February 16, Dear Administrator Rosekind:

Transcription

1 February 16, 2016 The Honorable Mark R. Rosekind, Ph.D. Administrator National Highway Traffic Safety Administration 1200 New Jersey Avenue, SE Washington, DC Federal Motor Vehicle Safety Standards 49 CFR Part 571, Rear Impact Guards, Rear Impact Protection; Notice of Proposed Rulemaking; Docket No. NHTSA Dear Administrator Rosekind: In February 2011, the Insurance Institute for Highway Safety (IIHS) petitioned the National Highway Traffic Safety Administration (NHTSA) to upgrade the Federal Motor Vehicle Safety Standards (FMVSS) on rear impact protection for semi-trailers (49 CFR , 224). The petition was based on IIHS research of real-world large truck rear impact crashes as well as crash tests of semi-trailers conducted by IIHS. NHTSA has responded by issuing a Notice of Proposed Rulemaking (NPRM) to upgrade FMVSS 223 and 224 by adopting the Canadian underride guard requirements. The agency also has tentatively proposed to require the guard and its attachments to remain intact during testing. While these changes represent an improvement over the current U.S. regulation, they fall short of addressing most of the concerns raised by IIHS in its petition. NHTSA has denied IIHS s requests to improve protection in small overlap crashes, require guards to pass the strength tests when fitted to a trailer, and reduce the number of trailers exempt from the standards. This is despite the fact that IIHS research has found each of these changes already have been adopted by some manufacturers, demonstrating their feasibility. As outlined below, NHTSA s denials were based on questionable assumptions or were reached after analyzing University of Michigan Transportation Research Institute (UMTRI) telephone survey data to a level that UMTRI previously cautioned against. Because of these denials, and because 93 percent of new semi-trailers sold in the United States already comply with the Canadian standard, the overall effect of what NHTSA has proposed will be minimal. IIHS is concerned that the proposed rulemaking misses an opportunity to substantially improve the rear underride protection of trailers on U.S. roads. Quasi-static force requirements Currently, FMVSS 223 requires that an underride guard design be capable of sustaining specific force levels in quasi-static testing at three locations on the guard. The main component of the NPRM is the plan to replace the test with the highest force requirement, currently applied at or near the intersection of the horizontal member of the guard and one of the verticals, with the Canadian Motor Vehicle Safety Standard (CMVSS) 223 distributed load test. While the distributed load test allows both vertical members of the guard to contribute to the tested strength, the minimum force level would be increased from 100 kn to 350 kn. IIHS supports the decision to increase the force requirements, but believes even higher levels are feasible and would provide better protection. In its 2011 petition, IIHS included results from crash tests and quasi-static compliance tests of three guards. The petition requested that NHTSA consider rulemaking that would ensure all guards were as strong as the Wabash guard, which exceeded the CMVSS force requirement by more than 70 percent. In its NPRM, the agency did not consider force requirements beyond those in CMVSS 223, stating that testing conducted by IIHS and Transport Canada with small and midsize cars show that the Canadian

