Is NHTSA Ready to Strengthen Seat Backs?

The National Highway Traffic Safety Administration and the automotive industry have long agreed that the Federal Motor Vehicle Safety Standard for seat back strength is inadequate. And, most safety advocates argue that FMVSS 207 has nothing to do with seat performance in real world crashes.

Today, many vehicle seats are built to far exceed the demands of FMVSS 207’s 1967 static pull compliance test on an unoccupied seat, which requires that it withstand a load equal to 20 times the weight of the seat with the load applied in a forward and rearward longitudinal direction. In 2016, David C. Viano, a former General Motors scientist, now a private engineering consultant and the industry’s go-to defense expert witness, documented a steady increase in conventional seat strength and resistance to seat rotation over the past 50 years. By the 2000s, seats were significantly stronger, even though questions about their real-world performance remain.

Yet, for half a century, the battle over FMVSS 207’s effectiveness was shaped thus: Did the standard allow seat backs to collapse in rear impacts, causing serious injury to the seated occupant and any rear-seated passenger in its path, as the safety community claimed? Or did the injury data show that seat-back collapse, euphemistically dubbed “yielding” seat backs actually protect the front-seat occupant from injury, as industry maintained?

Automakers argued that “yielding seats” limited injuries in low-speed rear impacts, and that stiffer seats would result in more severe injuries – an argument based on 1950s-era seat designs with short, flexible seatbacks and ineffective head restraints. However, in real-world, higher-speed rear impacts, weak seat backs don’t yield in a controlled manner. They simply break and collapse – a result readily observed in FMVSS 301 Fuel System Integrity rear impact tests with dummies in the driver’s seat.

Industry successfully forced this false frame – much like automakers did with their claim that serious rollover injury and death was not caused by weak, collapsing roofs but by occupants “diving” into them – around the issue, effectively stymieing any solution for decades. The concept of designing a seating system that protected occupants in both low and high-speed crashes was considered a pipe dream. 

But a recent paper published by NHTSA shows that weak seats lead to serious injuries, and there are effective countermeasures. In July, NHTSA released Front Seat Modeling in Rear Impact Crashes: Development of a Detailed Finite-Element Model for Seat Back Strength Requirements, which concluded that in rear impacts, seat back dynamic rotation should be reduced to less than 35° to prevent injury to the seat occupant and occupants seated directly behind it. NHTSA commissioned the study, conducted by EDAG, Inc., to re-examine the feasibility of increasing seat back strength by using computer finite-element (FE) modelling. EDAG’s goal was to validate the FE modelling of a current vehicle front seat design that could be used with existing anthropomorphic rear-impact test dummies to study seat performance in rear impact crashes.

EDAG researchers used the 2014 Honda Accord mid-size sedan, which they had developed for other projects, to represent typical front seats. Researchers then compared their Finite Element Analysis (FEA) computer simulations to two physical tests: a more rigorous version of the FMVSS 207 test and the dynamic FMVSS 301 Fuel System Integrity high speed rear impact crash test. For the FMVSS 207 rearward seat back pull test, the researchers loaded the seat back until it collapsed, on both a manual seat and a power seat. For the FMVSS 301 test (a 55 mph moving deformable barrier with 70 percent off-set into the rear of a vehicle), researchers substituted the barrier-to-vehicle impact with a rear impact sled test, using an acceleration pulse computed from the 2014 Honda Accord model, and placed a 50th percentile male BioRID II dummy in the front and rear seats.

EDAG used the seat back strength and dynamic motion of the front seat to determine injury potential during rear impacts, finding “significant injuries on the rear seat occupant,” in the head and the knee, both of which interacted with the front seat back as it rotated rearward. For example, in the scenario in which an unbelted occupant dummy was seated behind the driver, with seat in the full down and full rear position, the front seat rotated about 40° and hit the rear seat occupant knee. “The seat back rotation observed from this study was considered high potential to cause injuries to the rear seat occupants of all types such as children and adults.”  

These results led the researchers to conclude that a 40° seat back rotation in the FMVSS 301 sled test should be considered a failure because of the injury-producing potential, ultimately recommending that dynamic rotation of the seat back be limited to 35°.

Then, EDAG researchers used computer modelling to look at injury prevention measures to the seat back, the recliner, and the seat bottom that did not add significant weight or manufacturing costs. The FEA found that modifications to the seat back parts or recliner mechanism were ineffective, because “most of the seat back dynamic rotation was caused by weakness of the seat bottom frame parts and seat mechanism. EDAG found that a gauge increase to 3.0 mm from 1.8 mm and grade change to high strength steel on the seat bottom frames yielded the performance meeting the targets.”

