Automotive Seat Design and Collision Performance

760810 Automotive Seat Design and Collision Performance

D. M. Severy, D. M. Blaisdell, and J. F. Kerkhoff Severy, Inc.

ALTHOUGH SAFETY IS A BASIC CONSIDERATION research up to that time. Although some research in all aspects of automotive engineering, it is a has been conducted since that time, the development fact of life that safety advances are not uniformly of a safety seat still has not been accomplished. In accomplished. Some essential aspects of motorist the same study special devices, including the air protection have in the past, and probably will in the bag, were rated as only of moderate relative im- future, continue to be greatly outstripped by advances portance. made in other areas. In some instances, the design After 10 years of government regulated safety challenge simply exceeds current technology; on standards and nearly that many years of intensive other occasions an adequate safety standard and con- air bag development sponsored by both government certed industry interest can correct obvious safety and industry, the consensus of many automotive deficiencies. For example, although improved safety researchers in Europe (2), (3) and elsewhere tires resulted only after new materials evolved, a is to change the emphasis back to development of greatly improved motorist restraining system active restraint systems and to increase effective- resulted from industry initiated research that cul- ness by means of mandatory use laws. Air bags, minated in a standard requiring installation of com- if used, would serve a supplementary function in bination cross chest lap belts for all outboard front conjunction with active systems. Progress in seat locations. safer seat design has been impeded throughout a During the past 20 years, new materials and decade of automotive safety standard making, and techniques have improved the comfort and wear many lives have been lost that could have been fully resistance of automotive seats while simultaneously protected with a fraction of the inventive genius and reducing their weight and cost; however, significant funding lost to air bag development. A basic common safety related improvement in seat design during sense approach to motorist protection from collision this period has not been accomplished. A summary trauma calls for special attention to design of the of findings from prior research (l)* published in critically important structure nearest the motorist, 1966 classified a structurally redesigned Integral his seat. Seat with built-in S-point belt restraint as critically A vital consideration of seat design is its cap- important for reduction of motorist injuries; this acity to protect motorists, to the extent practical, summary also pointed out the lack of safety-seat from all types of collision injury exposures. This paper provides basic design data for crashworthy *Number in parentheses designate References automotive seat systems with integral active re- at end of paper.

Eighty-five laboratory full-scale force-deflec- tive significance to the requirements of automotive tion tests were conducted on passenger vehicle seat collision performance. seats, foreign and domestic, for purposes of evaluat- The foregoing research provided foundation for ing specific resistance to a collision environment modification of a production automobile seat into an and mechanisms of collision induced seat distortion. integral safety seat, based on a design concept that These tests evaluated seats that span the past thirty minimizes bending moments during collision. The years; additionally, seat design studies were con- modified seat was subjected to the same laboratory ducted evaluating basic features of automotive seat- test procedure applied to the 85 non-modified pro- ing during the past eighty years. duction seats and results of its performance is Data from full-scale collision experiments and given. from a large number of actual accidents facilitated Design concepts are presented that would serve the establishment of seat design criteria for greatly to mitigate undesirable seat distortions during col- improved collision performance. Evolution of seat lision and thus improve seat restraint capabilities and head support standards in the United States and without compromising the important factors of com- Europe are presented with evaluation of their rela- fort and cost.

