curves to a 100 mm (4 in) radius of curvature. To minimize hooking
potential and property damage in some applications, the final rule
adopts the TTMA's suggestion and allows a guard with rounded ends to
begin curving 255 mm (10 in) inboard of the side extremity of the
trailer. This will allow a radius of curvature of 150 mm (6 in), or 255
mm (10 in) if the guard end extends all the way to the side
extremities. To make the same allowances for forward-curving guards,
should guard manufacturers want to produce them, NHTSA is allowing
those guards to begin curving forward 255 mm (10 in) inboard of the
side extremities, even if the guards are already mounted as far forward
as possible--305 mm (12 in) forward of the rear extremity.
6. Guard Attachment
The SNPRM did not specify a particular guard attachment method. To
assure an adequate interface between the guard and the trailer, the
SNPRM proposed to require that the guard be attached to the trailer
chassis in accordance with the instructions provided by the guard
manufacturer.
Several commenters thought the SNPRM inadequately addressed the
issue of guard attachment and discussed the merits of certain guard
designs. Citing a study by Vehicle Safety Consultants (VSC) Ltd.,
Advocates stated that attaching the horizontal member of the guard to
the vehicle with vertical members is not ideal because the guard tends
to pivot forward and up if it is struck from the rear by a passenger
vehicle and fails. It said that the vertical members then form an
inverse ramp, thus aggravating any underride tendency by pushing the
passenger vehicle down and the trailer up. To solve this problem,
Advocates appears to recommend either guards with diagonal hydraulic
struts or the use of hinged, pivoting energy absorbing guards that can
fold up for rail or other intermodal transportation. IIHS also believed
a diagonal strut would improve guard strength without adding weight and
would make it more likely that the guard will move downward as it
deforms, thus helping to stop the passenger vehicle.
The agency agrees with IIHS and Advocates that designs employing
diagonal struts are strong yet light, but believes it would be
inappropriate to require such designs. There is no evidence that only
designs with diagonal struts perform adequately. To the contrary, the
design used in the VRTC tests did not have diagonal struts and
performed acceptably. Diagonal struts may also be impracticable in some
cases, due to trailer construction and use.
Likewise, while the pivoting, fold-away design that Advocates
recommended has obvious practical advantages in some circumstances, the
agency does not believe that there is any necessity for mandating that
all guards incorporate that design. Such designs would be unneeded by
many trailer operators since most trailers do not travel by ship or
train. If trailer operators need fold-away guards for intermodal
transportation or other

[[Page 2013]]
operational environments, they may specify such guards when ordering
new trailers.
NHTSA believes that specifying a particular attachment
configuration, as suggested by Advocates and IIHS, would unnecessarily
restrict design flexibility on the part of guard manufacturers.
Adequate performance may be achieved by a variety of attachment
methods. Moreover, it is impracticable for NHTSA to attempt to
anticipate all the factors that may go into the choice of attachment
method, given the variety of possible guard and trailer configurations.
The agency's decision not to specify a particular attachment method
leaves the guard manufacturers free to choose an appropriate design.
Some commenters had conflicting impressions that the SNPRM required
a particular attachment method. Transamerica Leasing interprets the
SNPRM's reference to ``attachment hardware'' as meaning that the
proposed rule contemplates only bolt-on guards. It thinks that guards
that are welded on should also be allowed. In contrast, Advocates
suggested that the SNPRM requires guards with vertical supports for the
horizontal member and welded steel construction.
No specific attachment method was proposed in the SNPRM. Nothing in
the SNPRM nor in this final rule requires vertical supports or welded
construction. Similarly, the agency did not intend its references to
attachment hardware in the SNPRM to imply that only bolt-on guards are
permitted. The agency's intent was to require that any necessary
attachment hardware be included with the guard when a guard
manufacturer sells the guard to a trailer manufacturer if the guard
manufacturer's method of attachment involves attachment hardware, as in
the case of bolt-on guards. Weld-on guards are also permitted. However,
if the guard manufacturer's installation instructions do not adequately
specify the welding procedures, welds of poor quality could break in
NHTSA's compliance testing. Weld strength could probably be assured
through incorporating by reference welding industry standard practices.
Some commenters believed that the guard-trailer interface was
inadequately addressed by the SNPRM. IIHS noted that the SNPRM proposed
no minimum strength for the chassis or the attachment method, and
concluded that the attachment may fail before the guard. It stated that
NHTSA's static tests showed that the trailer frame rails failed without
a doubler plate and that, even with a doubler plate, the flange welds
failed in dynamic tests. It also believed that NHTSA should require
installation instructions that are specific to each make and model of
trailer. IIHS reiterated these comments in a September 16, 1994 letter
that pointed to failures of the guard attachment hardware and trailer
structures resulting in PCI in two of the VRTC crash tests. IIHS urged
NHTSA to either require minimum strength levels for the guard
attachment hardware and frame rail or require that the guard be tested
together with the type of trailer frame rail to which it would be
attached.
Mr. John Tomassoni suggested that the preamble to this rule should
encourage manufacturers to install guards with due care so that the
attachment is as good as the guard. He said that the trailer frame is
the ``weak link'' in crashes today, and that adding ``doubler plates''
to trailer frame members helps to maintain the integrity of the
attachment in a crash.
NHTSA's test results show the importance of considering the
strength of the attachment point when designing a guard. The agency
does not at this time believe that it is necessary to define strength
requirements for the chassis or the attachment hardware because the
necessary strength is dependent on the design of the guard. For
example, a guard that is attached to the rear of the frame rail with
two vertical supports (i.e., the commonly used cantilever design used
in the VRTC tests and on most trailers) would require a stronger
attachment site and attachment hardware than a guard with many
attachment points or with diagonal struts. Therefore, without knowing
the design of the guard, NHTSA cannot readily specify minimum strengths
for the trailer frame or the attachment hardware, as suggested by IIHS.
