Search Results

During recent years, the Highway Safety Research Institute (HSRI) has developed and validated two- and three-dimensional models describing the motions and forces acting upon an occupant during a collision. These inexpensive-to-operate models are performing with approximately 90% accuracy in parametric studies of classical crash configurations. In our own validation procedures, contacts with automobile development and design groups, and discussions with federal agencies, certain shortcomings of mathematical modeling procedures have been isolated. These include primarily the inability of the user to determine and input data to the computer programs and also to specify force, motion, velocity, and acceleration output data in a form applicable to the various vehicle design, human tolerance, and compliance tasks for which the models have been developed.

This paper presents the various features and operational properties of a two-dimensional mathematical model of crash victim motions. The earliest forms of this model can be traced to the early 1960s. Developmental work on two-dimensional models then continued both within the automotive industry and in independent organizations such as the Highway Safety Research Institute (HSRI). The most recent product of this activity is the MVMA two-dimensional mathematical crash victim simulation developed at HSRI for the Motor Vehicle Manufacturers Association. The features of this model include: 1. An eight mass representation of the human body where contact between the crash victim and the vehicle is represented in terms of independent force-deformation properties of the victim and the vehicle. 2. An extensible multi-joint neck and a realistically flexible shoulder joint. 3. A real-line representation of the vehicle interior or exterior where shape is given as a network of points. 4.

Drivers carried out a lane-changing and passing maneuver using convex and plane exterior mirrors alone or in combination with a plane interior mirror. The data showed that the addition of the plane interior mirror compensated for judgmental errors found when convex mirrors were used alone. When the speed difference was 15 mph between the overtaking car and the subject's car, subjects accepted gaps that were too short irrespective of the exterior mirror type. The data suggested that exterior convex mirrors of radii greater than 30 in. may be used reasonably safely by drivers and would have the advantage of providing a considerably increased field-of-view compared to currently used exterior mirrors.

The object of this research program has been to extend the scope of earlier work to include long-duration head impacts and to develop new scaling relationships to allow extrapolation of impact data from infrahuman primates to living humans. A series of living primate side impacts to the head and torso was conducted in parallel with a series of impacts to human cadavers. Dimensional analysis techniques were employed to estimate in vivo human tolerance to side injury. The threshold of closed brain injury to humans was found to be 76 g for a pulse duration of 20 ms and an impact velocity of 43 ft/s (13.2 m/s). The maximum tolerable penetration to the chest was found to be 2.65 in (6.72 cm) for both the left and right sides. Scaling of abdominal injuries to humans was accomplished by employing a factor that relates impact contact area, animal mass, impact force, and pulse duration to injury severity.

Direct impact to the larynx is usually prevented in accidents by the protective nature of the chin. In some situations, the occupant motions leave the larynx unprotected and susceptible to impact by the steering wheel rim or instrument panel. As one of the unpaired vital organs of the body, there is no easy way to provide an alternative for its functions when the larynx is lost or damaged. Information available on the tolerance of the unembalmed human larynx to force is quite limited. This paper describes a multidisciplinary study to determine the response of unembalmed human larynges to blunt mechanical loading and to interpret the response with respect to clinical data. Fresh intact larynges were obtained at autopsy and tested at either static or dynamic loading conditions utilizing special test fixtures in materials-testing machines. Load and deformation data were obtained up to levels sufficient to produce significant fractures in both the thyroid and cricoid cartilages.

An open-loop limit-maneuver test methodology was refined from an earlier study which hypothesized a relationship between vehicle performance and highway safety. Refinements in methodology were attained in the areas of test apparatus, test procedure, data processing, and performance interpretation. Open-loop response measurements were conducted on a representative sample of 12 contemporary passenger vehicles. Numeric characterizations of performance are presented, indicating the range and distribution of response properties exhibited by the vehicle sample.

Mathematical models of the human body subjected to an impact environment have been developed by many research groups in industry, government, private research organizations, and universities. In most cases, the models have not been verified by or compared with experimental results. The purpose of this paper is to show comparisons between the two- and three-dimensional crash victim simulators, which have been developed at the Highway Safety Research Institute of The University of Michigan, and front and side impact sled test results using anthropometric dummies.