2551 2552 D. M. SEVERY

straints, as determined from the authors’ collision a seat also should have adequate structure for research and laboratory studies as well as experi- housing safety and convenience accessories. Trade- ence gained from investigation of relevant accidents. offs are imposed by this complex mix of requirements; however, in seat design we should no longer BACKGROUND overlook the requirement for a reasonably safe, collision resistive structure with built-in active The first automotive seat, like the first restraint system. automobile was an adaptation from the horse-drawn SEATING CATEGORIES - Automotive seats carriage. Springs to absorb road shocks were for passenger vehicles have a number of basic primitive, effective padding was non-existent and configurations and sizes depending on intended use seat adjustability had not yet been considered. Com- and location. Position adjustment depends on mencing around 1900, motorist safety while travel- factors such as location (front or rear): availability ling over rough roads was improved by development of comfort features, including center armrests, of deeply contoured seats that reduced the likeli- depends on vehicle price, type, purpose and country hood of motorist ejection as the car body pitched of origin. and rolled. Generally, however, little considera- BENCH SEAT - SOLID BACK - Standard and tion was given for the safety of the occupants, compact 4-doer sedans are frequently equipped with Fig. 1. solid-back berth seats, front and rear. The front Front seat fore-and-aft adjustment was not seat has a single adjustment mechanism and the available until about 1929 when adjustable front entire seat is usually adjusted according to the seats for the driver became a feature of higher priced automobiles. Occupant comfort was given increased attention as engineering problems con- cerning motor vehicle performance and reliability became more effectively managed. Improvements in seat design continued and by the mid 1930’s, seats, tracks and runners closely resembled those of the mid 1960’s. During the period between the thirties and the sixties, the only significant in- novation in seat design was the introduction of power seats and adjustable reclining backrests during the early 1950’s, Table 1. Seatback height reached reasonable levels in the late 1960’s but by the mid- seventies the height of backrests on many models had declined to levels less effective than thirty v” years ago. Seatback strength has not increased BJtLAWt significantly over the past thirty years and remains Fig. 1 - Driver’s seat, horseless carriage inadequate to resist even moderate collision forces. The remarkable safety improvement facilitated by - the backrest head support standard FMVSS 202 has TABLE 1 been compromised through failure to upgrade criteria to maintain performance commensurate with SEAT DESIGN EVOLUTION intent. Seventy percent of adjustable head restraints are used in the downmost position (4). Little ef- INTRODUCED ITEM EXAMPLE fective protection is afforded the motorist unless 1890 - 1900 Automotive Bench Seats Philion* the head restraint is positioned behind or slightly 1900 - 1910 Deep Bucket Seats Thomas* 1910 - 1915 Fold-forward Backrests [Model-T Ford* above the head and remains in such support posi- 1910 - 1915 Console Between Seats Wescott* tion during collision. 1910 - 1915 Pedestal Seat Argo Electric* 1915 - 1920 Swivel Seat Cole* SEAT FUNCTION AND CONFIGURATIONS 1920 - 1925 Fold-down Armrest Dusenberg* 1925 - 1930 Fore-and-aft Adjustment Viking* 1950 - 1952 Power Seats Packard It is not a simple engineering task to design a 1960 - 1963 Optional Head Restraints All U.S. good automotive seat; it must provide comfort, 1968 Integrated Head Restraint Volkswagen style and safety, and yet be sufficiently light weight 1969 Standard Head Restraints All U.S. to facilitate vehicle fuel economy and to minimize collision inertial stresses. Seat designs and mater- _ ____ -- .------ials must be affordable and durable to give accept- *Based on authors’ study of vehicles made available through coopera- able service over the life of the car. In addition to tion of Harrah’s Automotive Museum, Reno, Nevada. provisions for comfort and position adjustments, AUTOMOTIVE SEAT DESIGN AND COLLISION PERFORMANCE 2553 needs of the driver. A bench seat, if adequately their elevation, direct restraint system attachment designed for three occupants, must be capable of requires special design considerations. sustaining a rearward static moment, calculated FIXED, BUS-TYPE BENCH SEATS - Although about the H-point of the seat, equal to three times similar to the rear seats of passenger sedans these that of a single seat. seats are elevated and generally provide open BENCH SEAT - SPLIT BACK - Fold-forward space beneath; structural limitations and remoteness backrests provide access to the rear seat of 2-doer of seat from floor increase the difficulty of direct vehicles. Since only the upper portion of the seat restraint attachment. Seats used in buses that is divided, the fore-and-aft adjustment can accom- transport children require special considerations, modate only one person. Backrest angle, however, (91, (10). can be made adjustable according to the individual needs of driver and passenger. SEAT SYSTEM COMPONENTS BUCKET SEATS - Contoured seats for individual occupancy are called bucket seats, although A brief description of seat components will this term originally applied to more deeply con- allow a better understanding of basic seat require- toured racing-type seats. The containment feature ments and the special seat structure necessary for of deeply contoured bucket seats has been identified adequate passenger protection from collision trauma, by the authors and others (5), (6), (7) as a consi- Fig. 2. deration in design of a safer seat, owing to improved STRUCTURAL FRAME MEMBERS - The seat lateral retention; however, the extent of contouring framework is usually constructed of steel that has must be based on consideration of comfort and been formed into tubular configurations or of stamped practicality as well as safety. or rolled sheet metal. Historically, the function Adustable backrest angle is commonly available of this structural backbone has been limited to pro- and provides improved comfort and reduced muscle viding shape for the cushioning members and support fatigue. Direct attachment to the seat of both lap for its own weight and that of its occupant. Rede- and shoulder belts is a safety and convenience goal sign and strengthening of the seat framework in that has been identified previously, (l), (5)) (6), conjunction with its anchorages can provide the (7), (8) and is further evaluated in this study. force resistance necessary for occupant restraint FOLDING SEATS - These seats require special during moderate and severe front, side and rear- design considerations to allow them to be folded end collisions. flat when hauling cargo. These requirements include NON-STRUCTURAL SEAT MATERIAL - Cushions, special sizing, latching devices and flat, durable springs and upholstery provide the necessary means cargo surfaces. Seatback height and strength as of load distribution between occupant and seat frame; well as provision for isolating cargo are collision they also provide contour and geometry necessary safety factors to be considered. for occupant comfort. A redesigned integral seat would include energy absorbing (E-A) padding at PEDESTAL BUCKET SEATS - These seats are the sides, top and back to provide additional force common to buses, vans, motor homes and trucks. moderation and load distribution during collision. In addition to the normal fore-and-aft position ad- This concept is typified by the Crandell-Liberty justments, variable backrest angle and swivelling Mutual Survival Car II (11). mechanisms are sometimes provided. Because of

D RESTRAINT /INTEGRAL OR ADJU.STAlsLE/

UPHOLSTERY STRUCTURAL FRAMEWORK

LUMBAR CONTOUR ADJUSTMENT ACKREST ANGLE ADJUSThfENT SEAT/BACKREST JUNCTION w c PIVOT ARM Fig. 2 - Seat nomenclature PIVOT FOR REAR SEAT ACCESS LFORE AND AFT AND V.ERT/CAL ADJUSTMENT 2554 D. M. SEVERY

SEAT ADJUSTMENT MECHANISMS - These and attachments require a design of adequate strength mechanisms should be strengthened to withstand to accommodate seat and occupant inertial forces collision forces as well as the rigors of everyday identified specificaIly in other sections of the paper. usage. Considerations of driver anthropometric Besides the usual floor, sill and tunnel anchorages, differences necessitate a multiplicity of seat other potential anchorage locations for strengthened adjustments to assure that a given seat provides seat systems include side and roof attachment. the multiple functions of driver positioning for correct reach and view, as well as for adequate SEAT DESIGN CONCEPTS comfort. Desired seat position may be accomplished by means of longitudinal, vertical and tilt adjust- Restraint compliance, is the change in re- ments . Other adjustments include backrest angle, straint geometry induced by motorist’s inertial head restraint position and lumbar support stiffness. forces during collision. For example, a lap belt Head restraint height adjustment is a common may initially course from its floor anchorage to- feature of current production seats. Unfortunately, ward the lap of the wearer at 45 degrees but as a many recent seat designs provide no head support frontal impact develops, this angle is reduced owing protection from even minor rear-end collisions to a combination of factors that result in the when the head restraint is in its most commonly used, occupant’s buttocks being compressed downward and retracted position. Proposed legislation (12) would forward on his seat as the belt restraining forces modify head restraint requirements to ensure pro- increase, (13). Similar ily , variations in support per utilization by preventing mis-adjustment or force, restraint angle and resulting backrest complete removal. Backrests with integral head pressure modulations occur for the seated occupant restraints eliminate adjustability deficiencies, but as the seat backrest undergoes restraint compliance should be designed to protect 95th percentile adult during a rear-end collision. This is illustrated for male occupants. Adjustable headrests should extend restrained and also for inadequately restrained to 80 cm (32-in) and not allow adjustment below occupants, by the geometric areas within the curves 70 cm (28 in); the headrest in its adjusted elevated depicting displacements of occupants relative to position must not collapse from head impact. their vehicle, Fig. 3. Also shown is the added SEAT ANCHORAGES - Most vehicle designs physiological trauma associated with unstructured, depend on seat attachments at the floor or at the inefficient restraint compliance. Variations in occu- sill and tunnel to transmit forces between the pant collision-induced accelerations (slopes, Fig. 3) vehicle and the seat. This anchorage is generally depend on the effectiveness of occupant restraints interposed between the seat adjustment mechanism which determines in large measure the trauma and the vehicle floor pan or lower structure. during ride-up/ride-down accelerations; the mech- Under static conditions, the seat anchorage anism of transmission of accelerative forces re- transmits compression, tension and shearforces present the other operative factor for such trauma- from the seat to the floor or side structure. In tic exposures. The problem faced by a seat designer the event of collision or handling accelerations, the is to determine the force resistive levels, and there- resulting forces are transmitted in reverse direction fore the degree of collision severity, a specific seat from floor to seat. The seat anchorage structures design should be able to accommodate before back-

