Search Results

Rollover crashes are complex events that generate motions in all six degrees of freedom (6DOF). Directly quantifying the angular rotations from video can be difficult and vehicle orientation as a function of time is often not reported for staged rollover crashes. Our goal was to evaluate the ability of using a match-moving technique and consumer-grade video cameras to quantify the roll, pitch and yaw angles and angular velocities of a rollover crash. We staged a steer-induced rollover of an SUV at 106 km/h. The vehicle was fitted with tri-axial accelerometers and angular rate sensors, and five consumer-grade video cameras (2 on tripods, 2 on drones, 1 handheld, ~30 fps) captured the event. Roll, pitch and yaw angles were determined from the video using specialized software.

The purpose of this study was to characterize the kinematics of four Chevrolet Tracker rollover tests and to determine their average and intermediate deceleration rates while traveling on concrete and dirt. Single vehicle rollover tests were performed using four 2001 Chevrolet Trackers fitted with six degree of freedom kinematic sensors. Tests were conducted using a rollover test device (RTD) in accordance with SAE J2114. The test dolly was modified (resting height of the vehicle wheels was raised) between tests 1, 2, and 3. The RTD was accelerated to 15.6 m/s (35 mph) and then decelerated rapidly by an energy absorbing crash cushion (EA) to cause the vehicle to launch and roll. The vehicles initially rolled on a smooth concrete surface and continued into loose dirt. This paper adds to the body of work identifying phases of constant deceleration during staged RTD tests and compares these phases to the overall deceleration rate.

Vehicle rollovers generate complicated damage patterns as a result of multiple vehicle-to-ground contacts. The goal of this work was to isolate and characterize specific directional features in coarse- and fine-scale scratch damage generated during a rollover crash. Four rollover tests were completed using stock 2001 Chevrolet Trackers. Vehicles were decelerated and launched from a rollover test device to initiate driver's side leading rolls onto concrete and dirt surfaces. Gross vehicle damage and both macroscopic and microscopic features of the scratch damage were documented using standard and macro lenses, a stereomicroscope, and a scanning electron microscope (SEM). The most evident indicators of scratch direction, and thus roll direction, were accumulations of abraded material found at the termination points of scratch-damaged areas. Abrasive wear mechanisms caused local plastic deformation patterns that were evident on painted sheet metal surfaces as well as plastic trim pieces.

Estimation of target-to-background luminance ratios is a powerful method by which human detection of objects can be assessed. In the forensic community, evaluation of the detectability of a pedestrian to an automobile driver is often of interest. With calibration, the modern digital camera employing a CCD or CMOS light collection device can be a convenient and economical luminance estimation tool. Certain CCD or CMOS sensors will linearly report the impinging incident light pixel by pixel over a range of intensities. The device becomes nonlinear at low and high intensities; however, the linear region can be adjusted to the specific lighting conditions of interest by modifying the shutter speed, ISO setting, and aperture size. Image noise, sensor non-uniformity, temperature sensitivity, camera color sensitivity, and the spectral power distribution of the illuminant require treatment for direct comparison to the luminance.

The sampling rate and synchronization of the pre-impact data stored by General Motors (GM) sensing and diagnostic modules (SDMs) have not been experimentally determined. The goals of this study were to measure the time shift between the SDM-reported data times and algorithm enable, sampling rate variation and the synchronization of the sensor data. In this study, two experiments were performed. First, the SDM of a 2002 Pontiac Sunfire was artificially triggered while the throttle position, engine speed, vehicle speed and brake signals were also being monitored at their source sensors. Second, the throttle and vehicle speed sensors were replaced with artificially generated inputs so the timing of the SDM recorded values could be compared to that of the known inputs. Sampling rate and data synchronization were determined by fitting the SDM recorded values to the measured sensor outputs.

Crash data stored in the airbag sensing and diagnostic modules (SDMs) of General Motors vehicles can provide useful information for accident investigators. To quantify the accuracy and sensitivity of select 2003 to 2004 SDMs, two types of tests were performed. First, three 2004 vehicles underwent 136 vehicle-to-barrier and vehicle-to-vehicle collisions with speed changes up to 8 km/h. Second, 2003 and 2004 model year SDMs underwent a range of crash pulses using a linear sled. In all of the tests the speed change reported by the SDM underestimated the actual speed change. The speed change underestimates ranged from 0.2 to 2.9 km/h except for several anomalous tests in which the underestimate was as high as 12.3 km/h. The magnitude of this error varied with crash pulse shape. Increasing crash pulse duration and decreasing peak acceleration increased the difference between the actual and SDM reported speed change. The threshold accelerations for the SDMs tested ranged from 1.1 to 2.7g.