However, the guard manufacturer must consider frame and hardware
strength in order to have a basis for certifying the guard for use on
the types of vehicles specified in the installation instructions. NHTSA
agrees with Mr. Tomassoni that, if a cantilever design is used, guard
manufacturers should consider doubler plates or other appropriate frame
reinforcement to prevent frame failure. NHTSA does not want to require
such features, however, because a different attachment design or a
sturdier trailer frame may eliminate the need for reinforcement. It is
not a requirement of this rule that guard manufacturers specify frame
strength or reinforcement procedures in the installation instructions.
However, as a practical matter, to have a basis for certification, they
must consider frame strength using testing, engineering analysis, or
both, to be assured that the guard attachment is appropriate for the
types of vehicles specified in those instructions.
The VRTC test experience illustrates why guard manufacturers should
appropriately design the strength of the attachment. In one case,
attachment bolts which were marginally weaker than those used in the
quasi-static test sheared under the sudden onset of force in the
dynamic test. In another case, the proximity of the guard to the rear
edge of the frame rail resulted in tearing of the trailer frame rail
webbing. In each case, the guard itself was not really exercised
because the attachment failed. In each case, simple modifications
solved the problem. The importance of careful attachment hardware
material selection and attachment design cannot be overemphasized.
Although guard manufacturers are free to issue separate
instructions for each specific make and model of trailer, as IIHS
recommends, it is not necessary for NHTSA to require such instructions.
An efficient way to specify trailer type would be to list specific
make/model combinations. However, as long as the instructions are
adequate to identify which vehicles are appropriate for the
installation of the guard, specification of the make and model of the
trailer may not be necessary. One reasonable alternative for a guard
manufacturer with a very adaptable guard design is to show in its
instructions the types of trailer, types of chassis configurations, and
frame strengths that are necessary to the functioning of that
particular guard. For example, the guard manufacturer might specify
that any flatbed or van trailer with longitudinal frame rails extending
to within 305 mm (12 in) of the rear, spaced between 760 mm and 1,270
mm (30 and 50 in) apart, and with the bottom of the frame rails
configured as a horizontal surface at least 100 mm (4 in) wide,
composed of steel that is at least 6 mm (1/4 of an inch) thick, would
be an appropriate trailer for mounting the guard.
Some commenters believed that defining ``chassis'' as the ``load
supporting structure of a motor vehicle'' was too restrictive or
otherwise inadequate. NSWMA asked NHTSA to modify S5.3.2 of the vehicle
standard to allow vehicle manufacturers with ``unique design
considerations'' to attach the underride guard ``to a load supporting
structure of the vehicle or body, or through other means that provide
equivalent protection.'' It believed that this change is necessary to
take into account body designs that do not use a conventional chassis
frame.

[[Page 2014]]
Mr. John Tomassoni also suggested that NHTSA further define the term
``load supporting structure'' because the longitudinal frame members
don't extend all the way to the rear end of some trailers.
Although NSWMA did not provide any specifics on its vehicles, NHTSA
agrees that there may be some trailers that do not have adequate
chassis structure, in terms of a frame structure, to support a
conventionally designed rear impact guard. However, no change to the
requirements is necessary. Although the frame components are the
obvious attachment point in the case of most trailers, attachment to
this chassis member is not required by this rule. In certain cases, an
unconventional guard design that is attached to other parts of the
chassis may be necessary. In rare cases, custom-designed guards or even
extension of the trailer chassis may be necessary to mount the guard.
The TTMA suggested changing the installation requirements in S5.3
to apply to ``guards that are produced or modified and installed by a
vehicle manufacturer * * *,'' so that a trailer manufacturer can modify
stock guards to fit its particular trailers. It assumes that the guard
manufacturer is unlikely to provide installation instructions for the
wide variety of trailer configurations. It reasons that, since the
trailer manufacturer has to certify that the trailer is in compliance
with all Federal motor vehicle safety standards anyway, why not let it
modify the guard?
Vehicle manufacturers are allowed to modify purchased guards to
suit their own trailers. There may be minor modifications to widely
available guard designs that will make them suitable for trailers for
which they were not designed. However, if a vehicle manufacturer
modifies the guard in a way not contemplated by the instructions
provided by the guard manufacturer, that vehicle manufacturer becomes a
guard manufacturer. The vehicle manufacturer may no longer rely on the
certification of the original guard manufacturer, because the original
manufacturer presumably did not intend its guards to be so modified. As
a guard manufacturer, the vehicle manufacturer would have to certify
that the guard, as modified, complies with the equipment standard.
Also, the vehicle manufacturer would have to affix its own
certification label and prepare modified installation procedures. The
installation procedures are necessary both to ensure that the guards
are modified and installed the same way each time, and to allow NHTSA
to duplicate the modification when conducting compliance testing.
The original guard manufacturer's installation instructions may
provide for some flexibility in the installation. For example, they may
specify that a certain kind of spacer may be used to achieve a proper
fit, or that a doubler plate be installed if the thickness of the
chassis is below a certain amount. However, NHTSA may employ any of the
installation options provided to the vehicle manufacturer when
subjecting a guard to compliance testing. Any test failure of a
properly installed guard will represent noncompliance by the guard
manufacturer.
7. Compliance Test Requirements and Procedures
a. Dynamic Versus Static Testing. Several commenters, including
Advocates, urged that NHTSA require that the guards be tested
dynamically, that is, by crashing cars into the rear of trailers
equipped with the rear impact guard. The agency agrees that dynamic
testing more closely simulates the conditions in which underride
crashes occur in the real world than the quasi-static testing does.
However, dynamic testing is also far more expensive. To test one guard/
trailer combination with a dynamic test for strength and energy
absorption would entail total test costs of approximately $30,000.
Dynamic tests would be so expensive that specifying such testing of
trailers could raise practicability concerns regarding those trailer
manufacturers that are small businesses. A requirement based on such
tests would place these small manufacturers, which are numerous, at a
competitive disadvantage, relative to larger companies, and would
represent a significant financial burden.