This research, in part, buttresses the findings of many independent researchers, such as Dr. Kenneth J. Saczalski, Carl E. Nash, and the engineers at ARRCA, who have long argued that excessive seat rotation is a serious safety hazard to the seat occupant, and those positioned behind a collapsing front seat.

“There are some excellent points in this paper,” says Saczalski. “And I’m really surprised that NHTSA said that we need to limit the amount of rearward rotation. It was nice to see after all of this time, because the [seat back standard] doesn’t address the safety of the occupant, and we’ve published many papers showing this. You’ve got to make the seat stronger, and reliable under the various levels of use.”

To understand a vehicle seat’s safety performance, it must be adjusted and its performance assessed at the low and high position, and with dummies representing the heaviest occupants –the 95th percentile male, he adds. Some seats are strong, but when challenged by crash forces with certain occupants, and with the seats in certain positions, the small linkages fail and the seat back collapses.  

Dr. Teo Forcht Dagi, a neurosurgeon and expert in brain and spine injury, notes that “the move to reconsider the seat-back standard reflects an increasing appreciation for the importance of complex dynamic factors in injury causation and prevention. Design for passenger safety must contend with the fact that the mechanisms of injury change under different circumstances, at different velocities, and under changing loads. Not only is it incumbent upon automotive designers to re-examine their assumptions as additional data become available, they must rethink their designs and their standards for adequate safety performance.”

Other support can be found in a 2009 real-world crash study by researchers from the Center for Injury Research and Prevention at the Children’s Hospital of Philadelphia, who looked at the effect of front seat back strength on the injury risk to children seated in the rear in a rear-impact crash. CHOP examined cases from 2000-2006 of 1,035 restrained child occupants under 12 years old, seated in a second-row outboard position in rear crashes, and weighted the sample to represent 10,079 children. These data came from their Partners for Child Passenger Safety Study, a collaboration with State Farm Insurance. 

The researchers analyzed the data to quantify the overall injury risk in relation to the presence of a front seat occupant and reported front seat-back deformation. Researchers found 2.3 percent of the children sustained an AIS 2+ injury; 71 percent of those crashes had a front seat occupant and 8 percent of the cases reported front seat-back deformation. For those children with reported seat-back deformation occurring directly in front of them, there was a doubling of the injury risk. The researchers, who presented the results of this survey at the 52nd AAAM Annual conference, provided the first population-based estimates of the injury risk of rear row-seated children in rear impact crash events.

At the time of its release, Kristy Arbogast, CHOP’s  Co-Scientific Director and Director of Engineering, said  “In the automotive safety community, the debate centers around mitigating injury for front seat occupants through design of the front seat, and little focus is placed on the role of seat back design on injuries to other occupants. Using two population-based samples, this study points to at least a two-fold increase in injury risk for children seated behind yielding seat backs in rear impact crashes, after adjusting for potential confounders such as crash severity.”

Similarly, Viano’s early research into seat back strength in the 1980s included tests showing how seat back rotation injured the front seat occupant. His paper, published in 1992, documented the results of his study of the influence of seat-back angle on occupant retention in a rear impact. The paper concluded that front seat occupant retention became difficult in rear impacts once the seat back angle rotated rearward more than 45 degrees from vertical and that, at 60 degrees, occupant retention was not possible – even if a seatbelt was worn.

Rear Seat Safety Diminished as Front Seat Safety Increased

This latest paper offers another strategy for increasing rear seat safety, which has long been ignored. NHTSA, as well as seat experts, automotive injury researchers and other agencies have all acknowledged this gap in occupant protection.

In January 2017, the National Transportation Safety Board, an independent non-regulatory agency which investigates crashes and makes policy recommendations, published the results of workshop on rear-seat safety it sponsored in 2016. The report noted:

 …with the focus on advancing the safety of front seat occupants through improvements in vehicle design, regulations, and crash testing, some recent studies have indicated that the protection offered to rear seat occupants is not advancing as quickly as protection for front seat occupants. Advances in front seat design and technologies have created an environment where, for some occupants, such as older children and older adults in certain crash situations, the front seat may be safer than the rear seat. This development is in contrast to the longstanding belief that the rear seat is always the safest position for these occupants.

The workshop brought together 50 scientists, researchers, engineers, and representatives of industry to discuss short-term and long-term solutions and challenges to the many factors contributing to the trend of rear-seat injury and death, including rear seat design and low adult seat-belt usage rates in the rear.