Collision data stored in the airbag sensing and diagnostic module (SDM) of 1996 and newer GM vehicles have become available to accident investigators through the Vetronix Crash Data Retrieval system. In this study, two experiments were performed to investigate the accuracy and sensitivity of the speed change reported by the SDM in low-speed crashes. First, two SDM-equipped vehicles were subjected to 260 staged frontal collisions with speed changes below 11 km/h. Second, the SDMs were removed from the vehicles and exposed to a wide variety of collision pulses on a linear motion sled. In all of the vehicle tests, the speed change reported by the SDM underestimated the actual speed change of the vehicle. Sled testing revealed that the shape, duration and peak acceleration of the collision pulse affected the accuracy of the SDM-reported speed change. Data from the sled tests were then used to evaluate how the SDM-reported speed change was calculated.

The purpose of this paper was to compare the damage to pickup truck bumpers produced by vehicle-to-barrier and vehicle-to-vehicle collisions of a similar severity, in order to determine whether vehicle-to-barrier tests can serve as surrogates for vehicle-to-vehicle tests in accident reconstruction. Impact tests were conducted on the front and rear bumpers of five pickup trucks. Each truck was subjected to an impact with a fixed barrier and with a passenger vehicle. All impacts resulted in pickup truck speed changes of about 8 km/h. Damage produced in the barrier and vehicle-to-vehicle collisions was similar if both collisions resulted in bumper mount damage on the pickup truck. If there was no bumper mount damage, then the bumper beam deformation depended on the shape of the impactor.

The severity of low-speed front and rear impacts is commonly assessed through testing of the actual or exemplar vehicles or by comparison with available barrier test data. While tests conducted on actual or exemplar vehicles are most accurate, barrier testing also provides dynamic properties of the bumper isolators in a controlled manner. The goal of this paper was to determine if barrier testing of a single isolator can be used to assess the behavior of an intact bumper on an actual vehicle. Replacement bumper isolators for 15 vehicles were individually mounted to a moving barrier and subjected to low-speed impacts with a fixed barrier to correlate their dynamic compression with impact severity. The single isolator results were compared with actual vehicle barrier test results. A total of 1624 tests were conducted. A 2nd order linear spring-mass-damper model was used to predict impact severity characteristics of the actual vehicle from the single isolator data.

During vehicle collision testing an accurate measure of the pre- and post-impact vehicle dynamics is necessary for analytical purposes. Sensors typically used for measuring vehicle speed change in low-speed collisions include 5th wheels, high-speed video, bumper-mounted load cells, and accelerometers. The method used is often based on equipment availability, the involved vehicles, and the type of tests being performed. The purpose of this paper was to quantify the relative accuracy of these four methods in aligned low-speed rear-end collisions. Data from 73 such collisions (clustered in two groups at target vehicle speed changes of 4 and 8 km/h) showed that all four instruments yielded statistically similar results for a target vehicle speed change of about 4 km/h, and that data derived from the 5th wheel and high-speed video were different than data from the other two sensors at the 8 km/h level.

Limited data exist which quantify the kinematic response of the human head and cervical spine in low-speed rear-end automobile collisions. The objectives of this study were to quantify human head/neck kinematics and how they vary with vehicle speed change and gender during low-speed rear-end collisions. Forty-two human subjects (21 male, 21 female) were exposed to two rear-end vehicle-to-vehicle impacts (speed changes of 4 kmlh and 8 km/h). Accelerations and displacements of the head and torso were measured using 6 degree-of-freedom accelerometry and sagittal high speed video respectively. Velocity was calculated by integrating the accelerometer data. Kinematic data of the head and C7-T1 joint axis in the global reference frame, and head kinematic data relative to the C7-T1 joint axis are presented. A statistical comparison between peak amplitude and time-to-peak amplitude for thirty-one common peaks in the kinematic response was performed.