Quasi-static tests provide similar information far more
economically than dynamic tests. The VRTC research project demonstrated
that quasi-static testing generates similar forces to those generated
in an actual crash test, albeit at a slower rate. The project also
demonstrated that guards only ten percent stronger than the minimum
level of strength necessary to pass quasi-static test requirements
performed adequately in dynamic tests. The quasi-static compliance test
for a single guard at VRTC cost only about $3,500. Based on the
foregoing and the discussion in the section above on separate equipment
and vehicle standards, the agency believes that dynamic testing of
underride guards is unnecessary and overly expensive. NHTSA further
believes that quasi-static testing is adequate to ensure the
manufacture of safe and effective rear impact guards and that it will
do so at a far lower cost. Therefore, the quasi-static testing
procedure has been retained in the final rule.
Some commenters commented on the definition of ``rigid test
fixture.'' The TTMA assumes that a trailer can be used as a rigid test
fixture, and other commenters urged that testing be permitted on
trailers. The Institute for Injury Reduction commented that the terms
``sufficiently large,'' ``appropriately configured,'' and ``no
significant amount of energy'' in the definition of rigid test fixture
are vague, imprecise, ambiguous and in no way ``stated in objective
terms.''
NHTSA notes that a trailer may meet the equipment standard's
definition of a rigid test fixture, but because of slight flexing of
the vehicle structure, in other cases, they may not meet this
definition. NHTSA is persuaded that the benefits of testing on trailers
outweigh the possible effect on testing repeatability and does not want
to discourage testing on trailers by conducting its compliance testing
only on a rigid test fixture. The TTMA comment indicates that, although
it is not required, some vehicle manufacturers will conduct quasi-
static guard testing on trailers or trailer portions. NHTSA sees no
reason why this should not serve as a basis for manufacturer
certification even if the trailer is not a rigid test fixture. The use
of a trailer would be desirable because there is nothing more
``appropriately configured'' for guard mounting than the actual trailer
the guard will be installed on and because the structural integrity of
the trailer chassis will also be tested. However, caution must be
exercised to assure that the trailer is secured so that it does not
move during the test. If the guard is mounted to a trailer, the trailer
chassis will be secured so that there is no rotation or translation of
the trailer tires during the tests for guard strength and energy
absorption.
When conducting compliance testing, the agency will give the guard
manufacturer the option of designating testing on a rigid test fixture
or on a trailer. NHTSA notes that it may test on any trailer described
as appropriate in the guard manufacturer's installation instructions,
even if the guard manufacturer based its certification for that trailer
not on actual testing but on engineering analysis.
NHTSA agrees with the Institute for Injury Reduction that the
definition of ``rigid test fixture'' needs a slight modification. The
reference to size has been eliminated because size is not really as
important as rigidity. However, it is not necessary to define the
amount

[[Page 2015]]
of energy the fixture can absorb, because, like the ``fixed collision
barrier'' defined in 49 CFR 571.3, the guards will be expected to pass
the test no matter how little energy is absorbed by the fixture. Also,
the term ``appropriately configured'' has been clarified. There is no
way to precisely define how the test fixture will have to be configured
because that will depend on the design of the guard being tested. There
may be a number of appropriate configurations. As long as the guard can
be attached to the test fixture in the same way that the guard
manufacturer's instructions specifies the guard is to be attached to
the vehicle, without either modifying the guard or adding adaptive
parts to obtain a better fit between the guard and the fixture in a way
that is inconsistent with the instructions, the test fixture is
appropriately configured.
The agency had modified the strength test procedures to promote
ease of testing. Paragraph (b) of S6.5 now requires the application of
the force to the loading device to achieve a constant deflection rate,
rather than a constant increase in force, as proposed in the SNPRM. In
other words, rather than increasing the force at a constant rate, the
deflection rate is required to be held constant and the force will vary
depending on the resistance offered by the guard. Specification of a
deflection rate procedure is consistent with existing agency practice.
For example, the quasi-static compliance tests in S4(d)-(e) of Standard
No. 214, Side Impact Protection and S6.3 of Standard No. 216, Roof
Crush Resistance utilize this technique for force application.
b. Test Sites. Several commenters recommended changes in the
language 'specifying the test sites to be used during the compliance
tests. Mr. John Tomassoni recommended defining the P1 test site such
that the ``3/8 L'' lateral dimension (see Figure 1) is defined relative
to the side extremities of the trailer, as opposed to the center of the
guard. He suggested that this change would account for newer 2,600 mm
(102 in) wide trailers which have a 1,270 mm (50 in) longitudinal frame
rail span, or for any other width trailer. This approach, however, is
inconsistent with a separate equipment standard because the exact width
of the trailer may not be known at the time of testing. Moreover, the
requirement that guards extend to within 100 mm (4 in) of the side of
the trailer should assure that the P1 site will be sufficiently
outboard on the trailer, because wider guards will be required for
wider trailers, and the P1 location is dependent upon guard width.
Mr. Tomassoni also suggested that S5.2.2 and Figure 1 should be
modified to specify that the vertical center of force should not be
more than 560 mm (22 in) from the ground, rather than at ``the
horizontal plane that passes through the vertical center of the
horizontal member,'' as proposed in the SNPRM. Mr. Tomassoni indicates
that a guard with a horizontal member of cross sectional vertical
height greater than 100 mm (4 in) would result in higher test points.
Higher test points would yield test results that are not indicative of
the guard's effective impact strength near the bottom edge, where force
is likely to be concentrated in real world crashes. Although it is not
possible to define the test points relative to the ground because the
guard is not required to be mounted on the vehicle during testing,
NHTSA has modified the rule to define the test points relative to the
bottom of the guard itself. This should assure adequate strength and
energy absorption at the level of likely impact force.
Mr. John Kourik pointed out that the P1 test site was defined
incorrectly in the SNPRM, although it was correctly portrayed in Figure
1. The text of S5.2.2(a) (redesignated S6.4(a) in this rule) read ``3/8
of the transverse horizontal distance * * * between the * * * vertical
centerline of the guard [and] and the outermost edge * * * of the
guard.'' The P1 definition has been corrected to reflect that the point
is located 3/8 of the total guard width outboard of the centerline. Mr.