TIME, MILLISECONDS Fig. 3 - Relative abruptness of occupant “Ride-down/Ride-up ,I’ according to efficiency of restraint AUTOMOTIVE SEAT DESIGN AND COLLISION PERFORMANCE 2555

rest restraint compliance reaches values where burdens remain yet unsolved. With this in mind, occupant forced displacements expose him to serious the authors redesigned a Volvo production bucket or critical injuries. seat into an integrated safety seat. This approach When not faced with the realities of cost and should facilitate recognition by automobile manu- operational practicality, as well as the need to pro- facturers and the Federal Government’s safety vide design features that favor public acceptance, standards making authority, that the goal for greatly a simplistic approach is to set safety design criteria improved motorist safety is reasonably attainable at some arbitrarily high level, one considered suf- through development of an integrated front seat. ficient to encompass virtually all probable collision The Volvo seat was stripped of its upholstery exposures. Such arbitrary criteria is insensitive and subjected to an evaluation of modifications con- to orders of priority and cost for safety design im- sidered essential to developing performance speci- provement and disregards the interactive response fications for an integrated safety seat, Fig. 6. it may impose on other related, or concurrently Next, this production seat frame was modified to important, passenger and operational safety features. provide the added structural support required for The front seat backrest and the steering column the special protective hardware, Fig. 6b. have in common the engineering design problems Seat Modifications - Changes were made in the that result from dealing with cantilevered structures production seat by additions of components to aug- acted upon by occupant collision forces. Additional- ment existing safety and comfort features and to ly, backrest angles of 15 to 25 degrees increase introduce new protective functions as follows: design difficulties by incorporating angles that allow less efficient crash force moderation. (a) High section-modulus tubular steel backrest, from seat attachment through head support. LABORATORY SEAT TESTS (b) Double seat tracks, for more positive anchorages. Passenger vehicle seats of differing make (c) Lap and cross-chest safety belt inertial vehicles are as varied structurally as the motor reels mounted to seat frame for ease of fastening vehicles within which they are mounted. Aside from and for optimum occupant restraint for all adjust- their apparent commonalities consisting of seat, ment positions. backrest, head support, covering, padding, springs, (d) Energy absorbing back panel, to moderate and frame, the multiplicity of design and size and pocket occupant inertial loads during medium variations account for their significantly differing and high-speed rear-end collisions and to provide weights (under 9 kg to over 45 kg or about 20 to knee pocketing for kinematic control of rear seat over 100 lbs.) height of backrests (from 51 cm to occupants thrown forward during frontal collisions. about 76 cm or from 20 to about 30 in) and maximum (e) A pair of seat-mounted inertial reels for bending moment before excessive backrest deflect- belt fabric which connects the base of the seat at ion (from 45 m-kg to about 95 m-kg or about 4000 its rear to the roof; the belt fabric is passed through in-lbs to 17,000 in-lbs), Fig. 4. As remarkable the backrest to the roof anchorage. This arrange- as these variations would appear, they represent ment allows the seat to be adjusted fore-and-aft little difference in motorist protective capacity and folded for entry accommodation; however during owing to the fact that none of them are capable of a front or rear-end collision, the dual “safety effectively resisting motorist inertial forces for belts” are capable of immediate seat inertial snub- any substantial impact exposure. bing. Based on laboratory tests of seats from cars (f) Deeply contoured bucket seat geometry to large and small, foreign and domestic, and from assure occupant positive lateral support and to vehicles 30 years old to near new, backrest strengths improve comfort. were found to be remarkably alike, Fig, 4. Force- (g) Retention of the most advanced comfort deflection performance curves of representative pro- and safety features of backrest angle adjustment, duction seats as well as recommended performance lumbar spine firmness adjustment and louvered see- for passenger vehicle seats capable of protecting through head restraint. their occupants during severe collisions are shown, Most prior integrated seat concepts have Fig. 5. utilized the backrest cantilever principle which INTEGRATED SEAT - The concept of an in- necessitates strengthening floor and seat anchorages tegrated seat for motorists has been considered along with seat and backrest framing in an effort and evaluated theoretically; experimental seats to combat the extreme bending moments, especially have been constructed and tested experimentally, attending front and rear -end collisions, (7)) (8)) (I), (2), (5), (6), (7), (II), (14). In general, the (14). The advantage of this approach is that seat approach has been to design an entirely new seat, fore-and-aft positioning and backrest folding for inducing the argument that practical aspects of access to the rear seat can be accommodated with- manufacturing, occupant comfort and similar design out roof or side attachment; disadvantage is that 2556 D. M. SEVERY

1940 1950 1960 1970 1960 YEAR

z 1200- (RECOMMENOEO STRENGTH) \\a\\\\ c.3 _1104.000, $2 IOOO- LT 1 ~87,000, k-i U-J = ,,&?~

Fig. 4 - Automotive seat strength and backrest height (PREVAILING STRENGTH) variations over the past thirty years

