Rapidly expanding test programs in the field of Automotive Crash Safety Resea-rch have resulted in an influx of new engineering and photographic personnel who often lack an adequate background in the data gathering techniques presently being employed at major test sites.

What is the position, velocity, and attitude of an automobile at time T1 at T2? The same question might be asked of an object or occupant inside the auto during an impact test. How much damage was incurred, both in a qualitative sense, and in a quantitative sense? Manufacturers need answers to these questions to improve the safety of the automobile, and they frequently use photography to collect the test data.

While Hollywood and summer tourists primarily attempt to make pretty pictures - and sometimes succeed -, the instrumentation photographer basically endeavors to record data.

The Safety. Research. and Development Laboratory. of General Motors is responsible for all full scale vehicle collision tests, and impact sled programs, that are required by the Staffs and Divisions of the Corporation. The Laboratory is located at the GM Proving Ground in Milford, Michi gan. (Figure 1). In addition to full scale and sled tests, many component, static crush, and low speed impacts are conducted. All of these are required for engineering development and evHluation of occupant safety systems, or certification of compliance with the Federal Motor Vehicle. Safety. Standards.

The terms technology utilization and technology transfer are very much over-worked - a large number of books, reports, and papers have been written over the last ten years on these subjects. However, the majority of this effort is related to the fall-out from the military and space programs. There is no question that there is (and always has been) considerable fallout from these areas. A number of examples can be cited: the earliest of these is perhaps that of the development of tinned food (..or should I say canned food) to ensure the maintenance of the necessary supplies to Napoleon's army. The computer and radar are certainly important items from World War II that have had a tremendous impact on our lives; the orbiting weather and communication satellites are undoubtedly spin-offs from the NASA space programs.

A technique has been developed for measuring the six possible independent motions (three translational and three rotational) of a rigid body. The technique was developed specifically to measure head motions of an anthropometric dummy during simulated oblique and lateral crashes using an impact sled. Under such test conditions, body motions may be highly non-symmetric making a complete three-dimensional analysis necessary.

The dearth of comprehensive anthropometric and biomechanical information is a serious problem for automotive safety engineers and researchers. In this brief paper I want to focus particularly on the lack of quantitative three and four dimensional data about human body form and how stereometric analysis can be used to help fill the gap.

An important and substantial part of the large research effort presently devoted to automotive safety involves human protection in vehicles. Human protection requires knowledge of the motion of the occupant during a crash, that is, his displacement, velocity, and acceleration as they are influenced by seating, occupant size, and direction of impact. With a detailed description of the occupant's motion, restraint systems and other structural components can be designed to take best advantage of man's ability to resist force with minimal injuries.

As a preface to my presentation, I want to say a word about equipment. Like most people in industry we tend to use straight-forward, reliable equipment such as the PhotoSonics 1B and LOCAM cameras and the Vanguard Film Analyzer, believing that the advantages of more sophisticated systems are usually more than offset by the long down times required for repairs and by the increased likelihood of failure. We use relatively simple crystal controlled pulse generators for time-bases, for example, rather than real-time clocks that encode the time in some form and print it out in the gate at the center of each exposure. The pulse generators we use put out a pulse every millisecond, a double pulse every tenth time, and a triple on each hundredth. In regard to targets, we use the round, four sector, alternate yellow and black ones with which many of you are familiar. We feel this is extremely important because it's as unbiased a target as you could hope to find.

The Vignette Anomaly Monitor or VAM is an instrument for performing non-contacting dimensional measurements. These measurements are attained by multiple evaluation of shadows produced in broad daylight or in a typical manufacturing area. The VAM is not a laboratory curiosity but is rather an instrument which has been designed for production and for installation in stringent environments. The VAM has been selected over other instruments for dimensional measurement when high frequency response was required, when immunity to scintillation was desired, where it is of advantage to have no moving parts in the measurement system, where accuracies of the order of .001 inches are required, where it is of advantage to consider the shadow rather than the object, and where a stand-off distance of up to three feet provides or permits isolation between the measured object and the measurement reference frame.

