NHTSA Rear Guard NPRM Comments Guide

NHTSA is proposing to adopt requirements of the Canada Motor Vehicle Safety Standard (CMVSS) for underride guards (CMVSS No. 223, “Rear impact guards,”) that became effective in 2007.

NHTSA is stealing a 2007 safety regulation from another country to apply to U.S. trailers in 2018 and beyond! We are looking at adopting an 11 year old regulation during the time of the swiftest change in automotive technology including safety in history. We cannot afford to do our own research and write our own regulations.

The CMVSS No. 223 requirements are intended to provide rear impact guards with sufficient strength and energy absorption capability to protect occupants of compact and subcompact passenger cars impacting the rear of trailers at 56 km/h (35 mph).

We recognize, however, that benefits may accrue from underride crashes at speeds higher than 56 km/h (35 mph), if, e.g., a vehicle’s guard exceeded the minimum performance requirements of the FMVSS. NHTSA requests information that would assist the agency in quantifying the possible benefits of CMVSS No. 223 rear impact guards in crashes with speeds higher than 56 km/h (35 mph).

About 26 percent of fatal light vehicle crashes into the rear of trailers is at speeds 56 km/h (35 mph) or less, this means 74% of fatal light vehicle crashes are at speeds exceeding 56 km/h (35 mph). This means that NHTSA for the next twenty years or until around 2040 will not try to save 3/4 of fatal victims. Only the Canadian MANAC guard passed all tests at 35 mph (56 km/h), we believe this guard is too heavy at 300 lbs (136 kg) for the American market so calling this a 35 mph (56 km/h) standard is misleading. We believe it is truthfully a 30 mph (48 km/h)proposed standard.

MUARC tested energy-absorbing guards to 75 km/h or 47mph in the early 1990’s and the Impact Project tested energy-absorbing guards to 40 mph full and offset with computer models showing performance possible at 50 mph and more.

MUARC Monash University Australia energy absorbing guard design
Strong energy absorbing struts to outside edge for offset crashes














UNICAMP Impact Project in Brazil energy absorbing guard test
Angled high strength reinforcing struts to outside edge



“The size of an energy-absorbing truck front structure directly correlates to the survivable closing speed between car and truck in head-on collisions (e.g. 75 km/h survivable closing speed requires a 400 mm long energy-absorbing structure, 90 km/h (56 mph), requires 800 mm).” From Volvo Report

NHTSA Study: ” A study by Minahan and O’Day of fatal car-truck accidents in Michigan and Texas found evidence of underride in 90 percent of rear-end impacts and 70 percent of side impacts. Underride was found typically to occur at night on straight rural roads. Impact speeds were generally greater than 30 mph (48 km/h). The authors characterized this type of crash as a “surprise event” in which a passenger vehicle came upon a slower or stopped truck unexpectedly.”

VC-COMPAT Project The analysis revealed that approximately 57 % of the fatalities and 67 % of seriously injured could be prevented from their injures due to improved rear underrun protection systems (RUPS).

VC-COMPAT: “Plans are underway to extend the front of the truck 300mm to 500mm or more to create a crash zone or deformable soft nose that would absorb crash energy and might reduce serious injuries and fatalities another 10% from the current standard on trucks with energy-absorbing guards, and survivable speeds would be increased to 80-90kph (56 mph).”

NHTSA refuses to crash test at real world speeds including highway speeds to keep this information from the public to encourage uninformed adoption of these corporate preferred cheap and low speed guard standards.

Based on information from the Truck Trailer Manufacturers Association (TTMA), NHTSA estimates that 93 percent of new trailers sold in the U.S. subject to FMVSS Nos. 223 and 224 are already designed to comply with CMVSS No. 223. The agency estimates that about one life and three serious injuries would be saved annually by requiring all applicable trailers to be equipped with CMVSS No. 223 compliant guards.

The NPRM intends to upgrade the Standard as basically existed in 1996, 2018 is suggested as the final date of implementation. It takes twenty to thirty years to agree on a new regulation, by the time 2018 rolls around we would probably see 98 % of trucks on the road voluntarily meeting the Canadian Standard due to crash testing guards at IIHS providing some pressure for guard design improvement. This means we will be legalizing 98 % of current guards on the road for the next twenty plus years with no technological improvement! We will have 1950’s technology in 2040!

Further, an evaluation of trailers manufactured in 1998 and later in the 2008-2009 TIFA data files from UMTRI showed that the average ground clearance of rear impact guards for newer (MY 1998+) trailer models was 457 mm (18 inches).