2 Mark R. Rosekind February 16, 2016 Page 2 compliant guards are able to prevent passenger compartment intrusion in 56 km/h light vehicle impacts into the rear of trailers with 100 percent and 50 percent overlap with the guard. IIHS believes that protection for larger vehicle classes and at higher speeds could be possible. IIHS has used a midsize car in its testing while Transport Canada set the CMVSS 223 force levels based on testing with a small car. Even more fundamentally, however, there are two reasons why simply meeting the CMVSS 223 requirements does not guarantee that a guard will prevent underride in the crash scenarios NHTSA has claimed. First, it is possible that a guard design could meet the force requirements in the CMVSS 223 quasi-static test while being damaged to an extent that limits its ability to continue to absorb the energy of a crash. The IIHS petition identified the 2007 Vanguard trailer as an example of such a case. While the guard exceeded the CMVSS 223 force requirements, several of the bolts attaching the horizontal and vertical members of the guard were sheared in half during the quasi-static test. In a 56 km/h crash test, the same behavior was observed with the result that a midsize car sustained severe underride of the Vanguard. IIHS encourages NHTSA to adopt the NPRM s tentative proposal to require all portions of the guard and attachments to remain intact during FMVSS 223 certification tests, and to refine the language in the regulation to make this more explicit. For example, the NPRM states that partial separation is acceptable as long as there is still some degree of mechanical connection between the members. Would NHTSA consider a joint where 3 of 4 bolts are sheared to be a partial separation? IIHS believes that robust guard designs are those that withstand the quasi-static test loads without any separation of the attachments. Modifying the regulation accordingly would promote designs that are more effective in realworld crashes. Certifying guards while attached to trailers The second reason CMVSS 223 force requirements alone are insufficient to guarantee underride prevention, even in 56 km/h center crashes with small or midsize cars, is that the guards can be certified while attached to a rigid fixture rather than an actual trailer. NHTSA has denied IIHS s request to make this change based on several flawed arguments. The agency stated that trailer deformation during crash testing does not indicate that the guard-attachment-trailer system is weaker than a guard-attachmentfixture system, but that the trailer structure along with the guard offered resistance to the dynamic loads and that is why the trailer structure also deformed. The agency also claimed that testing on a rigid fixture could be more stringent because, When the guard is attached to a rigid fixture, it has to resist all the loads and absorb all the energy, whereas when it is installed on a trailer, the designs could be such that the trailer structure could resist a portion of the load. NHTSA is misunderstanding the problem. The question is not which scenario is more stringent for the guard itself, but which scenario better represents the real-world loading case. In both real-world crashes and in compliance testing, whatever the guard is attached to will offer resistance to the applied load. A rigid fixture offers resistance to the quasi-static test load. However, the fixture does not permanently deform because it has a higher strength than the guard itself. The problem arises when the same guard, which has been certified to a specific force level, is then attached to a trailer structure that does not have the demonstrated capability to resist the same force level. In a real crash, the deformation will occur at the weakest point of the guard-attachment-trailer system, and it is possible that the force levels mandated by the standard will never be realized because of this deformation. Figures 1-3 provide an example of this behavior. Figure 1 shows a Hyundai Translead underride guard attached to a rigid fixture prior to CMVSS 223 certification testing by the manufacturer. Figure 2 shows the same guard after certification testing, reoriented as it would be attached to a trailer. The main point at which the deformation occurred was near the center of both vertical members. Figure 3 shows the same guard design after being crash tested by IIHS. There was no visible deformation of the vertical members of the guard. Rather, the main deformation occurred at the trailer cross-members and slide rails. The

3 Mark R. Rosekind February 16, 2016 Page 3 difference between deformation patterns indicates that in the crash test, the peak forces on the guard (and the guard s corresponding resistance to the striking vehicle) likely never reached the force levels required by the compliance test. In this specific case, despite the deformation there was no separation of the trailer structure and underride was prevented. However, by allowing rigid fixture testing, NHTSA has left open the possibility that guards will be attached to trailer structures that are too weak to withstand the forces of a crash, in which case the strength of the guard itself is irrelevant. In fact, to the extent that trailer manufacturers fit stronger guards to the same trailers, trailer weakness will become a proportionally greater problem than what IIHS previously observed in its study of real-world rear underride crashes (Brumbelow and Blanar, 2010). Figure 1 Hyundai Translead underride guard on rigid fixture prior to CMVSS 223 compliance test. The arrow indicates the horizontal member of the underride guard. Figure 2 Hyundai Translead underride guard on rigid fixture after CMVSS 223 compliance test. Image is reoriented to represent guard in position on a trailer.