Safety aspects of automotive systems are now being greatly emphasized by the pub lic, industry, and the government. Increased reliance on current and newly developing methods of nondestructive testing is required to better insure the safety related performance of automotive components.

Vehicle impact tests have three basic differences from most other automotive tests: 1. the duration of the test is short -less than one-quarter of a second; 2. the test item is destroyed -it is impossible to go back and run the test again if something goes wrong; and 3. they tend to be complex -several items are tested for any given impact.

When utilizing high speed motion picture cameras to collect data for engineering analysis the most difficult problem frequently centers around the uncertainty of the event. We have had ample opportunity to establish that the actual photographic problems in vehicle crash testing are relatively simple. However, with high film consumption rates, the camera must be triggered with very precise timing and synchronization becomes the real problem in crash testing. This is especially true with long duration events such as roll-overs. Economy suggests using short film loads, fast action requires high frame rates, and large numbers of cameras precludes manual starting. Some form of automatic triggering is required.

The studies involving the measurement of human eye movements are not new. In fact, their origin can be traced back by the work of early researchers such as Javal (1879), Landolt (1891), Dodge and Cline (1901), etc. (Reference 1, 2, and 3). The early methods for the measurement of eye movements involved direct observation of the eye in a mirror. Since then, a large variety of techniques have been used to record eye movements and these include techniques such as:

This paper presents a method of full-field displacement measurement which is very advantageous in auto-safety studies. A linear ruling of parallel lines is projected on the object with an ordinary projector. A view camera completes the apparatus and is used to image the object in its local focal plane. If the object is of arbitrary shape, the image of the ruling becomes "warped" when the camera axis is offset from the projector axis. This "warped" ruling contains information about the object contour which can easily be extracted. Superposition of a linear ruling with the warped ruling yields absolute contour information directly. Superposition of two warped rulings yield contour difference information. The latter process may be conventionally used for studies of deformations due to impact. Examples dealing with knee impact upon instrument panels will be shown. The sensitivity of the method may be varied over a wide range and the need for extreme degrees of vibration isolation is eliminated.

In 1968 a program of experimental and analytical studies on high-speed vehicle redirection by cable barriers was initiated at the National Research Council of Canada. The objective of these studies was to provide a rational basis for analysis of the design of highway cable barriers.

Planning for lighting for high-speed motion picture photography of test-sleds must be considered as a vital part of the entire project. The lighting to be used will be influenced by the camera used, the frame rate required, the film selected, the location and many other factors. If the lighting is put off until the very last as a relatively unimportant factor, you may be in for some unpleasant surprises.

An engineering picture, though impossible to beat for communications information, is incomplete! Lets say an "engineering happening" is completely described by 'who', 'what', 'when', and 'where.' Then the pictorial subject material is an excellent description of 'what', and fair on 'who', but is sadly lack-ing on 'where' or 'when.' 'Where' is often important for aerial photography at varying bearings and altitudes. Azimuths and evaluations are important, particularly with Cine theodolite. "Who' may be necessary information in multiple camera situations. "When', however, is vital!

At night, car driving requires as much - or perhaps even more - visual information as driving during day-time. The fact that generally speaking the night-time accident rates are some three times as high as the corresponding day-time rates must be at least partially contributed to the lack of visual information at night (swot/ 1969), more likely so if the fact is taken into account that on roads with good overhead street lighting the accidents are some 30% lower than on unlit or poorly lit roads, where car-drivers have to rely on their headlamps (OECD 1971, Box 1971, Cornwell and Mackay 1972). With present-day technology, driving a car involves the driver as an active element, and it can be done only when a large amount of visual information is presented. Details regarding the driving task may be found elsewhere (Anon 1971, Asmussen 1972), together with details regarding the role therein of transport lighting (Schreuder 1970a, 1971). For the present discussion, however, it is enough to summarize this task in two sub-tasks: primo following the road and secundo avoiding obstacles. As indicated, at day and with good overhead lighting these sub-tasks can be performed reasonably well; major problems, however, arise when vehicles have to proceed on unlit roads. They can be related to the general lighting requirements that can be described as: sufficient high and uniform road luminance, glare restriction, and visual guidance (Schreuder 1970b).