Ten years earlier, at a May 2006 government-industry meeting sponsored by SAE, NHTSA staffer Sashi Kuppa presented an analysis of frontal (non-rollover) crashes from the FARS, NASS and State Data System databases showing that the rear seats—often touted as the safest positions in a vehicle—offer inadequate protection to their occupants. Kuppa examined rear outboard positions in 1991 and later vehicles (equipped with lap/shoulder belts) and concluded that for rear seat occupants the belt was often the source of injury. Kuppa also found that for newer cars front seats were safer for restrained adults—especially adults who were 50 years old and above—as advanced restraints were added. Other crash data analyses presented by the agency similarly concluded that rear seat belts were the major cause of injuries to rear seat occupants in frontal crashes. The results of this study showed that rear-seating positions are not receiving the same attention by manufacturers as front seats, which have had increasing scrutiny and improved regulation.  

Petition after Petition, Still No Rulemaking

Will this study move the needle on long overdue rulemaking to improve rear seat safety? The history of regulatory inaction isn’t encouraging.

Researchers have been asking NHTSA to strengthen rear seat requirements for 45 years. In 1974, Nash of the Public Interest Research Group petitioned NHTSA to upgrade FMVSS 207 by adding a dynamic rear-end impact test requirement to the standard. And, in 1998, NHTSA officially acknowledged on its website that the 207 standard was inadequate:

There have been several valid criticisms of the current Federal Motor Vehicle Safety Standard No. 207 which addresses seating systems. Generally it is acknowledged that the current standard requires inadequate seat strength to insure that the seat does not fail when a car is subject to a severe rear impact.

Yet, it ignored or denied subsequent requests for a more effective regulation. In 1989, Saczalski petitioned NHTSA to amend FMVSS 207 to address the problem of inadequate seat strength and seatback failure, noting that “during rear impact the seat backs are loaded by the inertia of the occupants upper body and the current seat back requirements result in collapse of the seat back which allows the occupant to slide out from under the lap belt thus rendering the restraint system ineffective.”

Saczalski requested that NHTSA specify that the testing load be both 20 times the weight of the seat back and 20 times the weight of the occupant, and that the seat back torque criteria be increased to 56,000 inch-pounds. 

 A year later, occupant restraint expert Alan Cantor, of the consulting firm ARCCA, Inc., petitioned the agency to amend FMVSS 207 to prohibit occupant “ramping” up the seat back during seat deformation. These petitions prodded NHTSA to so some research. In 1992, it launched a Seat Back Strength project to gather information, acknowledging the current standard was inadequate to ensure that the seat does not fail when a car is subject to a severe rear impact. Despite periodic announcements that improved seat strength and head restraints was an agency priority, among its priorities, it has proposed no new rulemaking.

In August 2004, Cantor filed another rulemaking petition with attorney Larry E. Coben, this time for amendments to FMVSS No. 208 Occupant Protection requesting the inclusion of rear seat belted dummies in the dynamic crash testing requirements and the adoption of European ECE Regulation 17 which requires unrestrained cargo behind rear seats in frontal crashes.   

Coben and Cantor recommended that the agency select dummies of various sizes and adopting FMVSS No. 208 injury criteria for the head, neck, chest, and femurs. They also recommended adopting a new method of assessing abdominal injury risk. Coben and Cantor argued that applying the same injury criteria to rear seat as front seat dummies in frontal crashes was warranted, and would not cause any undue expense.

In early December 2006, the agency rejected the petition as “premature,” because it was actively engaged in a research program examining rear seat occupant protection, that included analytical and sled tests, and simulations with different-sized test dummies in the rear seats to determine advanced restraint system feasibility and improved restraint geometry.  

In 2015, after nearly a decade of NHTSA apathy, Cantor and four of his colleagues at ARCCA Inc., re-petitioned NHTSA to amend FMVSS 207:

Since 1989, hardly a day passes that we are not faced with an issue related to seat failure in a rear-end crash and the resultant serious injury that such a failure causes. As automotive seating and restraint experts here at ARCCA, we have been involved in hundreds of seat back failure litigation cases, the vast majority of which have settled prior to trial. As part of our work on these cases, we have had the opportunity to review seat strength data from various auto makers, and we have been involved in the conduct of a  variety of both static and dynamic tests on the failing seats as well as on seats that can withstand the types of forces typically seen in today’s passenger vehicles that are involved in rear-end crashes. In addition, we have seen both the published and non-published research and data from many others, including most vehicle manufacturers. What is clear from all of this is that automotive seats are more than just “chairs” to allow people to comfortably drive cars or for passengers to be transported in luxury: seats are also safety devices that provide restraint and, in a rear-end crash, the seatback should afford the same kind of protection to the user that a seat belt provides in a frontal impact. 