4 Mark R. Rosekind February 16, 2016 Page 4 Figure 3 Hyundai Translead underride guard and trailer after 56 km/h midsize sedan center impact crash test. NHTSA also stated that in its own compliance tests, the rear impact guards contain part of the trailer frame rails and/or cross beams to which the rear impact guard is attached. While NHTSA may obtain guards in this configuration, it is not required by the standard so rigid fixtures remain an option for selfcertification (e.g., Figure 1). Additionally, without requirements for how a partial trailer section should be held, the test may be no more representative of the real-world crash loading than a completely rigid fixture. For example, if the underride guard shown in Figure 3 had been certified while attached to the two rear cross-members, it would be necessary that any fixture allowed those cross-members to rotate as they did in the crash test. NHTSA is opposed to removing the option for rigid fixture testing partly because it is expensive to conduct a full trailer test, which is a destructive test. While some trailer manufacturers voluntarily conduct full trailer tests (Manac and Stoughton, personal communication), as stated in its 2011 petition IIHS believes at a minimum, [guards] should be attached to sections of the trailer rear that include all the major structural components and that are constrained far enough forward that the load paths near the guard are not changed. Offset crash protection IIHS evaluated real-world rear underride crashes in the Large Truck Crash Causation Study (Brumbelow and Blanar 2010). There were 30 cases involving guards that met the FMVSS 224 geometric requirements. Of these, 30 percent (n=9) were crashes in which less than half of the passenger vehicle overlapped the trailer. In most of these cases, the vertical member of the guard did not engage the passenger vehicle and underride occurred, resulting in passenger compartment intrusion. Based on this finding, IIHS included a request in its petition that NHTSA adjust the outboard test location on the guard in order to improve offset crash protection.

5 Mark R. Rosekind February 16, 2016 Page 5 NHTSA has declined this request, stating that offset crashes appear to represent a small portion of the rear underride fatality problem. NHTSA based this conclusion on a report the agency commissioned from UMTRI (Blower and Woodrooffe, 2013). This analysis was an extension of the 2008 and 2009 Trucks in Fatal Accidents (TIFA) surveys conducted by UMTRI. These data were collected during phone interviews with someone who was familiar with each crash but may not have been at the crash scene, such as the truck owner or the truck carrier s safety director. In addition, the interviews took place 1-2 years after the crash. In their own earlier study using similar methods, Blower and Campbell (2000) stated, Collecting the data by means of telephone interview with people on the scene well after the fact probably is not sufficient to accurately measure degrees of underride. The inconsistency in the survey data is demonstrated by the fact that the 2013 report indicated that offset crashes had higher crash speeds than no-offset crashes, resulted in a slightly higher incidence of crashes with underride extent beyond the plane of the windshield, yet had a lower percentage resulting in major guard damage. Despite these inconsistencies, and UMTRI s previous warning against overanalyzing the telephone survey data, NHTSA is basing its decision to not promote improved offset crash protection on the survey s lower reported rate of offset crashes compared with center impacts (40 percent vs. 60 percent) as well as the difference in the proportions of offset and center crashes that were described as producing major guard damage (39 percent and 49 percent, respectively). While it is problematic to rely on such data in the first place, it also is unclear how someone who actually observed a trailer after an offset crash would answer the survey question about guard damage. On its 2009 survey form, UMTRI defined major guard damage as significant bending from original position, >45 degrees; includes torn off (Blower and Woodrooffe, 2013; it should be noted that the 2008 survey did not include definitions of what represented different levels of guard damage). Would someone responding to the survey have considered the bending of the guard as a whole (including the vertical members) or just the loaded end in an offset crash (e.g., Figure 4)? Furthermore, guard damage is not an adequate metric for the occurrence or severity of underride. Figures 4 and 5 show the same guard design after a 30 percent overlap and center impact crash test, respectively. While the guard damage could be defined as more severe in the center impact test, there was no underride in this case. In the 30 percent overlap test, there was severe underride with the driver dummy s head striking the rear of the trailer. NHTSA raised an additional concern that attempting to improve offset protection would not benefit safety overall. The agency based this on the IIHS crash test results. At the time the NPRM was issued, the 2012 Manac trailer was the only tested design able to prevent underride and passenger compartment intrusion in a 30 percent overlap crash. NHTSA was concerned that the Manac would not perform as well as other designs in center impact crashes at speeds greater than 56 km/h. This point is incompatible with NHTSA s approach to the NPRM as a whole. The Manac trailer complies with the CMVSS 223 force requirements that NHTSA is proposing. In addition, it performed well in all three IIHS crash test configurations (center impact, 50 percent overlap, and 30 percent overlap) with the same 56 km/h impact speed at which NHTSA expects their updated standard to be effective. To claim that the Manac design is deficient at higher impact speeds and, therefore, that its improved offset protection is an overall disbenefit to safety is unfounded and incongruous. It is unfounded because none of the guards were tested at higher speeds, and based on current performance it is difficult to say how the 2013 Strick trailer (Figure 5), for example, would perform better than the 2012 Manac in higher speed center impacts. And the claim is incongruous with the overall rule because if NHTSA s goal is to prevent underride in center impacts at speeds beyond 56 km/h, then the agency needs to propose force levels that go beyond the CMVSS 223 requirements. Additionally, NHTSA s belief that the guard configuration utilized by Manac is inherently less safe than other guard designs even in 50 percent overlap crashes is disputed by the fact that it was the only design tested by IIHS with the vertical member far enough outboard to overlap the transmission or engine of the striking vehicle in the 50 percent overlap condition.