~.~\\\\...~\\~

1940 1950 1960 I970 1960 YEAR

* STATIC REARWARD FORCE APPLIED 35 CM ABOVE SEAT/BACKREST JUNCTION

Fig. 5 - Backrest force-deflection data for 85 production BACKREST DEFLECTION, CM (IN) automobiles, foreign and domestic, large and small sized vehicles

it is a very inefficient method of resisting large tests along the lines previously described as shown forces owing to the massive increases in structure by Fig. 8. The improvement in force-displacement required to offset extremes in bending moments. over non-modified production seats is shown by the The authors found that many production seats could dotted curve of Fig. 5. Considering the nature of resist between 900 and 1800 kg force (about 2000 to changes necessitated by using an existing seat, it 4000 lbs) horizontally rearward or forward at the is apparent that a somewhat lighter seat could be seat rail. If these floor anchorages were only made even stronger if it were an original design slightly improved, and connected by belt webbing rather than a retrofit. The composite design to the roof, the cantilever design could be replaced factors for a remarkably stronger, attractively by the basically greater efficiency of the tension contoured seat are shown by Fig. 7. Extensive semi-hoop design spanning floor to ceiling. The crash testing of such new concepts are recommended Experimental Safety Vehicle by Opel also uses a for sound development before they are used in pro- form of roof anchorage for its front seat backrests, duction vehicles. (2). Encouragement was given to the practicality of this concept by a preliminary test of roof strength SEAT COLLISION PERFORMANCE of a mid-sixties 2-doer hardtop sedan. This un- modified roof structure withstood over 1800 kg Relative to vehicle safety, performance connotes (4000 lbs) horizontally rearward at a localized back- a level of attainment, a plateau from which concepts rest to roof attachment location, without significant for improvement may be assessed, both as to prac- downward crush. ticality and possible adverse influence on other safety The integrated seat was subjected to laboratory requirements. Like the occupant that uses it, a AUTOMOTIVE SEAT DESIGN AND COLLISION PERFORMANCE 2557 2558 D. M. SEVERY

ROOF ATTACHMENT: ;;;

ROOF ANCHORED BELTS TRAVERSE DOWNWARD THROUGH SEAT INTEGRATED SEE- THRU HEAD RESTRAINT PADDED LATERAL SUPPORT BlJlL T- /N 3 PO/NT BELT / BACKREST PANEL DEFORMABLE MA ROOF ATTACHMENT INERTIAL EXTENDED LATERAL STRUC ROCKER S/L L Fig. 7 - The Integral Safety Seat conceptualized and tested by the authors

Fig. 8 - Integrated seat during testing; 2268 kg (5000 lb) force applied rearward seat has rather specific collision force tolerances: The basic collision performance criteria that exceeding its design criteria allows crash forces applies to a seat is its resistance to backrest yield to yield or fracture whatever component of the seat and its ability to transmit forces through the seat is most susceptible. This collision response is to its anchorages. Although seat design has in no common to all seats, as determined from destruc- way kept pace with advances made in other safety tive tests to over 85 different automobile seats related areas, there are some notable exceptions; mounted on their floor pans. Although individual certain Mercedes and Porsche automobiles, for designs determine which components may fail first, example, attach lap belts to the seat, rather than their level of susceptibility is remarkably similar. using the less satisfactory arrangement of floor Perhaps the foundation for this similarity in seat anchorages. Volvo and Saab have provided see- strength levels is generic to the basic design require- through head restraints that improve visibility. The ments that were dictated long before collision forces track record of most vehicles, however is not were identified as factors for consideration. For impressive. In general, backrest strength and example, an adult male can apply 90 to 135 kg (200 height has not improved, possibly because the need to 300 lbs) force to the backrest during emergency for stronger, full-support backrests is not fully braking. If passengers also brace for a possible understood by regulatory agencies and those making collision, the backrest of a bench seat can be called design decisions. upon to resist as much as 275 kg (600 lbs) force, SEAT COLLISION PERFORMANCE - FRONTAL It is understandable, therefore, that seats in the COLLISIONS - It is becoming increasingly apparent forties and fifties had backrest yield resistance of that passive restraint systems alone cannot protect about 275 kg (600 lbs) even though they were not motorists from the varied exposures encountered designed in a period when collision safety was a during motor vehicle accidents (2), (3). Systems prominant consideration: incomprehensible is the that protect during a barrier-type crash may be fact that seats in the seventies show no significant completely ineffective if the same frontal collision improvement. is accompanied by one or more minor impacts or AUTOMOTIVE SEAT DE,‘,IGN AND COLLISION PERFORMANCE 2559 an upset. The cross-chest lap belt restraint system, the lap safety belt. This arrangement maintains on the other hand, provides effective motorist pro- the lap belt across the pelvic girdle during collision tection during most collisions or combinations of and, for non-integral seats, reduces mal-alignment collisions, including upset. If an integral three- of the cross chest strap as the torso flails forward point restraint system is available, in conjunction against it. with the strengthened seat to which it should be SEAT COLLISION PERFORMANCE - REAR- attached, it can provide levels of protection unat- END COLLISIONS - The most frequently occurring tainable with any other single system, however type of motor vehicle collision is the rear-end im- costly or complex. In addition to minimizing or pact. For example, forty-nine percent of the 14.6 eliminating impact forces to the vehicle interior, million accidents involving motor vehicles in the the combined restraint system appreciably reduces United States during 1968 were same direction or the forces that would be applied for a lap belt only rear-end type accidents; representing nine percent protective system, (15). of the total number of fatal collisions, (17). One advantage claimed by air bag proponents is Seat backrest performance during rear-end the additional ride-down distance provided as the collisions parallels a condition of walking on thin motorist crushes forward relative to the vehicle ice; minor torso inertial forces receive adequate interior. Direct attachment of belt restraints to support but as impact severity increases, the the seat minimizes uncontrolled belt slack and at resistive support approaches negligible values. The the same time allows the designer to incorporate mechanism by which this passive restraint is lost belt pre-stressing and other ride-down improve- is generally unimportant insofar as its effect on ments for the belt and seat anchorages. The belt motorist kinematics is concerned. Torso inertial system can thereby provide maximum protection for movement does not respond differently with differing front-end collisions and also provide more adequate mechanism of loss of back support; it simply re- protection during the wide range of other collision sponds inertially to fall-off of resistive force. Thus, exposures, including moderate to severe rear-end whether backrest resistance is subsequently lost collisions by keeping the occupant in the seat. Even because of excessive backrest yield or by combina- a floor mounted lap belt can provide considerable tions of baclcrest yield and anchorage separations, protection under these conditions, (16). the motorist inertial forces will act in the same Rear seat occupants that fail to fasten their manner. belt would obtain some degree of passive restraint Backrests are slightly reclined for comfort from a strengthened front seat backrest. The knee and to assist with torso support. However, as the pocketing concept evaluated and described previously seat is forced to recline additionally during rear- (6), will add considerably to rear seat occupants’ end collision, the torso is predisposed to ramping collision protection by controlling posture during up the plane of the backrest, Fig. 9. On reaching their “second collision. ” The spacing between rear backrest angles exceeding 45 degrees, the ramp occupants and front backrest provides a time delay component falls off according to the function “sin CY before the rear occupants strike the front backrest; cos CX, I’ where alpha (a) is the backrest angle rear- therefore, front seat occupant ride-down would not ward from the vertical axis. Simultaneously, the be compromised significantly from subsequent backrest support force, FSB, which accelerates impact to the backrest by the rear seat occupant. the motorist as the backrest bears against him Rear seat occupant protection for all but rear-end during a rear-end collision, falls off as a function collision exposures can best be accomplished by of backrest yield-angle, likewise depicted by Fig. 9. means of active restraint systems. Obviously, a zero or near zero angle provides the OCCUPANT RESTRAINT WITH A RECLINED most effective torso support by the backrest during SEAT - The cross-chest lap belt is less effective rear-ending and torso support tends to fall off in front-end impacts where the passenger has thereafter, as a cosine function of increases in the reclined his seat; the torso slides forward, shifting baclu-est angle. the lap belt upward across the viscera. When belt These factors are particularly relevant for snubbing eventually occurs, forces are transmitted production seats involved in light to moderate to portions of the body unable to sustain them with- rear-end collisions where bachest strengths out exposure to serious injury. Also, the neck or generally are adequate for those respective exposures. head may strike the cross-chest belt in an ineffec- However, for most moderate to severe rear end tive, if not dangerous manner. One solution is to collisions, all production seat backrests deflect to provide an additional pair of straps attached at the angles providing no effective motorist support. front of the seat and passed upwards through posi- This situation is illustrated by considering the re- tioning guides. Before reclining the backrest, the lationship of backrest support provided the torso as hanging ends of the straps are pulled up between a function of rear-end collision speed, Fig. 10. the legs and fastened to attachment loops sewn to Sixty-five percent of the body weight may be used D. M. SEVERY