Kourik also suggested that the four asterisks showing the P3 test sites
in Figure 1 be reduced to two asterisks. NHTSA has modified the figure
to make it clearer that there is only one P3 test site on each side of
the guard, but that the location of the site is within a range from the
centerline.
The TTMA and other commenters suggested broadening the range of
locations of the P3 test site to allow it to be ``any point selected by
the manufacturer * * * between 14 and 25 [rather than 20] inches
outboard'' of the guard centerline. Most new trailers are wider than in
the past with a frame rail span of 127 cm (50 in), and the frame rail
is a likely chassis structure for guard attachment. TTMA wanted NHTSA
to conduct the more demanding 100 kN (22,480 lb) P3 test near the
attachment point of the guard's supports. This was NHTSA's general
objective in specifying the P3 test location, and this objective is
furthered by accommodating TTMA's request in part. The rule has been
modified to provide that P3 is located 355 to 635 mm (14 to 25 in) from
the guard centerline. However, NHTSA will select any point within the
range for compliance testing, rather than permit a manufacturer to
specify a single test site within the 355 to 635 mm (14 to 25 in)
range.
c. Labeling and Certification. The TTMA suggested that affixing a
certification label is redundant in those instances in which the guard
is manufactured by the vehicle manufacturer because the vehicle
manufacturer has to certify compliance with all the safety standards
anyway. Although this is true, allowing some guard manufacturers to
omit the label would be impractical from an enforcement standpoint,
because vehicle inspectors would not be able to tell whether the guard
was certified by the guard/vehicle manufacturer as part of the vehicle
or whether the vehicle manufacturer installed a guard purchased from a
guard manufacturer who neglected to make a required certification.
Moreover, NHTSA does not believe that affixing the label is a
significant burden. Therefore, the final rule retains the requirement
of a separate guard certification for all guards.
The TTMA also recommended that the label be affixed to the roadside
vertical supporting member of the guard, instead of the center of the
horizontal guard member, to prevent damage and abuse. NHTSA believes
that docking and other routine operations could damage the label if
affixed in the proposed location. Therefore, the rule has been modified
to require the label to be affixed in a less vulnerable location. The
rule now requires the certification label to be placed on the
forwardmost surface of the horizontal member of the guard at an offset
location 305 mm (12 in) inboard of the right side end of the guard.
The TTMA also suggested changes in the label format. Specifically,
it recommended that the letters and numbers should be 2.5 mm (\3/32\ of
an inch) high, which is the same as the trailer certification label,
rather than 13 mm (\1/2\ inch) high as proposed in the SNPRM. TTMA also
asked that NHTSA require that the label be furnished to the vehicle
manufacturer with a protective cover that can be removed after
painting.
The agency believes that the smaller letters suggested by TTMA are
sufficiently legible for inspection purposes, and has changed the rule
to adopt this suggestion. However, market forces should determine
whether protective covers are provided. Vehicle manufacturers will
probably cover the labels themselves when painting to avoid having
their guard confused with a noncomplying guard.

[[Page 2016]]

C. Standard for Vehicles

1. Configuration Issues
a. Maximum Guard Ground Clearance. One of the major issues
addressed by nearly all the commenters was the maximum ground clearance
of the horizontal member of the rear impact guard. The SNPRM proposed a
maximum guard height of 560 mm (22 in). Consumer safety groups and
private citizens generally favored lowering the guard to within 405 or
460 mm (16 or 18 in) of the ground in the belief that doing so would
provide more complete protection for low profile vehicles such as sub-
compact and mini-compact passenger cars. Since the real issue is not
ground clearance, but guard height relative to the front structure of
colliding passenger vehicles, some of these commenters addressed
related issues such as the height of the engine block, hood, and cowl
(windshield base) of those vehicles. Except for the consumer safety
groups and a few private citizens, few provided a rationale or any data
to support a lower guard height. The organizations and private
companies related to the trucking industry generally supported a 560 mm
(22 in) height, but offered a variety of reasons not to lower the guard
further. Most of their concerns related to operational difficulties
that would be caused by lower guard heights.
The consumer safety groups focussed their comments on guard
effectiveness. Advocates advanced several reasons for reducing the
guard height in order to achieve better engagement between the guard
and the engine block, bumper, and tires of colliding passenger
vehicles. Advocates stated that lower engine block heights on modern
automobiles, combined with the lowering of the passenger vehicle's
front end due to suspension compression during severe braking, will
result in the rear impact guard passing over the engine and engaging
only the hood and fenders of most cars. In addition to the front end
lowering caused by braking, Advocates claim that additional frontal
lowering will occur on downgrades due to forward weight transfer.
Citing a random survey it made of subcompact cars and urging NHTSA to
conduct a more thorough survey, it said that no engine block is higher
than 560 mm (22 in) above the ground and bumpers are in the 430 to 535
mm (17 to 21 in) range. It stated that earlier NHTSA data using the
average hood height above the ground was misleading because its
``casual'' survey of subcompact hood front edges showed none higher
than 635 mm (25 in). It interprets these data to mean that only fender
top and hood sheet metal would be engaged, and concluded that air bag
sensors probably will not be triggered. Advocates also maintains that,
even if the top of the engine were engaged, the underride guard will
cause the blocks of transversely-mounted engines used in most
subcompacts to rotate (roll) rearward, crushing the car occupant's
legs. Based on British research, Advocates recommends a guard height of
no more than 405 mm (16 in), and ideally 305 mm (12 in). Both Advocates
and Mr. Byron Bloch, of Auto Safety Design, cited the 1980 study by
Dynamic Science which concluded that the guard height should not exceed
510 mm (20 in). Mr. Bloch recommended a height of 405 to 460 mm (16 to
18 in). CRASH solicited many private citizens to send in petitions,
letters, and pre-printed cards stating that the guard height should be
set at 405 mm (16 in), but none provided supporting technical
information.
