By Derwyn M. Severy, Harrison M. Brink, and Jack D. Baird
Institute of Transportation and Traffic Engineering,
University of California, Los Angeles. 1967

Three thousand seven hundred children were injured in school bus accidents during the year 1965, representing an increase of 75% over 1960. During the same period, there was only a 45% increase in the number of pupils being transported by school buses. Prior to these experiments no organized research program had been established to determine design criteria for school bus body structures and for safety equipment included within the bus to reduce passenger injuries during collision.

During the past sixteen years, ITTE-UCLA has conducted ninety full-scale automobile collision experiments and several hundred related laboratory experiments; additionally, the Research Staff has investigated school bus and passenger bus accidents during the past ten years (1)* for the purpose of gaining background information on the factors accounting for passenger injuries, as well as accident causation.

A series of school bus crash tests using store-type mannequins without instrumentation was conducted at Little Rock and Conway, Ark., in 1964. This pioneering work provided only generalized observations owing to lack of instrumentation and the application of research techniques. In another effort, a commercial intercity bus was crashed by General Motors (2) and carried a moderate amount of instrumentation; this GM test represents the only known sophisticated experiment conducted on bus collision performance prior to the UCLA study, but did not concern school bus conditions.

[Author's note: *Numbers in parentheses designate references at end of paper.]

ABSTRACT

The following categories relating to passenger injury causation were studied: location and type of impact, structural integrity of vehicles, vehicle size, seat design, type of restraint or force moderator, type of safety glass, passenger size, standing versus seated passengers, passenger kinematics and interactions, force sustained by passengers, and many related factors.

Electronic instrumentation consisted of 61 transducers positioned in the anthropometric dummy passengers, on the safety belts, and on the vehicles to record accelerations and forces during collision. Photographic instrumentation included thirty-three high speed motion picture cameras and special photographic devices that were arranged within, around, and above the colliding vehicles to provide detailed observation of all aspects of these collision experiments.

OBJECTIVE

CONCLUSIONS

METHODOLOGY - The foundations of scientific inquiry is its methodology. It is the procedures devised for evolving information not commonly available and not readily verifiable. The confidence in the data subsequently developed is dependent on the reputation of the investigators and the soundness of their methodology.

1. The three collision experiments reported in this paper provide conditions sufficiently realistic and severe to adequately evaluate the relative merits of various school bus passenger protective devices and to identify other injury-producing factors.

2. The three school bus collision experiments reported this paper are representative of nearly all collision exposures by school bus passengers.

3. The use of 3-, 6-, 13-year-old, and adult sizes of anthropometric dummies provided a practical evaluation of the effect of passenger size on school bus passenger collision safety.

4. On a selective basis, variations in seat types, restraint types and related protective devices were evaluated under realistic conditions.

5. Comprehensive instrumentation was used in these studies to provide detailed specifics required for performance evaluation of school bus passenger safety.

6. The utilization of instrumented trauma-indicating anthropometric dummies simulating passenger collision, induced movements, high speed motion picture color photography overlapping all collision movements, and full size vehicles providing realistic collision conditions, represents a most practical and reliable method for determining school bus collision performance.

7. The presence of double or higher order variables was controlled through appropriate methodology to ensure reliability of findings.

COLLISION PERFORMANCE - The susceptibility to passenger compartment encroachment, the comparative resultant accelerations, the influence of vehicle components on injury causation, and the preservation of passenger compartment integrity while undergoing moderate levels of impact acceleration are examples of vehicle collision responses characterizing collision performance. In the ordinary use of a vehicle, these deficiencies do not usually manifest themselves and frequently escape observation by accident investigators owing to the transient nature of these deficiencies or the investigator's lack of familiarity with levels of collision performance. Adequate collision performance provides the passenger with a protective shield from the crashing structures of the primary impact; adequate passenger compartment safety protects the passenger from the injury-producing forces of the secondary impact, the one in which the passenger may be hurled against the compartment interior or ejected. This section relates to the former and the following section, to the latter.

1. All heavy vehicle designs and more particularly those classified as buses, must include provisions for collision-resistant structure at the passenger car bumper height as well as at the heavy truck bumper height.

2. Bus design should ensure that the passenger compartment is securely attached to the frame of the bus by appropriately sized shear bolts at frequent intervals from front to rear and along both frame members.

3. The adverse performance of bus steering wheel assemblies during front-end collisions corresponds to that of other pre-1967 vehicles.

4. The new laminated glass having a high-energy interlayer with controlled adhesion, represents a substantial improvement for collision performance of windshield.< ul>

BUS INTERIOR - The school bus, fundamentally, is a "kingsize" passenger relating to passenger safety developed by the Society of Automotive Engineers, the automotive safety standards evolved by the United States General Services Administration, the Federal Motor Vehicle Safety Standards published by the Department of Commerce, as well as those standards currently under development, should be applied to buses, with very little revision required.

1. Tubular struts, protruding hand grips and similar protruding rigid structures should be eliminated.

2. Thin padding, less than 1/2 in., applied to tubular struts and similar rigid structures serves little practical value.