6 Mark R. Rosekind February 16, 2016 Page 6 Figure Strick trailer and underride guard after 56 km/h midsize sedan 30 percent overlap crash test. Figure Strick underride guard after 56 km/h midsize sedan center impact crash test. Since petitioning NHTSA, IIHS has continued to conduct crash tests of trailers and their underride guards. Some trailer manufacturers have voluntarily made changes to their guard designs in order to improve protection in rear impacts. In January 2016, after NHTSA released its NPRM, IIHS crash tested a 2015 Vanguard trailer with the manufacturer s newest underride guard design in the 30 percent overlap

7 Mark R. Rosekind February 16, 2016 Page 7 condition. The new design was able to prevent severe underride with passenger compartment intrusion (Figure 6; see also attached crash test report), joining the 2012 Manac as only the second such result. Manac achieved this performance by locating the vertical members of the underride guard farther outboard than the other trailer designs. While NHTSA s concern with this strategy is not consistent with the rulemaking as a whole, the performance of the redesigned Vanguard demonstrates that there are other options for improving offset protection than the approach taken by Manac. Figure Vanguard trailer and underride guard after 56 km/h midsize sedan 30 percent overlap crash. Exemptions In its petition, IIHS requested that NHTSA reduce the number of truck and trailer types exempt from the underride regulation. This request was based on the fact that more than half of the truck units in the realworld crashes studied by IIHS were exempt. Wheels-back trailers and straight trucks accounted for most of these exemptions. The agency has issued a separate Advanced Notice of Proposed Rulemaking (ANPRM) discussing the possibility of requiring straight trucks to be equipped with underride guards. The ANPRM indicated that such a rule would not be cost-effective. However, IIHS has submitted comments to NHTSA (2015) expressing its concern that the ANPRM overestimates the costs and underestimates the benefits of requiring straight trucks to be equipped with underride guards. With regard to trailers, the current NPRM declines to reduce the number of exempt types. IIHS believes this decision was based on questionable data and that removing the exclusion for wheels-back trailers would be feasible and beneficial. According to the UMTRI TIFA data, more than one-fifth of the trailers involved in fatal rear-end crashes were wheels-back and therefore exempt from the underride regulation. Despite this substantial proportion, NHTSA believes that excluding them may not have significant safety consequence based on UMTRI s telephone survey data, which reported only 16 percent of fatal crashes into the rear of wheelsback trailers involved severe enough underride to produce intrusion of the passenger compartment. As noted above, UMTRI previously cautioned against defining degrees of underride using the telephone