The petition pointed out that the current test – using a static load in an empty seat – ignored the laws of physics: occupants of different weights or mass will result in different loading of the seatback under the same rear-end crash conditions. ARCCA recommended that NHTSA establish a FMVSS 207 dynamic test – using the FMVSS 301 crash test as a guide – and a New Car Assessment Program (NCAP – NHTSA’s five-star rating system that manufacturers use to tout the safety of their vehicles and that consumers use in making buying decisions) rear-impact test. Or, more simply, restore paragraph S4.1 (b) of FMVSS 209, which required the lap belt portion of the seat belt to remain on the pelvis under all crash conditions.  

In 2016, The Center for Auto Safety petitioned the agency to modify its “child seating recommendations by adding the following or similar warning language and that such language be required in Owner’s Manuals under 49 CFR § 571.208 S4.5.1(f): If Possible, Children Should Be Placed In Rear Seating Positions Behind Unoccupied Front Seats. In Rear-End Crashes, the Backs of Occupied Front Seats Are Prone To Collapse Under the Weight of Their Occupants. If This Occurs, the Seat Backs and Their Occupants Can Strike Children in Rear Seats and Cause Severe or Fatal Injuries.”

These petitions are likely to meet the fate of their predecessors, dying quietly in obscurity.

Improving Rear Seat Safety through NCAP

The one action NHTSA has been willing to consider is to add a rear impact test to the NCAP, and let the universe of scarce stars scare the manufacturers into either improving their seat designs, or risking their markets. April 2013, the agency published a request for comments on possible changes to the New Car Assessment Program, and noted the need to improve rear seat safety.

In recent years, improvements that have been made to the front seat crash environment have significantly decreased the risk of injuries and fatalities for front seat occupants involved in frontal crashes. While exposure and injury rates for rear seat occupants overall are still relatively low, there is an emerging need to further understand the rear seat environment in recent model year vehicles, particularly in consideration of lighter and more compact vehicle designs. Comments are requested on the availability of any data that illustrate whether safety benefits can be realized through encouraging additional safety improvements and/or technologies including rear seat belt reminders targeted at protecting the rear seat environment.

The agency held out the possibility of dynamically testing rear seats and seat belts in the NCAP frontal crash tests, with a 5th percentile adult female Hybrid III dummy in the rear seat behind the ATD in the driver’s seat. The notice attracted nearly 60 commenters, many of whom generally supported the idea of enhancing safety for rear seat occupants and for additional NCAP tests with the 5th percentile ADT in the rear, but disagreed over the test parameters.

The agency, however, did not move forward with rear seat dynamic tests. Instead, it announced in January 2015, it planned to include NCAP ratings pertaining to automatic emergency braking systems.

In May 2016, the trade press, from Automotive News to the Insurance Institute for Highway Safety’s Status Report, reported that NHTSA planned to add a new oblique frontal crash test, simulating two mid-sized vehicles crashing at 56 mph, with a 50-percent overlap to measure how well vehicles protect people in an angled frontal crash. The test, using a stationary vehicle and a moving barrier, would feature a THOR 50th percentile male dummy in the driver seat and the modified Hybrid III 5th percentile female dummy in the right rear. The agency planned to implement this new test in time for the 2019 model year. And the impetus for this change was not to protect children, who have heretofore been frequent rear seat passengers, and vulnerable to injuries and deaths, but to protect adult passengers using ride-hailing services, such as Uber and Lyft.

This apparently did not happen, because in 2018, the agency published a notice of yet another  public meeting to gather input from stakeholders to discuss New Car Assessment Program. The October 2018 meeting, in part, “focused on crashworthiness strategies” for NCAP, including “rear seat occupant protection.” Last month, The Safety Record Blog, with no real confidence that the agency would be able to manage a “Yes” or “No” answer, asked if NHTSA ever made good on its 2016 promise. After nine days of e-mail exchanges, a NHTSA spokesman sent us a link to a press release announcing that NHTSA would publish another notice sometime in 2020, unveiling its proposed changes to the NCAP. Would these changes include a rear seat dummy in a frontal oblique crash test? Final word from the public information specialist: “The agency will determine the upgrades for the NCAP test after its careful review of the comments and feedback submitted to the 2020 Federal Register notice.” Well, that clears everything up.  

Despite this sad, long history of agency inaction, Saczalski is happy to see the agency publish a paper acknowledging the problem of weak seat backs and using Finite Element Analysis, a valuable engineering design tool, to build a better seat.

“This the paper is very good with FEA, you can find the most appropriate design, go to prototype part and run the tests. Then you know how to tweak the system. But this process has been so slow and the children who have died over all these years because of it – it’s a significant number.”