Fig. 9 - Degradation of backrest support as function of collision induced deflection /

I_---, cI, IO 20 30 40 50 60 ‘70 RO SEAT BACKiEST 4NGLE ,o< , DCWCES as a minimal approximation of the torso and requirements for individual components. For ex- adjacent flailing body components that operate any ample, the extremely difficult design task of instant to inertially act on the backrest and may be providing 75,000 inch-lbs yield resistance at the characterized as the occupant inertial force, F . bachest pivot for a 50% adult male in a 30G seat The seat backrest yield point, or point of maxizim system can be easily managed by utilization of a battiest resistive force, F , is reached when the roof ant hor . If one-third of the 6540 pound hori- occupant inertial force plus 7he seat backrest zontal force is applied at the top of t,he seat and inertia force exceeds FR as shown by F acts 32 inches above the backrest pivot, we now have 8’t;eFsI = Based on collision experiments (1 the following moment at the pivot arm: FR’ torso inertial force increases with collision speed, MI = 75,000 - (32) (2180) = 5300 in-lbs, a in the manner shown by Fig. 10. Regardless of reduction to 7% of the original required bending speed of impact, the backrest can provide only a moment. Most production seats can provide this given maximum resistance to yield; however, as level of bending resistance without need for modi- rear-ending speeds increase, the collision severity fication. With a roof attachment, the maximum is reached where the baclu-est can at best support bending moment will be at a higher elevation on the solely its own weight, but none of the occupant, seat and in a location of the backrest free of com- Fig. 10. Thus the effective backrest strength is plications of pivots and adjustment mechanisms. dependent on a combination of backrest resistance Correspondingly, the reduction in reaction strength to yield and backrest weight, referred to as back- required at R , RH and RF, if a roof anchorage is rest weight, referred to as backrest inertial yield added, prove*% es a requirement of only 2%, 65%, resistive force. The lighter the weight of a seat, and 2% respectively of the values necessary without relative to a given resistance to deflection, the a roof attachment. better able the seat to resist torso displacement SEAT COLLISION PERFORMANCE - SIDE IMPACTS - during rear-end collision. Direct side impacts into the passenger compartment Given a sufficiently severe collision, any seat represent the gravest danger to motorists at a given will fail; the type of failure however, depends on speed of impact for any type of collision, (18). design. Rear-end collisions induce five basic Usually less than a foot of intervening structure, types of failure; a particular collision exposure for much of which is non-structured space, separates a particular seat design may result in one or a the motorist on the impacted side from the front combination of these five types of failure: end of the impacting vehicle. During collision, (a) Excessive backrest yield the near-side occupant suffers a considerably more (b) Track-runner vertical separation severe acceleration than does the vehicle in which (c) Track-runner longitudinal separation he is riding. This impact exposure can be lessened (d) Compressive crush at seat base somewhat by increasing the amount of side structure (e) Floor yield at seat attachment stiffness and thereby minimizing passenger compart- All automotive seats have collision induced ment intrusion. However, the point of diminishing failure modes that are caused by moments and return is soon reached as intervening structure is forces exceeding one or more of the seat compo- added, since crush distance is still limited, and nent yield points. Critical reaction requirements delayed occupant response serves to magnify the can be calculated for various occupant sizes and resulting exposure, A considerably simpler acceleration exposures using free body analyses of solution is to attach the motorist firmly to a deeply the seat system, Fig. 11. contoured seat by means of an integral 3-point belt For rear-end collision exposures, the addition and to provide intermediate structural members of a roof anchorage for the backrest serves to between striking car and seat to cause the seat, distribute forces and considerably lessens design AUTOMOTIVE SEAT DE:XGN AND COLLISION PERFORMANCE 2561