A parameter study is performed involving several analytical vehicle occupant models in current use, with investigation of neck representations a primary goal. Side, oblique, and rear impact situations are investigated. Attention is given to the effects of varying head-neck mass and moments of inertia, anthropometry, muscle strength, and location, as well as well as strength, of motion-limiting “stops.” A model that replaces the conventional simple ball-joint neck with a two-joint, extensible neck is studied. This model also makes use of joint-stop ellipses to approximate the anatomical range for relatively free angular motion at a joint. Allowance is made for the effect of muscle contraction on occupant dynamics as a function of the degree of voluntary or involuntary tightening of the muscles, based upon experimental findings. A discrete parameter neck model that treats the cervical spine as a linkage of rigid vertebrae and massless, deformable discs is discussed briefly.

This paper discusses the problem of occupant protection in severe rear-end collisions from the standpoint of high performance seat structures and head restraints. Consideration is given to both fixed head restraints and to deployable head restraints. Two-dimensional computer simulations of occupant kinematics in a variety of rear-end collisions are utilized to provide initial performance criteria for head restraint design configurations. The resulting prototype system underwent a test and development program on an impact sled. The results of the various prototype performances and general criteria for high performance head restraint systems are discussed.

A series of mathematical models of the interaction between an occupant and the interior of a vehicle is presented. The following parameter studies using an eight-mass, two-dimensional model are discussed: belt material properties, belt slack, belt geometric configuration, and comparison of seats with and without headrests in rear impact. In addition, it is demonstrated by example that simple mathematical models can perform a valuable service in laying the groundwork for more sophisticated analytical and experimental work as well as yielding short term results. Finally, three-dimensional models are discussed. It is shown that a three-mass, three-dimensional model is a logical extension of current simulation efforts in order to provide insight into occupant response in oblique and lateral impact as well as nonsymmetric restraint systems.

This paper describes an injury criteria model implemented in computer language, and a restraint system effectiveness index for evaluating the degree to which the vehicle environment can prevent or reduce occupant injuries. The need for criteria of this type is based on the fact that if the degree of protection offered to a vehicle occupant by a restraint system or a vehicle interior (a function of the distribution and magnitude of the forces transmitted to the occupant) could be expressed in quantitative terms, then, more meaningful comparisons could be made between restraint configurations, and, areas of needed biomechanical research and statistical accident investigations could be more readily identified on the basis of the sensitivity of the results when the injury or effectiveness criteria are applied. The injury criteria model consists of three parts: 1.

A set of safety-relevant performance qualities for the passenger car has been defined as a first step in the development of objective measures of precrash safety performance. Measures were sought that stress the performance produced by a passenger vehicle when it is operated under emergency crash-avoidance conditions. This goal has led to the identification of six limit maneuvers and associated limit responses to serve as a first-order means of assessing the safety quality of a motor vehicle. The viability and the discriminatory power of the proposed test procedures have been demonstrated by applying these procedures to four separate vehicles.

The current federal accident data collection system is inadequate. It does not produce representative data essential for answering cause-and-effect questions concerning accidents, injuries, and fatalities, and it does not produce adequate data essential for conducting cost-benefit analyses of changes in vehicle designs, highway designs, or driver licensing policies. A proposed federal data collection system (SIR) can solve those problems at a total cost of about $6 million a year. The SIR system would include 30 investigating teams precisely located throughout the U.S., and would include a Sampling program, an In-depth program, and a Rapid-response program. The sooner this system is established, the sooner government and industry will begin to obtain accurate and reliable answers to pressing questions in the field of highway safety.

This paper describes the development of an improved neck simulation that can be adapted to current anthropometric dummies. The primary goal of the neck design is to provide a reasonable simulation of human motion during impact while maintaining a simple, rugged structure. A synthesis of the current literature on cervical spine mechanics was incorporated with the results of x-ray studies of cervical spine mobility in human volunteers and with the analysis of head-neck motions in human volunteer sled tests to provide a background for the design and evaluation of neck models. Development tests on neck simulations were carried out using a small impact sled. Tests on the final prototype simulation were also performed with a dummy on a large impact sled. Both accelerometers and high-speed movies were used for performance evaluation.