8 Mark R. Rosekind February 16, 2016 Page 8 survey data. NHTSA further justified its decision to continue excluding wheels-back trailers using estimated crash speeds. This figure also is based on the UMTRI report. Without vehicle crush measurements (or truck stiffness data), the speed distribution was calculated using a method that relied on the reported pre-skidding travel speeds recorded on police crash reports or during interviews. Reported travel speeds prior to any crash must be considered speculative. This is even more the case for crashes in the UMTRI report because the driver of the striking vehicle often was killed. Interestingly, the TIFA found that one-half of wheels-back trailers involved in any fatal crash (not just rear) were equipped with an underride guard despite being exempt. This raises several important points. First, it suggests the feasibility of requiring guards on all wheels-back trailers. Second, by including these trailers in the regulation, they would then need to be certified to the same strength requirements as other guards. The guards already fit to wheels-back trailers have unknown strength levels and effectiveness at reducing underride occurrence. Third, the presence of so many wheels-back trailers with guards in the TIFA data adds a further concern to those already mentioned about the validity of the comparisons of underride severity by trailer type. Conclusion In conclusion, IIHS believes rear underride protection for large trucks and trailers must be improved. Although NHTSA s NPRM to upgrade the current FMVSS 223 strength requirements is welcome, it falls short of ensuring meaningful improvement in rear underride protection. NHTSA should incorporate testing of guards on trailers, improve offset protection, and allow fewer exemptions from compliance. The tentative plan to prohibit separation of the guard components during testing should be adopted after making it more explicit. IIHS s crash testing and other research has found that these changes already have been implemented by some trailer manufacturers, demonstrating their feasibility. Each of these issues represents an opportunity for an upgraded standard to reduce the number of deaths and injuries in rear underride crashes. Sincerely, Matthew L. Brumbelow Senior Research Engineer Attachment Insurance Institute for Highway Safety Large truck underride research: 2010 Chevrolet Malibu front into 2015 Vanguard trailer rear. Report no. CF Arlington, VA. References Blower and Campbell Underride in fatal rear-end truck crashes. SAE Technical Paper Warrendale, PA: Society of Automotive Engineers. Blower and Woodrooffe Heavy-vehicle crash data collection and analysis to characterize rear and side underride and front override in fatal truck crashes. Report no. DOT HS Washington, DC: National Highway Traffic Safety Administration.

9 Mark R. Rosekind February 16, 2016 Page 9 Brumbelow and Blanar Evaluation of US rear underride guard regulation for large trucks using rear-world crashes. SAE Technical Paper Warrendale, PA: Society of Automotive Engineers. National Highway Traffic Safety Administration Comment from the Insurance Institute for Highway Safety on ANPRM Underride Protection of Single Unit Trucks. Docket Document ID NHTSA Washington, DC: U.S. Department of Transportation.

10 Insurance Institute for Highway Safety Large Truck Underride Research 2010 Chevrolet Malibu Front into 2015 Vanguard Trailer Rear 56 km/h; 30 percent overlap CF Vanguard Vehicle identification number: Body style: Empty weight: Test weight: 2010 Chevrolet Malibu Vehicle identification number: Body style: Engine/transmission: 5V8VC5320FM ft. dry van semi-trailer 6,384 kg 25,084 kg 1G1ZB5EB5A Midsize 4-door sedan Transverse 2.4-liter 4-cylinder, 4-speed automatic, front-wheel drive Vehicle Specifications (Provided by Manufacturer) Wheelbase: 285 cm Overall length: 487 cm Overall width: 179 cm Curb weight: 1,549 kg Vehicle Specifications (Measured) Curb weight: 1,558 kg Test weight: 1,707 kg (56% front, 44% rear) Nominal Test Parameters 56.3 km/h, 30% overlap of Malibu s width with Vanguard s rear underride guard Dummy Seating Protocols IIHS Guidelines for Using the UMTRI ATD Positioning Procedure for ATD and Seat Positioning (Version V) (IIHS, 2004) Crash Test Date January 20, 2016 February

11 Figure 1 Video Frame Capture 2015 Vanguard Trailer and 2010 Chevrolet Malibu Figure 2 Postcrash 2015 Vanguard Trailer and 2010 Chevrolet Malibu February