Accurate assessment of the severity of a low speed impact between two vehicles can sometimes only be accomplished through staged collisions with the actual or exemplar vehicles. However, the cost of obtaining, colliding, then repairing the vehicles often precludes this option. For this paper, two prototype moving barriers were constructed to test three different bumper assemblies separate from their vehicles. Candidate bumper assemblies were mounted to the moving barriers for low speed impact testing with a stationary barrier and three other vehicles. Forty three test series of 701 total impacts were done to compare bumper performance in moving barrier tests with their in-situ counterpart. Vehicle-to-fixed barrier, vehicle-to-vehicle, moving barrier-to-fixed barrier and moving barrier-to-vehicle tests were done using four different vehicles. The actual vehicle and moving barrier results were statistically compared.

This paper continues the analysis of data published previously, focusing on steering wheel behavior and its correlation with driver fatigue (as measured by EEG, heart rate, and subjective evaluation of drowsiness from video). New steering-based weighting functions devised from observed changes in steering wheel motions are presented. Significant correlations between the weighting functions and the measures of driver fatigue suggest that some of the functions could form the basis of a fatigue-detection algorithm.

This paper evaluates the accuracy of different methods for determining the speed change during vehicle-to-vehicle collisions from isolator compression and low-speed barrier data. A controlled regimen of 938 aligned, low-speed collisions was completed, including a series in which collision force data were collected to compare vehicle-to-barrier and vehicle-to-Vehicle Collisions. Five vehicles (four with isolators and one with a foam-core bumper) were tested against a rigid barrier and against each other in collisions below damage threshold. Three methods of assessing the speed change of a low-speed vehicle-to-vehicle collision are evaluated as alternatives to a fourth method: staging collisions with exemplar vehicles. For each of the three methods, the expected accuracy and limitations are presented.

The driving performance of 17 heavy-truck drivers was monitored under alert and fatigued conditions on a closed-circuit track to determine whether driver fatigue could be indirectly measured in the vehicle control inputs or outputs. Data were recorded for various potential physiological indicators of fatigue (EEG, heart rate and a subjective evaluation of drowsiness), for vehicle speed, steering, and accelerator pedal movements, and for vehicle position on the track. The objective was to determine whether a simple set of vehicle-based control measures correlated with the fatigue indicators. Correlations between other vehicle-based measures reported in the literature and the fatigue indicators were also calculated. The results indicate that there are measures which correlate sufficiently well with driver fatigue that they could potentially be used for an unobtrusive vehicle-based fatigue-detection algorithm.

This paper describes the instrumentation used to study how the control inputs of 17 long-haul truck drivers were affected by fatigue. The task required outfitting a test vehicle to accurately measure the following control inputs and resulting vehicle behavior: Vehicle speed, Steering wheel angle and angular velocity, Accelerator pedal angle and angular velocity, Perception/response time, Driver EEG and heart rate, Clinical assessment of driver fatigue, Vehicle lateral lane position, and Car-following distance. The location and mounting procedure of each instrument as well as the sampling requirements for each device are discussed. Also discussed are the methods of data handling and storage.

There are many bumper-to-bumper automobile collisions in which a vehicle occupant claims injury but where there is little or no outward damage to the vehicles. On vehicles equipped with shock-absorber-type bumper isolators, the only “damage” often consists of compression marks left on the isolator piston tube and scuffs on the bumper. This paper examines the behavior of specific automobile bumpers in aligned low-velocity collisions. Specifically, empirical data gathered during numerous (currently 660) vehicle-to-vehicle and vehicle-to-barrier collisions are presented and relationships between isolator compression and vehicle impact severity are developed. General trends among all types of isolators and trends specific to vehicle manufacturers are identified and discussed. Damage threshold data are also presented.

The objective of this paper is to examine automobile bumper systems in aligned low-speed impacts and provide data which correlate compression of bumper systems with the vehicle impact severity. A significant number of automobile collisions involve bumper-to-bumper contact at speeds which produce little or no permanent vehicle damage. Contemporary bumper systems predominantly consist of a fascia and impact beam, which span the vehicle width, and some form of impact absorber. A common impact absorber is the shock-absorber-type isolator. Foam cores, deformable steel struts, rubber shear blocks and leaf springs also exist. Data from 58 vehicle-to-barrier and 136 vehicle-to-vehicle aligned impacts are presented. Impact duration, speed change, isolator compression, and coefficient of restitution results are presented and discussed. Static and dynamic compression tests on several isolators have been carried out.