IL 400-

BACKREST PEAK RESISTAANCE TO YIELD

0 20 100 (124) u44093 ($33) 14% 162 I) REAR-END COLLISION CLOSURE VELOCITY, KWHR (MPH) Fig. 10 - The relationship of backrest torso support, as a function of rear-end collision speed

3 -1 8,000- $ (69,000) -

iz 2

0 IO 20 30 40 COLLISION ACCELERATION, G’s

Fig. 11 - Collision induced seat stresses, rear-end collision exposure with occupant attached thereto, to be accelerated up, ” is similar to the “ride-down” concept of more promptly but less abruptly. protection from frontal collisions, identified by Analysis of the velocity-time profile of a prior publication, (19). The more prompt typical side impact collision, Fig. 12, provides acceleration also reduces the impact abruptness or further understanding of the nature of accelerative “Stapp’‘-factor. Since the combined weight of the exposure and resulting injury mechanism as it seat and the occupant is often less than 90 kg exists for the unrestrained struck motorist. This (200 lbs), it is not necessary to provide significantly graph also shows the remarkable reduction in heavier structure than already present to accomplish occupant acceleration that can be accomplished if the objective of prompt seat acceleration during a the seat is structured and interlocked with side collision. It is only necessary to position the panels so that it is accelerated more promptly structure at the proper elevation and location to following initial contact by the striking vehicle. allow more direct collision force application to This improved side impact acceleration, or “ride- the seat structure to occur before substantial in- 2562 D. M. SEVERY

50 K#/l

kl; G ‘MAXIMUM ACCELERATIONS (APPROX.) 8 0 STRIKING VEHICLE C.G. ,...... _...... _... 20 G \ t IO 0 STRUCK V E HICLE C.G ._...... 20 G -(33) OUTER SURFACE _...... _... + @ G ? INNER DOOR SURFACE 1706 ;,.,li /‘m f : : ;,;z;;;;;pzg(,GENERGY ABSORPTION 60 G @ 4, u ,, 3tfl’ FULLY RESTRAINED I-X-CliPANT““““. _.. , IN. . . INTFGRAI,.. . L.,I.__ SEAT... __. __... ._. ._. __ .__. 30 G

Fig. 12 - Integral seat, side-impact exposure analysis ward crush has taken place. Additional motorist STANDARDS AND RECOMMENDED PRACTICES force moderation is provided by the pocketing effect of the integrated seat. This seat side-struc- In November 1963, a Society of Automotive ture tends to shield the motorist from direct Engineers Committee report on passenger seats impact by his vehicle side structure as well as to and adjusters was approved. This action, for the minimize “slack” between the motorist and his first time, extended the concept of recommended restraint system. This shielding also reduces practices and performance standards to include motorist exposure to punctures and shear forces. automotive seats. Two tests were specified by ROLLOVER PROTECTION - Injuries and recommended practice 5879, the first of which fatalities during rollover accidents occur most established minimum strength requirements for frequently as a result of ejection (20), (21); a horizontal inertial seat loadings for front and rear basic requirement for occupant restraint systems impacts. This 20G specification corresponds to is to retain the occupant within the vehicle during approximately a 900 kg (about 2000 lbs) seat an- upset. Even if adequate retention is achieved, chorage loading for a typically large domestic protection is compromised if there is inadequate bench seat and serves to evaluate the retention supportive roof structure. capability of the adjuster mechanism and seat Roof structures are sometimes flattened during anchorages. The second test prescribed by SAF upset sufficient to dangerously reduce the motorist 5879 concerned backrest strength and specified sur viva1 space. This crush can be reduced consider- a minimum moment of 4250 in-lbs. Although the ably by constructing the seat backrests with suf- scope of the recommended practice included an ficient strength to act as central pillars for the intent to submit it to a “continuing review” no roof. A strengthened seat with integrated head substantial changes were made until July 1968. restraint of a height necessary for adequate head In 1965 the Federal Register published General support for taller persons 80 cm or more (32 in) Services Administration (GSA) Federal Standard is recommended if central roof collapse resistance 515-6 governing “Anchorage of Seats. ” This initial is to be increased by backrest structural support. seat standard became effective in 1966 for 1967 mo- Seat backrest structure of sufficient strength and del automobiles. GSA adopted basically the original height to answer the requirements imposed by SAE recommendations with no major changes. other collision exposures will also prevent complete Although this transfer of 5879 from an SAE Recom- collapse of roof structure in nearly all collision mended Practice to a GSA standard was applicable exposures. only to government purchased vehicles, it was understood to be the forerunner for a standard directly affecting every passenger car sold in the United States. AUTOMOTIVE SEAT DE.;IGN AND COLLISION PERFORMANCE 2563