12 Summary On January 20, 2016, the Insurance Institute for Highway Safety (IIHS) conducted a crash test with a 2015 Vanguard 53-foot dry van semi-trailer and 2010 Chevrolet Malibu. The front of the Malibu struck the rear of the Vanguard at 56.2 km/h with a 30 percent overlap with the trailer s underride guard. The Vanguard was stationary and connected to a 2001 Kenworth tractor. A Hybrid III 50th percentile male dummy was positioned in the driver seat of the Malibu with the lap/shoulder belt fastened. During the crash, the horizontal and right outboard vertical members of the Vanguard s underride guard bent forward and were detached from each other. However, both members remained attached to the trailer and engaged with the Malibu, mitigating underride. The Vanguard s rear sill pushed the Malibu s hood into the windshield and directly loaded the driver side A-pillar. However, the Malibu s hood did not penetrate the windshield and A-pillar deformation was limited. The driver dummy was restrained by the three-point lap/shoulder belt and airbag. Analysis of the high-speed film indicated the maximum forward excursion of the Malibu s center of gravity relative to the Vanguard was approximately 171 cm. Measures of intrusion taken after the crash indicated that maximum deformation of the Malibu s A-pillar was 12 cm on its leading edge and 4 cm on the rearward (doorframe) edge. Deformation of the roof was negligible. The post-crash deformation of the horizontal member of the Vanguard s underride guard was 38 cm at the right outboard end. Test Conditions The pretest setup followed the Crashworthiness Evaluation Offset Barrier Crash Test Protocol (Version XIII) (IIHS, 2004) with several deviations to account for the different test configuration, including: In place of a deformable barrier face with slotted bumper, the Malibu struck the rear of a 2015 Vanguard 53-foot semi-trailer (Figures 3 and 4); The Malibu was aligned to produce a 30 percent overlap on the driver side with the horizontal member of the Vanguard underride guard at impact; The target test speed was 56.3 km/h; and Two onboard cameras with their respective power supplies and control units were installed on the Malibu. The Vanguard trailer was connected to a 2001 Kenworth tractor. The trailer was loaded with concrete blocks totaling 18,700 kg. The sliding rear axles were placed in their middle position, with a 192 cm clearance between the rear of the trailer and the rearmost surface of the rear tires. The ground clearance of the underride guard was 45.7 cm. Prior to the test, the trailer s brakes were pressurized to 40 psi to simulate being stopped in traffic. The Malibu s driver seat back, steering column adjustments, and Hybrid III dummy seating parameters were set according to the Guidelines for Using the UMTRI ATD Positioning Procedure for ATD and Seat Positioning (Version V) (IIHS, 2004). After final positioning of the dummy, measurements from various parts of the dummy to a number of vehicle interior points February

13 were made. These measurements and the seat back, shoulder belt upper anchorage, and steering column adjustments are described in the Appendix, Dummy Clearance Measurements. The Malibu s vehicle acceleration measurements were made by a triaxial arrangement of accelerometers mounted on the vehicle s longitudinal centerline and 72 cm behind its center of gravity (197 cm behind the front axle). The vehicle speed recorded just prior to impact was 56.2 km/h. Figure 3 Top View of Test Configuration February

14 Figure 4 Front View of Test Configuration Underride Guard Performance and Vehicle Structural Interaction Precrash static measurements indicated the Vanguard s underride guard had a ground clearance of 45.7 cm and a height of 9.8 cm. The Malibu s bumper bar had a ground clearance of 40.3 cm and a height of 12.4 cm. This resulted in a vertical overlap of 7.0 cm between the Vanguard s underride guard and the Malibu s bumper bar. Each of the guard s two main vertical support members are attached to the trailer s chassis with two bolts on the rear sill and with six bolts fastened to a crossmember that, in turn, is bolted to the bottom of the rear sill with two bolts and to the longitudinal slide rail with three bolts. There are additional vertical supports on the outboard ends of the guard that are fastened to the rear sill with one bolt and to a triangular gusset with three bolts. The gusset is bolted to the rear sill with three bolts and to the bottom of the trailer side rail with three bolts (Figure 5). Diagonal braces are attached to the inboard edge of each inboard vertical support with two bolts and to the top center of the horizontal member with two bolts. The horizontal member has flanges welded to its top surface that are attached to the inboard vertical supports with six bolts on each support; the horizontal member is attached to the outboard vertical supports with two bolts at each end. The guard installed for this test differs from that previously tested at a 50 percent overlap (CF14003). The vertical supports on the outboard ends were redesigned for additional strength. This change was implemented in January February