The IIHS, citing the same studies as Advocates, urged NHTSA to
adopt a maximum ground clearance of 460 mm (18 in). IIHS is primarily
concerned that a 560 mm (22 in) high guard will override car bumpers,
thus bypassing much of the potential front end energy absorption. Other
concerns expressed by IIHS were late air bag activation, braking-
induced bumper depression of two to 100 mm (4 in) or more, and possible
lifting of the rear end of the trailer as the car wedges under the
guard. IIHS implied that a 460 mm (18 in) requirement is practical,
noting that one U.S. freight carrier reportedly sets its guards at 495
mm (19.5 in).
IIHS believes NHTSA's estimate that trailers probably sit 50 to 75
mm (2 to 3 in) lower when loaded is wrong. IIHS tests on 11 trailers
showed the most heavily loaded trailers showed only 38 to 57 mm (1.5 to
2.25 in) of depression with an average of 28 mm (1.1 in). Four of the
trailers even raised in the rear, indicating that load distribution is
probably a factor in determining rear extremity compression height.
IIHS believes that modern air suspensions compensate for loading
depression. Even if loaded trailers are depressed, it believes that
passenger vehicles should be protected from partially loaded or empty
trailers, which it says are involved in 29 percent of fatal crashes.
Therefore, IIHS urges NHTSA to assume no depression of the trailer bed
due to loading.
Mr. John Tomassoni commented that a lower guard would be better
because engine block resistance to a rigid guard doesn't start until
460 to 610 mm (18 to 24 in) behind the bumper. However, Mr. Tomassoni
concluded that a 560 mm (22 in) requirement is a significant
improvement over the existing 760 mm (30 in) height, and one that can
be implemented with little or no difficulty. He notes that trailers 16
meters (m) (53 feet (ft)) or longer are currently being equipped with
560 mm (22 in) high guards.
Some municipalities sent comments in favor of lower guard heights.
For example, the City of Durham, North Carolina sent an unsigned
resolution that the height be set at no more than 460 mm (18 in). Its
Transportation Advisory Committee submitted a similar comment. About
2,300 private citizens recommended a guard height of 405 mm (16 in).
The industry groups focussed their comments relating to guard
height on operational restrictions that would result from the reduced
``angle of departure'' that lower ground clearance would cause. The
angle of departure is basically the acute angle formed by the ground
and a line connecting the point where the rear tires meet the ground
with the bottom of the guard. The lower the guard, and the further
forward the rear wheels are positioned relative to the guard, the
smaller the departure angle is, and therefore the more likely the guard
is to scrape or ``hang'' on the ground when the trailer mounts a steep
incline. The problem is exacerbated for the longer 16 m (53 ft)
trailers being used today, because they have correspondingly greater
rear overhangs, and thus smaller departure angles. Many trailers have
their rear wheels mounted on sliding tandems, or bogeys, that can be
moved forward or rearward on the trailer's frame, depending on the load
and the need for maneuverability. The further forward the wheels are,
the more maneuverable the trailer is and the more the rear end of the
trailer ``swings out'' in turns.
Changes in the industry since 1981 seem to have relieved the
concerns of the rail industry that the proposed ground clearance of 560
mm (22 in) would interfere with rail car loading and unloading
operations, in which trailers are driven up steep ``circus ramps'' onto
flat cars. The Association of American Railroads (AAR) and TTX Company,
a trailer-on-flat-car operator, opposed the 1981 NPRM, but now support
the 560 mm (22 in) requirement because there are few ``circus'' ramps
still operating. However, they caution that a significantly lower
height would interfere with intermodal flatcar operations. TTX asserted
that such a reduction in guard clearance could interfere with lift-on
and lift-off operations for one type of railroad car (TTAX ``spin
cars'') handling 16 m (53

[[Page 2017]]
ft) trailers. It added that there must be extra guard clearance to
account for loading depression and bouncing. To illustrate the
potential economic impact of lower guard clearance, TTX stated that
there are 2,300 such cars costing $340 million, which are only 1.8
years old on average. TTX estimates that lowering the guard clearance
could eliminate 75 percent of the capacity for 14 railroads.
In contrast, the 560 mm (22 in) guard height is still considered
low by the portion of the industry that transports trailers in ships.
Transamerica Leasing, Inc. recommends that NHTSA conduct further study
before issuing this rule because a 560 mm (22 in) high guard would
scrape loading ramps during roll-on/roll-off ship loading when the
wheels are positioned forward to provide the maneuverability necessary
in ships. The American Trucking Associations (ATA) supports the 560 mm
(22 in) proposed ground clearance, but stated that any lower clearance
would be unacceptable. It calculates that a loaded trailer driven onto
a barge or vessel, which it says have departure angles as high as 15
degrees, would drag the guard if the rear axle is 190 cm (74.5 in) or
more forward of the guard. It said that many states have restrictions
on trailer kingpin-to-rear- axle distances that result in a 245 to 275
cm (96 to 108 in) rear-wheel-to-guard distance on 16 m (53 ft)
trailers. It concludes that these trailers' guards would hang on such
vessel loading ramps or on any 20 percent grade. It finds the 560 mm
(22 in) clearance acceptable only because 16 m (53 ft) trailers are
rarely used on vessels, and because 20 percent grades are rare. The
Truck Maintenance Council of the ATA recommends a guard clearance of
560 mm (22 in) for general freight equipment. According to Mr. Robert
Crail, a trailer designer and manufacturer, the proposed 560 mm (22 in)
height is acceptable because, although many trailers are still driven
into ships rather than being crane loaded, vessel owners can adjust
their ramps, and because it is compatible with the dimensions
established by the trucking industry and loading dock restraint device
manufacturers. Ford Motor Company had no specific data, but is
concerned that 560 mm (22 in) may be inadequate ground clearance for
loading and unloading of long trailers in trains or ships. Ford also
noted that some single unit trucks are equipped with kneel-down air
suspensions to facilitate loading and unloading, which Ford says are
incompatible with a 560 mm (22 in) high guard.