Some trailers may have the rear axle further forward to improve maneuverability of the trailer. NHTSA believes that, for such trailers, rear impact guards that are lower than 560 mm (22 inches) may scrape and snag with the ground and get damaged.

NHTSA plans on keeping guards at 22 inches discounting numerous studies showing lower guard heights improve injury and fatality outcomes. NHTSA hides all favorable information and studies from the public. Average guard height is already 18 inches or lower. In violation of Vision Zero NHTSA places possible maneuverability over saving lives.

The guard bottom must be close to the road to align at the height of most car bumpers, around 400 mm or 16 inches. The bumper must be engaged to properly interact with car safety features such as crush zones which are designed to absorb crash energy. If the guard is not low enough the car can penetrate below the guard and create the wedge effect (mostly for straight trucks) lifting the trailer causing crush zone features of the car to not engage and sending the hard bottom edge of the trailer towards the windshield and car occupants.


Penn State simulated car-truck crashes with varying guard heights

400 mm = 16 inches   600 mm = about 23 inches

MUARC recommends 400 mm height or about 16 inches


NHTSA’s proposed guard height at 22 inches (560 mm) from the roadway will kill in tests at higher speeds. Testing at low-speeds misrepresents the safety of higher guard heights and road tests prove lower guard heights are safe and feasible. A guard bottoming out would be rare and cheap to fix, why change our entire safety philosophy for a rare problem that does not really exist as American trailers average guard height is 18 inches (460 mm).

(MUARC Guard 4 year road test at 450 mm (18 inches) from roadway). The rear overhang was 3.5m. The clearance of the rear of the barrier from the road surface was set at 450 mm (18 inches). Condition of the barrier after 4 years: the Australia Post truck with the energy absorbing rear underrun barrier, photo taken on October 18th 2002, some 4 years after fitting.


Requiring that the guard be tested when attached to the trailer would be an unnecessary and significant cost burden for the manufacturers, especially for small trailer manufacturers with low sales volumes.

Violates Vision Zero reducing fatalities over cost benefit analysis.

Guards can break off of trailers during a crash due to the massive force involved. Attachment hardware such as bolts must be extra thick and strong, attachment points such as the trailer frame must also withstand tremendous forces and if not strong enough the trailer frame should be required to be reinforced in the standard. Guards should be tested on the trailer so attachment failure which is common can be decreased.

IIHS states that ideally, FMVSS No. 223 should require guards to be certified while attached to complete trailers, and that at a minimum, guards should be tested while attached to sections of the trailer rear that include all the major structural components and that are constrained such that the load paths near the guard are not changed.

The Manac rear impact guard prevented PCI in 56 km/h (35 mph) crash tests with full overlap, 50 percent and 30 percent overlap of the Malibu.

Manac guards are heavier at 307 pounds (140 kg). Improved guards will not be the norm due to higher weights and costs for little improvement. History shows minimum compliance, not maximum as with the Manac guards. The Manac guard is an Canadian guard and we do not expect adoption in the American market as it is a third or more heavier than American guard designs. We should see about 30 mph (48 km/h) performance for minimum compliance with poor extreme offset performance of American market guards.

We have tentatively decided not to require used trailers be retrofitted with CMVSS No. 223 compliant rear impact guards. Our analysis indicates such a retrofitting requirement would be very costly without sufficient safety benefits.

MUARC in Australia recommended an extra angled brace attached to the outside edge of the guard running to the horizontal trailer frame in their study of rear guards in the 1990’s. The Impact Project in Brazil also incorporated this outside angled brace and recommendation in their guard designs and recommendations. These braces could be added to older guard designs to improve offset impact performance and enhance full impact strength. Using new high strength aluminum alloys these braces could be added as zero cost as part of normal maintenance budget operations, with low cost, and low weight.

While 20 percent of fatal light vehicle impacts into the rear of trailers are wheels back trailers, they only represent 8 percent of those fatal crashes with PCI into the rear of trucks and 30 trailers.

Wheels back trailers are now exempt from guard standards but modern cars require a flat surface to interact with their safety systems. Tires are hard surfaces and present an uneven surface hazard that can be mitigated with the flat surface and proper height and energy absorbing features of an properly designed underride guard. NHTSA Study: “Wide-base singles may become more widespread to improve the fuel economy of the truck population. Wide-base singles present a larger gap for smaller, narrower light vehicles, which in turn brings the wheels-back exemption into question.” High-speed crashes require energy absorbing guards, the brick wall approach means giving up on saving high-speed fatal victims which are the majority in real world crashes.