In 1966, by act of Congress, the Department of tion (NHTSA). However, with respect to automo- Transportation (DOT) was added to the Federal bile seat collision safety, little progress has been Government. The 1965 GSA 515-6 Standard, which made. A near obsession with air bag development was actually the SAE Recommended Practice of 1963, regrettably has diverted a tremendous amount of was adopted by DOT in January 1968 as Federal research time and resources from the readily Motor Vehicle Safety Standard (FMVSS) 207. The achieveable goal of making the automobile seat FMVSS 207 established backrest strength for crashworthy. As a result, many automobiles standard passenger vehicles at 3300 in-lb, differing being manufactured a decade after the onset of insignificantly from the GSA/SAE 4250 in-lb value, Federal Government standard making still have owing to a change in reference systems. FMVSS seats no stronger than they were during the pre- 207 references the Hip Pivot or H-point (as described regulatory period. Any real improvements made by SAE J826b) whereas the SAE 58’79 and GSA 515-6 in seat design have been on the initiative of the used a reference pivot point that was about 20 cm manufacturers because the seatback standards lower (8 in) at the interface between the seat frame relating to strength have remained virtually un- and adjuster. Although 3300 in-lbs may appear to changed. The presence of an inadequate, seldom be an impressive value, production seats from the upgraded standard limits initiative of automotive nineteen forties and fifties tested by the authors were manufacturers because of the implication that found to substantially exceed this standard. FMVSS satisfactory conditions prevail. 207 included the requirement that the fold-forward backrest locks withstand a 20G minimum inertial SUMMARY OF FINDINGS load which represented a modest improvement. In 1972, FMVSS 207 was amended to include multipur- 1. Improved automotive seat systems should pro- pose passenger vehicles, trucks, and buses. Addi- vide adequate structure and built-in active restraints tionally, the standard was expanded to include re- to insure a reasonable degree of protection from quirements relating to rear seats. collision forces. Seats need adequate contour to Head restraints, whether separable, or integral provide effective support and sufficient adjustments with the front seat backrest were regulated by to meet the varied and changing requirements of FMVSS 202 in 1969. This standard requires the their occupants. Redesign and strengthening of the head restraint to be capable of resisting a 200 pound seat framework and anchorages is required for force applied horizontally to it and actually consists direct attachment of lap and shoulder belts and to of a more severe loading to the backrest pivot than provide the force resistance necessary for occupant does FMVSS 207, to reduce the chances of head restraint during moderate and severe front, side support failure before backrest failure. and rear-end collisions. In Europe, seat strength standards applicable 2. All automotive seats develop collision-induced to non-exported vehicles were enacted by the Council yield or separation modes caused by forces and of European Communities in 1974; these are moments from motorist and seat inertia when such basically the same as FMVSS 207, except for an stresses exceed one or more of the seat component increase in the minimum moment resistance of yield points. Although the specific component of a the backrest from 3300 in-lb to about 4700 in-lb seat that fails first may differ, the degree of sus- (54 m kgf), both with respect to the same H-point ceptibility of all seats is remarkably similar. Lab- reference. oratory tests establish that production seats from Prior to the advent of the automotive standards cars large and small, foreign and domestic, and making authority by the Federal Government, the from vehicles 30 years old to near new, have back- automotive industry was largely self-governed, rest strengths remarkably alike. All are incapable relative to implementation of safety concepts. of effectively resisting motorist inertial forces for Slow, methodical but nevertheless significant any but light impact exposures without inducing progress was made during this self-governed period, excessive yield and/or component separation. as pointed out in a prior publication, (19). The 3. Variations in support force and restraint angle emphasis was on operational and performance occur for the seated occupant as the seat backrest safety; accordingly, collision safety advances were undergoes restraint compliance during a rear-end limited. collision. The mechanism by which this passive Following the mid-sixties, the responsibility restraint is lost is generally unimportant insofar for progress in collision safety was taken over by as its effect on motorist kinematics is concerned. the Federal Government and assessment of the Torso inertial movement does not respond dif- progress made in this more recent period depends ferently with differing mechanism of loss of back on the specific safety category being considered. support; it simply responds inertially to the degrada- In general, commendable improvement in collision tion in resistive force. safety has occurred during this period of regulation 4. Intended headrest height requirements are by the National Highway Traffic Safety Administra- not met for many seats when the headrest is in its 2564 D. M. SEVERY most commonly used, retracted position. Backrests 9. The integral safety seat protective capability with integral head restraint should protect 95th per- against frontal impacts is well established, and centile adult occupants; it is preferred that head- roof crush during rollover can be considerably rests be contiguous with the backrest and of fixed reduced by taller and strengthened front seat back- elevation; if adjustable, headrest adjustment should rests serving as central pillars for augmentation of provide extension to 80 cm (32 in) and not allow roof support: with an integral seat-restraint system, adjustment below 70 cm (28 in). regardless of the direction of impact, levels of 5. Passive restraint systems that effectively con- motorist protection can be achieved that are un- trol motorist collision dynamics during a barrier attainable with any other single system, however crash may be ineffective if the same frontal collision costly or complex. is complicated by multiple impacts or upset. The 10. Commendable improvement in collision safety cross-chest lap belt restraint system, on the other has occurred during a decade of regulation by the hand, provides effective motorist protection during National Highway Traffic Safety Administration. most collisions or combinations of collisions, However, with respect to automobile seat collision including upset. safety, effective progress has been negligible. A 6. Collision-induced occupant accelerations near obsession with air bag development regrettably vary according to the effectiveness of occupant has diverted a tremendous amount of research time restraints. These variations in “ride-up” and and resources from the readily achievable goal of “ride-down” velocity changes determine in large making the automobile seat crashworthy. Although measure the extent of trauma during collision; the the 3300 in-lbs backrest resistance feature of the mechanism of transmission of accelerative forces FMVSS 207 sounds impressive, production seats represent the other operative factor for such from the nineteen forties and fifties tested by the traumatic exposures. authors were found to substantially exceed this 207 7. The authors designed, constructed and labora- principal criteria. tory tested a high performance integral safety seat 11. Judged on any rational basis, whether as to that was made by structurally modifying an existing initial cost, maintenance costs, functional reliability, production seat design without compromising its exposure to dangerous malfunction or lack of advanced design comfort features. The roof-to- effectiveness for many types of collision exposures, floor anchored webbing for backrest crash force the air bag comes up as a king-size loser when com- attenuation remarkably simplifies seat structural pared with the collision protection of the integrated requirements because bending moments are de- seat which includes integral cross-chest lap belt creased more than ten-fold. By use of inertial restraint. While the United States endlessly ponders reels this front seat conversion is accomplished the air bag issue, the rest of the automotive world without inhibiting rear seat access for two-door has reached the conclusion that cross-chest lap vehicles or backrest reclining or seat fore-and- belts in combination with mandatory use laws aft positioning. Direct attachment of cross chest, represents the simplest, least costly and most lap belt restraints to the seat minimizes uncontrol- effective motorist protection measure. led belt slack and allows incorporation of belt pre- stressing and other restraint compliance improve- REFERENCES ments. Increased costs for this restraint system concept would be nominal considering the benefits, 1. D. M. Severy, “Collision Performance of however extensive crash test evaluation of any the Motorist’s Compartment, ” Highway Safety Re- safety seat design is required prior to production search Institute, University of Michigan, Ann Arbor, to make certain all safety design features act con- Michigan, April 1967. structively without degrading other aspect of 2. Anon. , “Fifth International Technical Con- motorist’s safety. ference on Experimental Safety Vehicles, ” sponsored 8. The integral safety seat provides the basis for by U. S. Department of Transportation, Washing- effective protective measures against direct side- ton, D. C., hosted by the U. K. Department of the impact collisions. With the motorist comfortably Environment, London, England, 1974. restrained by a 3-point belt and seated within a 3. K. Langwieder, “Car Crash Collision Types deeply contoured seat, structural members between and Passenger Injuries in Dependency Upon Car the seat and outside door surface act, during Construction, ” Sixteenth Stapp Conference, SAE collision, to accelerate the seat, with occupant No. 720968, Society of Automotive Engineers, Inc. , attached, during the initial contact and crush phase Warrendale, Pennsylvania, 15096, 1972. of the collision and therefore accelerate the motor- 4. J. D. States, et al, “Injury Frequency and ist less abruptly. Additionally, this design mini- Head Restraint Effectivenss in Rear-end Impact mizes “slack” between the motorist and his restraint Accidents, ” Sixteenth Stapp Conference, SAE No. system and shields him from direct puncture and 720967, Society of Automotive Engineers, Inc. , shear forces. Warrendale, Pennsylvania, 15096, 1972. AUTOMOTIVE SEAT DE! IGN AND COLLISION PERFORMANCE 2565