Even outside the context of intermodal loading and unloading
operations, some commenters were concerned about the reduced departure
angle that a 560 mm (22 in) high guard would create. The National Solid
Waste Management Association (NSWMA) emphasized the importance of
maneuverability for sanitation trucks in negotiating driveways and
backing into tight places. It estimated that a 560 mm (22 in) guard
mounted flush with the rear extremity of a sanitation truck would have
a departure angle of only 9 degrees, which it says is typical of many
driveway entrances. Although it appears that many of the trucks NSWMA
is concerned with are single unit trucks that are excluded from the
rule, NSWMA is also concerned about the guards getting hung up on the
ground when the trailers are taken off-road onto the soft, unpaved,
uneven roads at landfills and construction sites.
One additional industry concern is engagement of the guard with
``dock locks.'' When trailers back up to loading docks, these devices
engage the underride guard to keep the trailer from moving away from
the loading docks as forklifts repeatedly travel across the rear door
sill. Transamerica Leasing believes that the 560 mm (22 in) high guards
may interfere with ``dock lock'' engagement arms. Yellow Freight System
states that thousands of dock locks have been installed according to
the 560 mm (22 in) guard height recommended by the Maintenance Council
of the ATA, and urges NHTSA not to change now. However, Rite Hite
Corporation, a manufacturer of dock locks, submitted information
indicating that dock locks can accommodate guard heights between 355
and 760 mm (14 and 30 in).
One industry group endorsed a lower guard height. The AFL-CIO
Teamsters Union suggested that NHTSA could require a ground clearance
lower than 560 mm (22 in) because auto carriers and UPS trailer fleets
have reported no problems with lower guard heights. It also observed
that 16 m (53 ft) trailers in many states have no problem using 560 mm
(22 in) guards.
The question of proper guard ground clearance involves a balancing
of the effectiveness of the guard in providing protection against PCI
against the cost and operational restrictions that lower guard heights
could impose on the industry.
The effectiveness of the guards is a primary consideration.
Regarding Advocates' survey of bumper and hood heights on compact and
subcompact cars, NHTSA conducted a similar survey of engine block
height and front end profile of a sample of 40 vehicles. The results of
this survey were summarized in the agency's Truck Underride Report to
Congress, dated November, 1993. The NHTSA survey showed that the height
of the top of the engine block was between 660 and 790 mm (26 and 31
in), with an average height of 840 mm (28 in). The hood leading edge in
NHTSA's survey averaged about 685 mm (27 in) and the lower edge of the
windshield frame averaged about 840 mm (33 in). The agency is not aware
of the basis upon which Advocates selected the cars for its survey, but
NHTSA's survey was targeted preferentially at cars with the lowest
front end profile. Since NHTSA's average heights were higher than those
obtained by Advocates, NHTSA has no explanation for the discrepancy,
unless the survey methodologies were different. Hood heights have been
getting lower over the past few years, but that trend may have stopped
in the last two years. NHTSA believes that the average hood heights in
its survey are representative of the anticipated dimensions for new
passenger vehicles 5 to 10 years in the future. NHTSA concludes from
the VRTC test results that a 255 to 305 mm (10 to 12 in) overlap
between the guard bottom and the lower edge of the windshield will
ensure adequate structural engagement with the guard for the vast
majority of compact and subcompact cars.
NHTSA agrees with IIHS that a guard 560 mm (22 in) high will
override most bumpers, but disagrees that bypassing the bumper
sacrifices much of the potential front end energy absorption
capability. The bumper is designed to prevent cosmetic damage in low
speed crashes (less than 16 kph, or 10 mph) and provides only a small
portion of the energy absorption by a car crashing at higher speeds.
The bumper is mounted to the frontal crash energy management components
which extend rearward and upward to the rearmost section of the engine
compartment. These components will be adequately engaged by the rear
impact guard during a collision. Regarding IIHS's contention that NHTSA
should assume no loading-induced depression of the trailer bed, NHTSA
has not made such an assumption. The final rule regulates the guard
height only when the trailers are unloaded, and the 560 mm (22 in)
guard height was adequate in NHTSA's VRTC tests.
The agency conducted seven full scale crash tests with the proposed
guard in the course of the recent research project, using two types of
subcompact and two types of compact cars. These vehicles were
representative of average hood and engine heights for cars in those
size classes. The minimally compliant rear

[[Page 2018]]
impact guard was set 560 mm (22 in) above the ground. During these
tests, the cars had their front ends depressed to simulate the lowering
that would be experienced during heavy braking, but the guard was not
depressed to a level below the minimum clearance, as it might be if the
trailer were loaded. In some sense, therefore, these tests represented
a ``worst case scenario'' with regard to guard height. In each test,
the air bags were fully deployed before dummy contact and the
deceleration readings were much better than the minimum requirements in
Standard No. 208, Occupant Crash Protection. When there was no guard
attachment failure, they adequately engaged the structure of each car
and prevented PCI. There was little movement of the engine and no
contact between the engine and fire wall. The transversely mounted
engines did not rotate substantially, and none of the dummies legs were
crushed. Therefore, based on the docket comments, the recently
completed crash tests, and the assessment of late model passenger
vehicle frontal structure characteristics, NHTSA concludes that the 560
mm (22 in) maximum guard ground clearance is adequate to engage the
frontal crash energy management structure of most subcompact and
compact cars.
Although some small sectors of the industry may be affected, NHTSA
does not believe that there will be any insurmountable problems with a
560 mm (22 in) guard height. Several states have required 560 mm (22
in) maximum guard ground clearances in conjunction with the passage of
laws allowing 16 m (53 ft) trailers. NHTSA contacted several
distributorships/dealerships that sell heavy trailers in excess of 15 m
(50 ft) in length to the trucking industry and was unable to obtain
information documenting substantial operational problems due to guard
ground clearances of 560 mm (22 in) or less. The AFL-CIO Teamsters
Union did not give NHTSA enough information about the operating
environment of Carolina Freight Carriers Corporation, the trucking
company that sets its guards at 495 mm (19.5 in), to determine why they
have not experienced the problems that the other commenters expect with
guards lower than 560 mm (22 in).