Blunt abdominal trauma is a major cause of death in the United States. However, little experimental work has been done to clarify the mechanism of blunt abdominal injury and to quantify tolerance parameters for the abdominal organs. This paper describes a joint study by the Highway Safety Research Institute and the Section of General Surgery of The University of Michigan in which direct impacts were applied to livers and kidneys. The tests were performed in a high-speed testing machine at a controlled ram velocity and stroke limit. The organ was surgically mobilized in anesthesized Rhesus monkeys and then placed on a load cell while still being perfused in the living animal. Tests were performed at ram speeds of 120, 6000, and 12000 in/min (5, 250,and 500 cm/s). The resulting load-deflection data were normalized and average stress-strain curves plotted for each test. In addition, the resulting injury severity was estimated immediately after impact using an injury scale of 1 to 5.

Serious injuries are caused to the chest and thoracic organs both in front and side automobile collisions, and statistical surveys indicate that overall chest injuries are the third most frequent after head and the lower limbs. For safer design of restraint systems and vehicle interiors experimental data has to be obtained to establish chest injury criteria. Unembalmed human cadavers were used to conduct nine frontal and fourteen lateral impacts including four with a simulated arm rest. All impacts used a six inch (15.2 cm) diameter impactor with impact velocities ranging from 12 mph (19.3 kph) to 20 mph (32.2 kph). Chest impacts were also conducted on rhesus monkeys and baboons to establish primate-human injury scaling criteria. Four human volunteers were used to obtain static load deflection curves in the lateral and frontal directions. The results of the above experiments and those conducted by other investigators are presented and analyzed.

In a previous study, an extensive study of the dynamic performance of child seating systems indicated that little protection was provided under circumstances other than panic braking. This study was performed with impact test conditions of 30 mph frontal and 20 mph lateral and rear barrier impacts with seating systems meeting the requirements of FMVSS 213. Additionally, the performance of prototype seats developed under contract with the Department of Transportation under similar test conditions will be presented to compare the protective qualities available with seats of current design and those that could become available in the future. The performance of the child seats will be evaluated using two criteria, motion limits and acceleration limits. It is believed that the performance of child seats can be determined without the extensive test equipment and facilities required for adult seating and restraint systems.

Driver-vehicle tests were performed in which the deceleration/pedal-force ratio (i.e., gain), pedal-displacement level, speed, surface-tire friction, and driver characteristics were systematically varied in order to determine the influence of these variables upon minimum stopping distance and other performance variables. Tests performed on a low coefficent of friction surface showed that high values of deceleration/pedal-force gain result in a greater number of wheel lockups and longer stopping distances compared to results achieved with intermediate or low deceleration/pedal-force gains. Tests performed on the two test surfaces with high and intermediate levels of friction showed that low deceleration/pedal-force gains produced longer stopping distances than were obtained with high gain, even though a high-gain brake system causes higher frequencies of wheel lockup.

A sample population of 51 male and 57 female subjects ranging in age from 18 to 78 years was assembled and tested in six different vehicles for preferred seat positions under non-driving and driving conditions. Volunteer subjects were selected by age, stature, and weight criteria in order to match the U.S. adult population to the extent practical. Preliminary analyses of these data suggest that on a total sample basis there is little difference between seat positions selected under non-driving and driving conditions, but that individuals may show significant differences. The small differences in group mean positions observed in this study may be due to a seat belt and/or an initial seat position factor. Post-drive seat position results were analyzed in a variety of ways to identify factors that may influence a person's preferred seat position.

A study of the biomechanical factors concerned with the design of side structures and doors for crashworthiness has been made. Questions regarding optimum stiffness, location of reinforcing members, effect of armrests, and padding have been answered within the framework of injury criteria models. Results of animal studies, cadaver studies, and anthropometric dummies have been combined to produce injury criteria for lateral impacts to the head, thorax, and abdomen. Impacts were applied utilizing a specially designed “air gun” in a laboratory environment emphasizing reproducibility and control. Full-scale crash simulations were performed on an impact sled to verify the results of the more specialized tests and analyses. Scaled models of current production doors were used in the animal series. Scaling relationships for various species of animals have been developed and extrapolated to man. Significant differences in right and left side tolerances to impact were noted and detailed.