3. Force amplifying structures should be relocated, recessed, or eliminated.

4. Glass should remain in place, even after sustaining head or shoulder impacts.

5. Window-glass impact performance of buses depends on many factors; the relative performances of laminated and tempered glass for these collisions should not be compared without cautious consideration of the varying circumstances of exposure.

SEATING UNITS - Seat designs ranging from frameless air seats to nonpadded hard fiberglass shell seats with tubular frames were evaluated; some seats had no head support, no lateral support and no padded backrest (to protect passengers from the rigid knife - like seatback) while some were safety seats, designed and shown substantially to protect the passengers from abusive collision forces, regardless of direction. Properly designed bus seats provide an inner protective shield around their precious cargo while also compartmentalizing the passengers to reduce the possibilities of their interacting with each other during all but the most devastating of collisions. In general, seats in buses are the initial and very frequently the only structure contacted by passengers during collision. Special attention to their construction can contribute very markedly to bus passenger safety.

1. Low back seat units, seatback height less than 28 in., greatly increase chances to injuries during school bus accidents.

2. School bus seat anchorages and seat cushion fasteners should not fall from forward decelerations under 30 G and should comply with other related performance criteria that become a part of the Federal Motor Vehicle Safety Standards.

3. Seats not designed to accommodate the added stress of multiple lap belts attached to the seat can be retrofitted with a satisfactory structure to accomplish this modified performance.

4. Seatback strength should include allowance for passenger thrown forward against the backrest.

5. Elastic rebound of seatbacks increased the chances of passengers sustaining multiple impact injuries.

6. Plastic deformation of high seatbacks reduce lap-belted passenger's rebound towards seat ahead in rear-end collisions but greatly increase chances of injury for passenger thrown against them.

7. For the moderately severe collision exposures reported in this paper, it was established that a well-designed safety seat would protect passengers from sustaining more than minor injuries.

9. Seat belts recommended for safety seats. These bus experiments, the many actual school bus accidents investigated by the authors, the many types of collision experiments conducted during the past 16 years by the authors and investigations by others, clearly establish the value in passenger protection of lap belts when used with high back seats. The greatest single contribution to school bus passenger collision safety is the high strength, high back safety seat. Next in importance is the use of a three-point belt, a lap belt or other form of effective restraint. These restraints can be added to the safety seat at very little added cost and their presence provides the continuity needed for proper training of youth concerning habitual use of restraints when riding in any vehicle.

10. Low back seats (backrest than 28 in.) common to school buses built in 1966 and earlier should not be retrofitted with lap-type safety belts, unless the low backrests are replaced with adequately designed high backrests. During front-end impacts and following rebound from their seat-back for rear-end collisions, the lap-belted passenger pivots about his belt and slams his head, face, and if tall enough, his chest into the seatback ahead. The low back seat presents dangerous surfaces to the belted or unbelted passenger hurled forward against it during collision. In addition, exposure to serious back and neck injuries results when passengers in low back seats experience a rear-end collision. Forces to the passenger as a result of a rear-end collision are increased if a lap belt is worn because it secures the hips thereby intensifying the fulcrum-like action of the seatback forces.

11. Seatbacks and armrests should be designed using well-padded, broad surfaced metal frames designed to provide the required strength and attenuate head impact forces in accordance with the performance specifications of the Federal Motor Vehicle Safety Standard No. 201, S 3.2.

12. The narrow, this padding covering rigid tubular structures such as the tops of seatbacks, and so forth, represents an unsatisfactory solution to the problem of an inadequate design.

13. Seats should not be provided with rigid protruding structures such as handgrips, handrails and similar injury producing fixtures.

14. The air seat did not impress the authors as being a practical approach to school bus passenger protection during collisions.

15. School bus seats at the time of this study are grossly inadequate for protecting passengers.

RESTRAINTS - A list of passenger-protective devices would generally show restraints at the top, with the seat a close second. The reason passenger seats are regarded as more important than individual restraints for the protection of school bus children is the close proximity of the occupants to school bus seats as contrasted to passenger car seats. The compartmentalization provided by school bus seats, if they are high back seats, serves as a very valuable constraint for all horizontal directions of impact. The performance of safety belts and harnesses in this study followed the lines clearly established in prior experiments. (3-5) Properly designed restraining devices direct collision forces to the strong parts of the body in a manner least likely to produce injuries.

2. The cross-chest lap-belt combination when properly fitted provides significantly more passenger protection than does the use of only a lap belt. A comparison was made between performances of three-point and lap belts in the prior conclusion. In contrast with no belt, the three-point belt allows its wearer to sustain but one-third the crash forces received by an unrestrained passenger of the same size seated beside him. More importantly, the forces are directed by the three-point restraint system to strong parts of the passenger's body in a generally noninjury-producing manner, as contrasted with head and chest injuries commonly sustained for unrestrained passengers on direct impact with the structures around them.