In 2010, NHTSA published the results of a study, analyzing several data sources, to determine the effectiveness of trailer rear impact guards compliant with FMVSS Nos. 223 and 224 in preventing fatalities and serious injuries.5 The agency’s analysis of the Fatality Analysis Reporting System (FARS) could not establish a nationwide downward trend in fatalities to passenger vehicle occupants in impacts with the rear of trailers subsequent to the implementation of FMVSS Nos. 223 and 224.

Previous legalization of guards already on the road by NHTSA did not improve safety contrary to NHTSA claims which were disputed by all non-industry parties. In twenty years, we fear a similar outcome from legalizing guards on the road today. Thousands are dead that could have been saved with energy absorbing guards in 1992 and beyond, light weight simple designs existed since at least 1970.

We propose to replace the current definition of “rear extremity” in FMVSS No. 224 with that specified in CMVSS No. 223. The change is intended to ensure that aerodynamic fairings are located within a certain safe zone at the rear of the trailer. Aerodynamic fairings on the rear of trailers, also known as “boat tails,” are rear-mounted panels on trailers that reduce aerodynamic drag and fuel consumption.

Actually, common sense. We would recommend extending guards for high speed performance.

The average cost of a Canadian compliant rear impact guard is $492, which is $229 more than an FMVSS No. 224 compliant guard.

As shown in Table 10, upgrading from the FMVSS No. 224 compliant guard to the CMVSS No. 223 compliant guard would add an average incremental weight of 48.9 lb to the trailer, thereby reducing the overall fuel economy during the lifetime of the trailer. The incremental increase in lifetime fuel cost for a 48.9 lb weight increase of a trailer was estimated to be $1,042.2 and $927.7 discounted at 3 percent and 7 percent, respectively.

Excludes information on new light weight aluminum alloys purported to be low cost and as strong as titanium. Does not discuss weight and viability of new high strength plastics. Does not examine low cost imports from China which are increasingly available. NHTSA costs are outdated from relatively small suppliers. Fully energy absorbing guards in quantity should cost in the $500 range.

VC-Compat Improved Rear Underride Guard Costs in EU: Current RUP devices cost 100 € – 200 € (Approx. $134 – $268) per vehicle. Additional costs ranging from 20 € to 100 € (Approx. $27 to $134) are estimated for ‘low profile’ improved RUP, while additional costs for more complex folding devices may exceed 200 € (Approx. $268) per vehicle.

Upper Great Plains Transportation Institute at NDSU: “Cost-benefit analysis shows that the rear-guard safety equipment has injury severity benefits that far outweigh equipment cost. Given a 10 percent reduction in injury severity attributed to the rear-guard devices on agricultural trucks, in the relevant crash population, the benefit is estimated to be $14.4 million over the seven-year depreciable life of a truck. Total equipment and maintenance cost for the North Dakota agricultural truck fleet is estimated to be $8.1 million. An estimated safety benefit of $1.76 is generated from each dollar spent on rear guards for North Dakota’s agricultural truck fleet.”

As such, NHTSA does not intend that this proposed rule would pre-empt state tort law that would effectively impose a higher standard on motor vehicle manufacturers than that established by today’s proposed rule. Establishment of a higher standard by means of State tort law would not conflict with the minimum standard proposed here. Without any conflict, there could not be any implied preemption of a State common law tort cause of action.

Real world cases show these laws are used to exclude evidence against trucking companies and make lawsuits more expensive for victims. Strict Liability is almost impossible.

CMVSS No. 223 permits an option that a rear impact guard does not have to meet energy absorption requirements if it is able to resist 700,000 N of force using the distributed load application device without deflecting more than 125 mm.

Modern cars can survive crashes into solid walls at more than 40 mph (64 km/h) including in modest offset impacts. You basically allow the cars safety systems to absorb all of the crash energy and build the guard very strong and rigid to prevent PCI.

VC-Compat: “the energy absorbing capability and capacity of passenger car front structures has improved to such an extend that impact speeds up to 64 – 75 km/h (Approx. 40 mph to 47 mph) may well be survivable for passenger car occupants in collision with rigid FUPs.“

If you must give up on high-speed energy absorbing guards that will save more lives and prevent more injuries for the same cost then this might make sense. New guard designs extended from the truck are computer modeled to perform at 67 mph. Once again violating Vision Zero to kill more victims to gain modest cost-savings.



“Research into rear underrun crashes has shown that an underrun barrier needs to absorb a minimum of 50 kJ, though preferably the energy dissipation should be 100 kJ.”