5. D. M. Severy, H. M. Brink and J. D. Baird, motive Engineers, Inc. , Warrendale, Pennsylvania “Collision Performance L. M. Safety Car,” SAE 15096, 1959. No. 670458, presented at SAE Mid-year Meeting, 14. D. M. Severy, H. M. Brink and J. D. Chicago, Illinois, May 1967. Baird, “Backrest and Head Restraint Design for 6. D. M. Severy, H. M. Brink, J. D. Baird Rear-end Collision Protection, ” SAE No. 680079, and D. M. Blaisdell, “Safer Seat Design,” SAE No. presented SAE Congress, Detroit, Michigan, 690812, Proceedings of Thirteenth Stapp Car Crash January 1968. Conference, Harvard School of Public Health, 15. L. M. Patrick, et al, “Impact Dynamics December 1969. of Vehicle Occupants, ‘I Tenth Stapp Conference, 7. F. W. Babbs and B. C. Hilton, “The Pack- Society of Automotive Engineers, Inc. , Warrendale, aging of Car Occupants - A British Approach to Pennsylvania 15096, 1966. Seat Design, ” Chapter 32, Proceedings of Seventh 16. F. Preston, “A Comparison of Contacts Stapp Car Crash Conference, University of California, for Unrestrained and Lap Belted Occupants in Los Angeles, California, October 1963. Automobile Accidents, ” Seventeenth Conference 8. D. M. Severy, H. M. Brink, J. D. Baird American Association for Automotive Medicine, and D. M. Blaisdell, “Active Versus Passive 801 Green Bay Road, Lake Bluff, IHinois 60044. Motorist Restraints, ” SAE No. 700424, International 17. J. W. Melvin, J. H. McElhaney, V. L. Automobile Safety Conference Compendium, Detroit, Roberts and H. D. Portnoy, “Deployable Head Michigan, May 1970; Brussels, June 1970. Restraints - A Feasibility Study,” SAE No. 710853, 9. D. M. Severy, H. M. Brink and J. D. Baird, Fifteenth Stapp Car Crash Conference, Society of “School Bus Passenger Protection, SAE No. 670040, Automotive Engineers, Inc. , Warrendale, Pennsyl- Automotive Engineering Congress, Detroit, vania 15096, 1971. Michigan, January 1967. 18. J. H. McElhaney, R. L. Stalnaker, 10. A. W, Siegel, A. M. Nahum and D. E. V. L. Roberts and R. G. Snyder, “Door Crash- Runge , “Bus Collision Causation and Injury Patterns, ” worthiness Criteria, ” SAE No. 710864, Fifteenth SAE No. 710860, Proceedings of Fifteenth Stapp Stapp Conference, Society of Automotive Engineers, Car Crash Conference, Society of Automotive Inc. , Warrendale, Pennsylvania 15096, 1971. Engineers, Inc. , Warrendale, Pennsylvania 15096, 19. D. M. Severy, H. M. Brink and J. D. 1971. Baird, “Passenger Protection from Front-end 11. C. Clark and C. Blechschmidt, “Human Impacts , ” SAE No. 690068, International Automo- Transportation Fatalities and Protection Against tive Engineering Congress, Detroit, Michigan, Rear and Side Crash Loads by the Airstop Re- January 1969. straint,” Proceedings of Ninth Stapp Car Crash 20. D. F. IIuelke, J. C. Marsh, et al, Conference, Society of Automotive Engineers, Inc. , “Injury Causation in Rollover Accidents, ‘I Proceed- Warrendale, Pennsylvania 15096, 1965. ings of Seventeenth Conference of the American 12. National Highway Traffic Safety Adminis- Association for Automotive Medicine, Oklahoma tration, “49 CFR Part 571 (Docket No. 74-13; City, Oklahoma, 1973. Notice 1)) Federal Motor Vehicle Safety Standards, It 21, P. V. Hight, A. W. Siegel and A. M. Federal Register, Vol. 39, No. 54, Washington, Nahum, “Injury Mechanisms in Rollover Collisions, 11 D.C., March 1974. SAE No. 720966, Sixteenth Stapp Car Crash Con- 13. D. M. Severy, J. H. Mathewson and ference, Society of Automotive Engineers, Inc. , A. W. Siegel, “Automobile Head-on Collisions, Warrendale, Pennsylvania 15096, 1972. Series II, ” SAE Transactions, Society of Auto-