NHTSA does not believe that the number of trailers involved in ship
roll-on/roll-off and trailer-on-flat-car circus ramp operations is
significant. TTMA data indicate that less than 5 percent of trailers in
the U.S. are ever transported by ship or barge, and that between one
and less than ten percent of new trailers are produced for trailer-on-
flat-car use. Modifications of may solve these problems. Most of the
vehicles in the waste services fleet mentioned by NSWMA are single unit
trucks excluded from the rule. However, in those few cases where there
are still problems, movable or adjustable guards may be needed.
There is adequate evidence in the comments to conclude that
requiring a guard height lower than 560 mm (22 in) would cause an undue
burden on the industry. Of particular concern are the comments of ATA,
TTX, AAR, and Transamerica Leasing, indicating that any height below
560 mm (22 in) will cause interference in intermodal operations.
Moreover, a lower height will increase the probability that the guard
will scrape or snag during normal vehicle operations and be damaged as
a result. Therefore, because the 560 mm (22 in) maximum ground
clearance proposed in the SNPRM appears to be the lowest height that
provides adequate effectiveness without imposing an undue burden, it
has been retained in the final rule. The agency notes that guards may
be mounted with less than the maximum allowable ground clearance.
b. Guard Width. The SNPRM proposed that the horizontal member of
the guard be required to extend across the width of the trailer to
within 100 mm (4 in) of the side extremities, but not outboard of the
side extremities. Advocates commented that the 100 mm (4 in) allowance
appeared arbitrary, based on the rulemaking record, but did not
actually suggest that the guard should extend fully to the side
extremities of the trailer. The AFL-CIO Teamsters Union indicated that
it fully supports the SNPRM's 100 mm (4 in) allowance, while noting
much anecdotal information from drivers about the importance of a
``full width'' guard, especially for crashes that occur at an angle to
the rear of the trailer.
NHTSA notes that there is no requirement of a 100 mm (4 in) inset.
Vehicle manufacturers are permitted to install guards extending the
full width of the trailer. However, the 100 mm (4 in) allowance gives
trailer and guard manufacturers some flexibility in choosing and
providing guards, without sacrificing safety or effectiveness. From the
perspective of guard effectiveness, it is doubtful that the extra
lateral coverage would significantly increase the strength of the guard
at its extremities or its ability to protect passengers in an offset
collision.
In fact, a 100 mm (4 in) inset would decrease the previously
mentioned ``hooking'' potential during sharp turns of the trailer and
provide more clearance in certain passing situations. The Florida
Department of Transportation and the TTMA recommended allowing rounded
guard ends to alleviate this potential problem, but NHTSA notes that a
100 mm (4 in) inset on an unrounded guard will partially accomplish the
same goal. As discussed above in the section on shape of the horizontal
cross member, pursuant to the TTMA's suggestion NHTSA has modified the
rule to allow rounded corners on guards to begin curving at a point 255
mm (10 in) inboard of the edges of the vehicle, while retaining the 100
mm (4 in) requirement for straight guards. Curved guards still have to
meet the other requirements of the vehicle standard (i.e., extend to
within 100 mm, or 4 in, of the side extremity). This modification
merely removes for the curved portion of the guard the requirement that
the bottom of the horizontal member be within 560 mm (22 in) of the
ground, in the case of upward curving guards, and the requirement that
the rear surface of the horizontal member be within 305 mm (12 in) of
the vehicle rear extremity, in the case of forward curving guards.
c. Specification of the Rear Extremity. Some commenters requested
that NHTSA modify the proposed definition of ``rear extremity'' to take
into account vehicles with high protrusions in the rear. The SNPRM
defined the rear extremity as the rearmost point of the vehicle that is
located 560 mm (22 in) or more above the ground. The specification of
the rear extremity is important because the SNPRM also requires that
the rear impact guard be located no more than 305 mm (12 in) forward of
the rear extremity of the vehicle. Some trailers and semitrailers, such
as hopper trailers with V-shaped bins and trailers with liftgates or
refrigerator units in the upper rear, are shaped such that the rear
extremity of the vehicle is located well above the road surface. These
protrusions do not present a danger of PCI because they are located
well above the roof line of most passenger vehicles. Yet, applying the
rear extremity definition in the SNPRM, a rear impact guard would have
to be mounted such that it extends rearward from the base of the
trailer to a position within 305 mm (12 in) of the back of the high
protrusion. Such an extended guard might pose a safety hazard as well
as operational difficulties.
Several manufacturers of vehicles with high rear end overhang
recommended alternative definitions of ``rear extremity'' that excluded
portions of the trailer rear that were high enough to clear the roofs
of passenger vehicles. The TTMA and the ATA recommended that vehicle
structure with a ground

[[Page 2019]]
clearance of 1,680 mm (66 in) or more be excluded from the definition
of rear extremity. NSWMA recommended excluding that portion of the rear
of the vehicle located 1,520 mm (60 in) or more above the ground.
The agency acknowledges the potential problem with the proposed
specifications and believes that redefining the rear extremity to
accommodate these vehicles is possible without reducing rear impact
guard effectiveness or creating new safety hazards. NHTSA contacted
officials from TTMA and ATA to obtain more information about the
current number and future production plans for vehicles of this type.
According to TTMA, these are mostly highly specialized vehicles and the
high overhang often consists of equipment such as cranes in addition to
``bubble door'' type container trailers. TTMA estimates that these
vehicles constitute less than one percent of the annual trailer and
semitrailer production and there is no trend toward increasing the
numbers substantially. ATA also estimated that the number of vehicles
produced annually with high rear overhanging structure represents less
than 5 percent of the total annual production of trailers and
semitrailers. ATA did not provide information on the future trend of
production of these vehicles, but indicated that the number has been
fairly constant in the recent past with new vehicles brought into
service primarily to replace vehicles going out of service.
The NSWMA recommended that the rule specifically state that, for
roll-off/hoist type trailers, the containers on the hoist frame be
considered as part of the load and not as part of the vehicle for
purposes of rear extremity specification. It suggests that the rearmost
part of the hoist frame should be considered the rear extremity.