“A final crash test at 75 km/hr was carried out at Melbourne’s Autoliv facility, once again with passenger and driver crash test dummies as shown in Figure 12. The mass of the vehicle in this instance was around 1350 kg. The energy dissipated by the barrier was 50 kilojoules over a distance of 300 mm. Barrier forces of the order of 500 kN were measured. The comparison of the Hybrid III results at 75 km/h with the equivalent 56km/h NCAP full frontal barrier test, indicates that the energy absorbing system reduced the crash severity for the front passenger (HIC of 1205 vs 1223; chest deceleration of 48 g) to that of a 56km/h impact. For the driver the impact was a little more severe than the NCAP test (HIC 1842 vs 1499; chest deceleration of 56 g; high femur load of 14kN).”

U.N. Comment: (Increasing energy absorption reduces the strain on trailer frames and mounting hardware for high-speed crashes. High-speed crashes require high energy absorption (imagine bigger pillow) to decrease deceleration forces on car passengers i.e. a bigger pillow allows surviving higher speed crashes!)


However, the full overlap crash test results indicate that trailers that have the main vertical supports for the guard more outboard may not perform as well in full overlap crashes as trailers that have the vertical supports more inboard for crash speeds greater than 56 km/h (35 mph). Since full and 50 percent overlap crashes are more frequent than low overlap (30 percent or less) crashes, and since most fatal light vehicle impacts into the rear of trailers are at speeds greater than 56 km/h (35 mph), such guard designs may reduce protection against PCI in higher speed full and 50 percent overlap crashes.

NHTSA in 1992 successfully tested rounded aluminum honeycomb guards that deflected cars at high speed in severe offset crashes.

Offset crashes can kill in even a few inches of overlap with the trailer. It is common to have guards not extend to the outside edges of the trailer, when the car hits just a few inches from the outside edge it can completely miss the outside edge of the guard and at a slight angle can underride the trailer up to five feet until it contacts tires with catastrophic results. These offset from center crashes are common and the force of the impact can be concentrated in the outside inches of the guard away from the reinforcing struts that hold the guard bar to the trailer. It is important to have the guard extend all of the way to the edge of the trailer with angled high strength reinforcing struts close to the ends of the guard, probably not more than 4 to 6 inches from the trailer edge. Moving the P1 force test location further out to the edge of the guard such as IIHS recommends will help to guarantee the ends of the guard are strong for these offset crashes. We must require better bracing than vertical struts provide! The entire guard needs to be strengthened.












UNICAMP Impact Project in Brazil energy absorbing guard test
Angled high strength reinforcing struts to outside edge

SPEEDS ABOVE 30 MPH – John E. Tomassoni

“Comment: It is clear that offset impacts will result in greater underride magnitudes than in centric impacts, all else being the same. Underride is also expected to increase with increasing offset. But impacting vehicle rotation will also occur in offset impacts. This will, of course, depend upon the amount of offset and the interacting structural properties. It is very likely that the occupant responses will be less than with centric impacts, but this will be only if the occupant head and torso are not contacted by the intruding structure. Injury measures, however, will be greater for the occupant on the impacted side. It is possible that vehicle rotation can be either clockwise or counterclockwise depending upon the strengths of the interacting vehicle front structure and the guard. If the guard offset strength is less than the engaged portion of the car crush strength then the guard will deform and may cause the car to rotate with its front deflecting somewhat away from the centerline. On the other hand, if the. guard offset strength is greater than the car crush strength, then car rotation will be in the opposite direction where its rear end will displace away from the centerline. See offset impact data contained in Refs 5, 6 and 8 which indicate that a guard total strength of greater than 45,000 pounds is needed for adequate offset impact protection. It is expected that certain offset conditions could result in car rotation such that the passenger compartment may beneficially avoid intrusion entirely. The performance of the MCG has not yet been demonstrated by test for offset or angle impacts.”


The Canadian standard is only effective to about 30 mph (48 km/h) with minimum guards in offset impacts with a single guard manufacturer tested successfully to 35 mph (56 km/h) in offset tests. If we harmonized to the weak Canadian Standard most trucks and trailers will only meet the minimum requirements and we would be stuck for twenty years with an 30 mph (48 km/h) standard that protects only a quarter of victims. Increasing energy absorption requirements to 50 kj minimum would force adoption of energy absorbing guards that we feel strongly will save lives at 50 mph (80 km/h) and more with little extra weight due to modern materials and for similar cost. NHTSA must begin a robust crash testing program for guards to enhance safety design and encourage adoption of better performing designs.