3. The cross-chest lap-belt combination restraint is not recommended for use in school buses.

4. Seats having strong well-padded armrests provide important lateral constraint.

5. The restraint bar provides acceptable restraining action against front-end type collisions but does not provide restraint against the lateral forces of a side impact.

6. The restraint bars of the type tested in these experiments are not recommended for school buses.

7. The air bag provides good impact attenuation for passengers thrown against it; further research is recommended before a decision can be made concerning its practicality for school buses.

OCCUPANT KINEMATICS - The forced body movements of passengers during collision are not readily predictable, except by reference to observations from full-scale collision research. The complications of relative impacting masses; positions of contract and the resulting dynamic centers of rotation of vehicles; location of passenger and his size in relationship with the constraints he encounters within; and on ejection from the vehicle, each typify the variables that challenge credibility of information not based on full-scale research. An understanding of human body collision kinematics provides the basis for practical and effective design of passenger restraining devices and of vehicle interiors that provide the most protection for the money invested. These experiments have shown that, depending upon the nature and extent of passenger collision protection, school bus occupants may be killed or sustain no injury even though subjected to identical collision conditions. This subject has been extensively investigated by the authors (3-5, 7) and by other investigators (8).

1. Bus drivers and truck drivers should be required to wear at least a lap-type safety belt whenever their vehicle is in motion to assure that they remain "behind the wheel" during an accident.

2. Passenger collision kinematics are sufficiently varied from crash to crash that there is no "safe seat" and also there is no "death seat."

3. The path the body travels during collision is not definable in simple terms but rather depends on many factors, each including many variations.

EVACUATION - The orderly exiting of many people under emergency conditions poses a problem requiring special measures; this is true whether the evacuation relates to a burning building, a sinking ship, a crashed aircraft, or bus. The adversities of congested space, injured and dazed people should not be complicated by too few and too inadequate escape routes. In addition to evolving more emergency exits for school buses, studies are needed to give direction to design engineers for evolving expeditious yet safe and practical emergency escape systems. In addition, school districts should be required to conduct practice emergency exit demonstrations so that the younger passengers can manage their own escape during an accident that incapacitates the driver. The ever-present hazard of past-crash fire necessitates prompt and orderly evacuation by all able passengers in order to improve chances of rescue of those unable to help themselves.

1. Because of the hazard of crash-induced fire following moderate to severe collisions, passengers should be evacuated promptly and required to stand remote from traffic and the accident scene.

2. Because the driver is generally the only responsible adult in the bus, he should be protected from collision injury by at least a seat belt restraint and a well designed high back safety seat with padded armrests.

3. School buses need at least four full size emergency escape routes of a standardized design and location.

REGULATORY STANDARDS - Experience has shown that in matters affecting public safety, specific performance and functional guidelines evolved by specialists are necessary to assure that economical considerations do not suppress safety. This section is not intended to be complete; the authors felt that certain conclusions belong under this category but remain hopeful that all findings will be carefully examined by committees that are charged with formulating regulatory standards for buses, including school buses.

1. Safety belts should be worn at all times by the drivers of buses and other heavy vehicles.

2. Safety regulations for school buses should be standardized without allowances for the intended size of school passengers to be hauled.

3. The bumpers of all buses (and other large vehicles) should be capable of effectively transferring collision forces to heavier structural members of the bus frame and should be positioned with the base not more than 16in. above the pavement for the unloaded vehicle.

4. Seatback height for school buses should not be less than 28 in.

5. Built-in or improved seating in the aisle should not be permitted in buses.

REFERENCES:

1. D. M. Severy, "Application of Collision Research Findings to School Bus Passenger Safety." Transactions, National Safety Council, (October 1964), pp. 78-81.

2. D. J. LaBelle, "Barrier Collision and Related Impact Sled Tests on Buses in Intercity Service." Seventh Stapp Car Crash Conference -- Proceedings, D. M. Severy, ed., Charles C Thomas: Springfield, Ill. (1965), pp. 46-53.

3. D. M. Severy, J. H. Mathewson, and A. W. Siegel, "Automobile Head-On Collisions, Series II." SAE Transactions, Vol. 67 (1959), pp. 238-262.

4. D. M. Severy, J. H. Mathewson, and A. W. Siegel, "Automobile Side-Impact Collisions, Series II." SAE Pre-print 491A, SP-232, 1962.

5. D. M. Severy, "Engineering Studies of Motorist Injury Exposures from Rear-End Collisions." National Academy of Engineering Proceedings, Washington, D. C., April 1966.

6. D. M. Severy, "Human Simulation for Automotive Research." SAE SP-266, January 1965.

7. D. M. Severy and H. M. Brink, "Safety Glass Breakage by Motorists During Collision." Ninth Stapp Car Crash Conference -- Proceedings, University of Minnesota, Minneapolis (1966), pp. 226-234.

8. Arthur D. Little, Inc., "The State of the Art of Traffic Safety." A Critical Review and Analysis of Technical Information on Factors Affecting Traffic Safety, prepared for Automobile Manufacturers Association, Inc. (June 1966), p. 624.