15 Early in the crash, the horizontal member of the Vanguard s underride guard bent forward as it was loaded by the front bumper bar of the Malibu. As the Malibu continued to move forward, its bumper bar slipped below the horizontal member of the underride guard. The guard s outboard vertical member detached from the trailer sill and the horizontal member. However, it remained attached to the outboard gusset and continued to load the Malibu. At the same time, the Malibu s front left tire/wheel assembly loaded the horizontal member of the guard near its attachment to the right inboard vertical member. Late in the crash, the Vanguard s rear sill pushed the Malibu s hood into the windshield and directly loaded the driver side A-pillar, but did not separate it. The combined loading from the different components of the underride guard was sufficient to stop the Malibu without substantial intrusion of the passenger compartment. The trailer did not interfere with the restraint of the driver dummy by the seatbelt and airbag. Resultant acceleration of the Malibu reached a maximum of 15 g at 77 ms. At approximately 194 ms, the Malibu reached its point of maximum forward excursion relative to the Vanguard. Film analysis indicated that at this time the Malibu s center of gravity had moved forward 178 cm and the Vanguard had moved forward 7 cm. Table 1 describes the major structural events in the crash along with the dummy kinematics. Figures 6 and 7 show postcrash views of the Vanguard trailer and its underride guard. Figure 8 shows precrash and postcrash views of the Malibu. Figure 9 shows precrash and postcrash measurements of the exterior A-pillar and roof header. The maximum deformation of the A-pillar was 12 cm on the leading edge. Maximum deformation of the A-pillar on the trailing doorframe edge was 4 cm. Deformation of the roof was negligible. February

16 Figure Vanguard Trailer Underride Guard Attachment to Trailer Chassis Figure Vanguard Trailer Postcrash View February

17 Figure Vanguard Trailer Postcrash View February

18 Figure Chevrolet Malibu Precrash and Postcrash Oblique Views February

19 Figure Chevrolet Malibu Precrash and Postcrash Measures of A-pillar Exterior and Roof Header (from above) Precrash Postcrash Longitudinal distance from front of vehicle (cm) Lateral distance from vehicle centerline (cm) 20 Dummy Kinematics and Restraint System Performance During the crash, the dummy's head loaded the fully-inflated frontal airbag. During rebound, the rear of the head contacted the driver head restraint. Table 1 summarizes the major crash events. Tables 2-5 summarize the peak dummy injury measures. February

20 Table 1 Vehicle and Dummy Kinematics 2010 Chevrolet Malibu and 2015 Vanguard trailer Event Time (ms) Activation of seat belt crash tensioner 22 Deployment of driver frontal airbag 34 Malibu bumper bar slips beneath Vanguard underride guard 50 (approx.) Malibu front wheel begins to load Vanguard underride guard 52 Frontal airbag fully inflated 62 Malibu's firewall begins to load Vanguard underride guard 94 (approx.) Dummy face begins loading frontal airbag 98 Malibu's front wheel slips beneath underride guard 105 Malibu s windshield contacts rear corner of trailer 116 Malibu s driver side A-pillar contacts rear of trailer 142 Maximum forward excursion of Malibu 194 (approx.) Rear of dummy's head contacts driver head restraint 326 Table 2 Head Injury Measurements 2010 Chevrolet Malibu Measure Published Tolerance Threshold Vector resultant acceleration (g), during frontal airbag loading 80 Vector resultant acceleration 3 ms clip (g), during frontal airbag loading 80 Head Injury Criterion (HIC) Head Injury Criterion 15 ms interval (HIC-15) Result Time (ms) February

21 Table 3 Neck Injury Measurements 2010 Chevrolet Malibu Published Tolerance Threshold Result Time (ms) ± Axial compression force (kn) Axial tension force (kn) Nij Tension-Extension Nij Tension-Flexion Nij Compression-Extension Nij Compression-Flexion Flexion bending moment (Nm) Extension bending moment (Nm) Measure A-P shear force (kn) Table 4 Chest Injury Measurements 2010 Chevrolet Malibu Measure Vector resultant spine acceleration 3 ms clip (g) Published Tolerance Threshold Result Time (ms) Rib compression (mm) Viscous criteria (m/s) Sternum deflection rate (m/s) February