Recommended Reading:

ECF Position on Safer HGV Cabs
Concept Design of a Crash Management System for Heavy Goods Vehicles
fka – Design of a Tractor for Optimised Safety and Fuel Consumption
Volvo European Accident Research and Safety Report 2013
Chalmers Car-to-Truck Frontal Crash Compatibility

Improved Crashworthy Designs for Underride Guards
American Standards Rear Guards – Comments Luis Otto
American Standards Rear Guards
Tomossoni NHTSA Crash tests 1992


The Underride Network supports a high speed standard such as recommendations from MUARC in Australia would present.


1.    Barrier test Forces:

P1 (outer edge)   P2 (off centre)   P3 (centre)

              200 kN          200 kN          100 kN

2.    Barrier height: 400mm

3.    Barrier width: Within 100mm of the outer frame of the rear of the truck

4.    Energy absorption: 50kJ minimum



Underride Network position on high-speed crash tests:

NHTSA crash tests new cars to rate their safety in crashes and publishes performance based on a star system thru their NCAP rating for new cars.  You might buy a car that is 5 star crash rated that passed with flying colors the 35 mph or 56 km/h offset crash test. Why are trucks exempted? Undo influence from lobbyists?

When we crash test guards at 30 to 35 mph (48 to 56 km/h) we get guards for 50 years that perform at 30 to 35 mph (48 to 56 km/h). When you try something over and over and over again and get a negative result, why would you continue this activity. If we crash test guards at high speeds perhaps we will see guards that perform at high speeds. The FHWA tests crash attenuators in 62.2 mph (100 km/h) crash tests (Real World Crash Speeds) and attenuators protect cars and trucks in crashes at 62.2 mph (100 km/h) and more!

The FHWA Office of Safety considers that a 100 km/h (62.2 mph) crash test is representative of worst case run-off-road crashes. We agree, real world fatal crashes happen on 50 mph (80 km/h) roadways and between 50 and 60 mph (80 and 97 km/h). If we test at real world crash speeds we will get underride protection that performs at these speeds. When guards fail tests at real world speeds manufacturers will finally feel public and political pressure to increase crash effectiveness. We must see real world tests of guards at 50 and 62.2 mph (80 and 100 km/h) such as tests at FHWA for crash attenuators. We must use more extensive crash test criteria such as those used in the AASHTO Manual for Assessing Safety Hardware (MASH). MASH tests cars to 100 km/h or 62.2 mph and big trucks at 80 km/h or 50 mph. MASH tests crash attenuators at various speeds and we should do the same for underride guards to give the public a real world picture of their safety.

The Underride Network supports a similar criteria for underride guard crash tests as those submitted by Prof. Raphael Grzebieta and (Adj) Associate Professor George Rechnitzer and Transport and Road Safety (TARS) Research Centre in Australia based on the criteria used for MASH crash tests in the AASHTO Manual.


We would submit requiring multiple speed tests to include real world crash speeds and would not limit extension of guards to increase crush or stroke distance to increase guards effective speed while diminishing deceleration forces. Tests might be performed at 44 mph and 50 mph (71 and 80 km/h) and 62.2 mph (100 km/h) to test minimally compliant guards in low-speed test and using higher speed tests to monitor performance at real world crash speeds. We support testing for Practical Worst Case (PWC) scenario crashes that happen in the real world just as MASH includes PWC in it’s crash test series. We must include tests of offset controlled after crash direction of vehicle spin or VRU (Vulnerable Road User or bikes and pedestrians) after crash spin to assess high-speed crash avoidance for cars and prevention of running over VRU users in frontal crashes. John E. Tomassoni “It is expected that certain offset conditions could result in car rotation such that the passenger compartment may beneficially avoid intrusion entirely”. WE would encourage annual NCAP type testing of truck and trailer underride guards to encourage industry improvement of guards on an annual basis such as crash performance of cars improves on an annual basis using publication of the results of NCAP tests for cars to increase sales of better performing products.


Chalmers Car-to-Truck Frontal Crash Compatibility

Increasing the length of the HC (HoneyComb shaped front nose underrun guard) increases the critical impact speed, it requires extending the guard out from the truck or trailer 600 mm or about 2 feet to withstand high-speed impacts. We have fought for over 50 years for these life-saving 2 foot extensions, they require little cost and very little weight and yet would save so many lives. Two feet of honeycomb aluminum or piston extensions to save lives in highway-speed crashes, stiff guards kill, period!

To 95 km/h (59 mph) with a 300 mm – length HC structure (+27%)
To 102 km/h (63 mph) with a 600 mm – length HC structure (+36%)
To 107 km/h (67 mph) with a 900 mm – length HC structure (+43%)