Containers extend up to 1.5 m (5 ft) rearward from the end of the hoist
frame.
The agency has decided to revise the SNPRM's definition of ``rear
extremity'' to limit its ambit to the portion of the vehicle's rear
located between a lower and upper height limit. The lower limit
specification remains unchanged at 560 mm (22 in) (that is, guard
ground clearance). An upper limit for the area in which the rear
extremity is located has been specified at 1,900 mm (75 in) above the
ground surface for purposes of the vehicle standard. The portion of the
rear of the trailer that is located in the same horizontal planes as a
passenger vehicle windshield is the critical area for rear underride
protection. This is between 760 mm and 1,900 mm (30 and 75 in) above
the ground for almost all passenger cars, vans, and light trucks.
With regard to roll-on/hoist type trailers, the agency agrees with
NSWMA that there would be numerous regulatory problems involved in
considering the containers to be part of the vehicle, rather than part
of the load. Although the containers may extend beyond the end of the
vehicle and are capable of causing PCI just like the rear end of a
trailer, they are not part of a new vehicle as manufactured. Further,
the boxes, tanks, and other specialty containers are manufactured,
maintained, and in many cases owned separately from the vehicle. NHTSA
has no authority to regulate vehicle loads under 49 U.S.C. Chapter 301.
While NHTSA cannot require guards on the container on roll-on/hoist
type trailers, it can require guards on the rear of the trailer that
carries it. If the vehicle is designed to carry containers that do not
extend appreciably beyond the rear of the vehicles, the agency sees no
basis for excluding it. Casual observations indicate that the
containers do not usually extend beyond the rear of the vehicle, so
these trailers are required to have guards. The rear extremity will be
determined without the container.
d. Distance between the Guard Rear Surface and the Vehicle Rear
Extremity. Several commenters urged NHTSA to change the requirement
proposed in the SNPRM that the guard's horizontal member be mounted not
more than 305 mm (12 in) forward of the rear extremity of the trailer
and not rearward of the rear extremity. The distance between the guard
and the trailer rear extremity is significant because the sooner the
passenger vehicle engages the underride guard, the farther its occupant
compartment will be from the rear of the trailer when the guard is
engaged, and the better the chance that the passenger vehicle will stop
short of PCI.
Some commenters thought that NHTSA should allow the guard's
horizontal member to extend rearward of the rear extremity. Mr. John
Tomassoni stated that he saw no good safety reason for restricting rear
extension, since it is beneficial for preventing PCI. The TTMA also saw
no reason why the guard should not be located rearward of the rear
extremity. It also suggested a change in the language of S5.1.3 that
makes it clear that, even above 560 mm (22 in) the guard cannot be more
than 305 mm (12 in) from the rear extremity of the vehicle.
The Rite Hite Corporation stated that, for dock locks to function,
there must be no more than 230 mm (9 in) between the rear extremity and
the guard. It is concerned that the 305 mm (12 in) allowance will
render the dock locks useless.
NHTSA notes that the 305 mm (12 in) allowance is not a minimum, but
a maximum requirement. Casual observations by the agency indicate that
nearly all trailers currently have their guards mounted flush with the
rear extremity of the trailers. This practice is also specified as the
recommended practice in the ATA Maintenance Council guidance (RP 707).
It is also the configuration most compatible with dock locking
mechanisms. Based upon the TTMA's comment relating to mounting rearward
of the rear extremity, the industry appears to be in favor of mounting
as far rearward as possible. Therefore, NHTSA believes that trailer
manufacturers will continue to mount guards flush with the rear
extremity of the vehicle.
The main incentive to change the prevailing practice relates to the
smaller departure angle that will be created by lowering the maximum
guard ground clearance from 760 mm to 560 mm (30 to 22 in). Moving the
guard 305 mm (12 in) forward will slightly increase the departure
angle. However, nothing in this rule increases that existing incentive.
Therefore, the agency does not expect that a 305 mm (12 in) allowance
would have any effect on prevailing practice. Further, NHTSA does not
believe the benefit of moving the guard forward would be very
significant. Nevertheless, the agency had modified the requirement in
section 5.1.3. for guard rear surface location, or off-set, to state
that the guard should be mounted as close as practical to the rear
extremity of the vehicle. This will prevent vehicle manufacturers from
mounting the guard with up to 305 mm (12 in) of forward off-set from
the rear extremity of the vehicle unless the off-set is necessary and
not merely convenient. It should be noted that the requirement to mount
the guard as close to the rear extremity as practical is identical to
the requirements of ECE Regulation 58.
NHTSA agrees that having the horizontal member of the guard
positioned rearward of the rear extremity would be beneficial for
preventing PCI in the event of a crash. Some meritorious guard designs,
such as the Quinton-Hazel hydraulic energy absorbing guard and the Hope
rearguard underrun device, utilize horizontal members that are hinged
so that they are angled down and slightly rearward from the rear of the
trailer. This rearward positioning enables the guard to engage a
striking vehicle at a greater distance from the rear extremity and
gives the guard a greater distance to swing

[[Page 2020]]
forward and ``ride down'' the energy of the striking vehicle before PCI
occurs. If vehicle manufacturers want to provide this extra measure of
safety, this agency will not discourage it, as long as vehicle
manufacturers consider State laws governing overall combination truck
length. However, NHTSA does not want to require rearward positioning
because this configuration exacerbates the previously mentioned
potential for ``hooking'' adjacent vehicles during sharp trailer turns
and in other situations. Therefore, NHTSA has removed the SNPRM's
prohibition on positioning the horizontal member rearward of the rear
extremity. The new requirement that the member be as close to the rear
extremity as practical is limited so that it does not prohibit mounting
rearward of the rear extremity.
Advocates stated that NHTSA has no data to support the 305 mm (12
in) allowance because all crash tests were done with guards positioned
at the very rear of the trailer, thus implying that testing in the
forward-mounted position is required to support the allowance.
Even though the crash tests conducted by NHTSA had the rear impact
guards mounted in the usual posi

[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%3A96-682. Public record. Not legal advice.