22 Table 5 Leg and Foot Injury Measurements 2010 Chevrolet Malibu Published Tolerance Threshold Left Time (ms) Right Time (ms) Femur axial force (kn) 9.1* Tibia-femur displacement (mm) L-M moment (Nm) ± A-P moment (Nm) ± Vector resultant moment (Nm) Index Measure Upper Tibia Lower Tibia L-M moment (Nm) ±225* A-P moment (Nm) ±225* Vector resultant moment (Nm) 225* Axial force (kn) 8.0* Index A-P acceleration (g) ± I-S acceleration (g) ± Foot Vector resultant acceleration (g) * These published thresholds are for fractures of the tibia. Ankle and foot injuries have been associated with bending moments as low as Nm, and heel fractures have been associated with axial forces as low as 6.0 kn. February

23 References Backaitis, S.H. and Mertz, H.J. (eds) Hybrid III: The First Human-Like Crash Test Dummy. Warrendale, PA: Society of Automotive Engineers. Begeman, P.C. and Prasad, P Human ankle impact response in dorsiflexion (SAE ). Thirty-fourth Stapp Car Crash Conference Proceedings, Warrendale, PA: Society of Automotive Engineers. Begeman, P.; Balakrishnan, P.; Levine, R.; and King, A Dynamic human ankle response to inversion and eversion (SAE ). Thirty-seventh Stapp Car Crash Conference Proceedings, Warrendale, PA: Society of Automotive Engineers. Insurance Institute for Highway Safety Crashworthiness evaluation offset barrier crash test protocol (version XIII). Arlington, VA. Insurance Institute for Highway Safety Guidelines for using the UMTRI ATD positioning procedure for ATD and seat positioning (version V). Arlington, VA. Mertz, H.J. and Patrick, L.M Strength and response of the human neck (SAE ). Biomechanics of Impact Injury and Injury Tolerances of the Head-Neck Complex, Warrendale, PA: Society of Automobile Engineers. Parenteau, C.S Foot-ankle injury: epidemiology and method to investigate joint biomechanics. Gothenburg, Sweden: Chalmers University of Technology. Prasad, P. and Mertz, H.J The position of the United States delegation to the ISO Working Group 6 on the use of HIC in the automotive environment (SAE ). Biomechanics of Impact Injury and Injury Tolerances of the Head-Neck Complex, Warrendale, PA: Society of Automotive Engineers. Transport Canada Motor Vehicle Safety Regulations Canadian Motor Vehicle Safety Standards, Schedule IV Part III Standard 208, Occupant Restraint Systems in Frontal Impact. Ottawa, Ontario. Welbourne, E.R Vehicle performance requirements for head injury protection: a comparison of the head injury criterion with an 80 g limit on resultant acceleration. Technical Memorandum. Ottawa, Ontario: Transport Canada, Vehicle System Division. Zeidler, F The significance of lower limb injuries of belted drivers. Journal of Orthopedics [German]. February

24 Dummy Clearance Measurements Test Number: Vehicle: Seat Type: Upper Belt Anchorage: Steering Column Adjustment: Foot Pedal Adjustment: CF Chevrolet Malibu Adjustable bucket seat (manual fore/aft, electric height, and manual seat back angle) Set to topmost of 4 positions Tilt adjustment set to midpoint position and telescopic set to midpoint of range Fixed Location Code Measure Location Code Measure Head to header Nose to rim Chest to dash Rim to abdomen Knee to dash, left Knee to dash, right Steering wheel to chest, horizontal Steering wheel to chest, reference Hub to chest, minimum Pelvic angle Seat back angle Torso recline angle (H-point to Head CG) Neck bracket angle Neck angle, seated HH NR CD RA KDL KDR SCH SCR HCM PA SA TRA NBA NAS Striker to CG, horizontal Striker to CG, lateral Striker to CG, vertical Striker to knee* Striker to knee angle* Striker to H-point, horizontal Striker to H-point, vertical Ankle to ankle Knee to knee Arm to door H-point to door Head to A-pillar Head to roof Head to side window CGH CGL CGV SK SKA SHH SHV AA KK AD HD HA HR HS All distance measurements are in millimeters (mm). * These measurements were made in a vertical plane containing the striker and parallel to the driver door sill. February