Search Results

Rear-end crashes account for more than one in five fatal crashes in the U.S. The rear-end crash scenario most commonly associated with fatal crashes involves a following vehicle traveling 40 to 70 mph closing on a lead vehicle at a rate greater than 30 mph. The current research compiled an analysis of the literature to identify the kinematic factors, environmental factors, traffic-related factors and individual differences that are likely to influence drivers’ responses when closing on a slower-moving or stopped lead vehicle [LV]. In Part 1, several primarily kinematic-based methods for modeling drivers’ responses to a LV were compared for high-speed closing events. Methods utilizing looming (angular growth rate) equations were shown to predict drivers’ responses and time-to-contact methods (Inverse Tau) were conditionally accurate when applied to specific crash scenarios. However, the ratio or nominal response time methods did not predict drivers’ responses in most crash scenarios.

When the visibility of an object or person in the roadway from a driver’s perspective is an issue, the potential effect of moonlight is sometimes questioned. To assess this potential effect, methods typically used to quantify visibility were performed during conditions with no moon and with a full moon. In the full moon condition, measurements were collected from initial moon rise until the moon reached peak azimuth. Baseline ambient light measurements of illumination at the test surface were measured in both no moon and full moon scenarios. Additionally, a vehicle with activated low beam headlamps was positioned in the testing area and the change in illumination at two locations forward of the vehicle was recorded at thirty-minute intervals as the moon rose to the highest position in the sky. Also, two separate luminance readings were recorded during the test intervals, one location 75 feet in front and to the left of the vehicle, and another 150 feet forward of the vehicle.

Collision statistics show that more than half of all pedestrian fatalities caused by vehicles occur at night. The recognition of objects at night is a crucial component in driver responses and in preventing nighttime pedestrian accidents. To investigate the root cause of this fact pattern, Richard Blackwell conducted a series of experiments in the 1950s through 1970s to evaluate whether restricted viewing time can be used as a surrogate for the imperfect information available to drivers at night. The authors build on these findings and incorporate the responses of drivers to objects in the road at night found in the SHRP-2 naturalistic database. A closed road outdoor study and an indoor study were conducted using an automatic shutter system to limit observation time to approximately ¼ of a second. Results from these limited exposure time studies showed a positive correlation to naturalistic responses, providing a validation of the time-limited exposure technique.

This paper describes an investigation that seeks to understand how rollovers occur in real-world crashes, both by studying real world crashes and by analyzing vehicle handling tests to gain insights into potential mechanisms of pre-crash loss of control. In particular, this study focuses on one type of rollover, namely single-vehicle rollovers that follow a pattern of the vehicle first leaving the roadway and then returning to the roadway typically out-of-control. Aims of this study included the following: To describe the frequency and characteristics of single-vehicle rollovers involving an off-roadway excursion followed by a complete, if only temporary return to the roadway. To the extent possible, given available data, to assess the nature and consequences of driver inputs during the crash sequence. To define characteristics of crash scenarios which include a substantial proportion of this subset of single-vehicle rollovers.

Numerous studies have validated EDSMAC4 as an effective method of reconstructing automobile collisions; however, little has been done to investigate the effect of varying stiffness coefficients on the results of accident reconstruction and simulation analyses. When comparing simulations to staged collisions, the stiffness coefficients are frequently well defined; however, this is not always the case in real world accidents. Six vehicle-to-vehicle test impacts were modeled using EDSMAC4. Stiffness coefficients for the vehicles were obtained from test data of exemplar vehicles. After modeling the impacts with the base stiffness level, the stiffness coefficients were modified for both vehicles either plus (+) or minus (−) 25%. The impacts were re-run and the predicted vehicle damage (maximum crush and pattern), impact severity (Delta-V), peak acceleration, impact duration, post impact trajectory, and impact force was compared.

The purpose of this paper is to present an updated methodology for validating the visibility of objects in low-illumination digital photographs. A procedure is presented for creating prints from digital images that accurately represent the observers' view at a low-illumination scene. In addition, procedures are presented that allow representative displays of digital images on CRT and projector devices. The results indicate that the selected luminance settings of experimenters closely matched the selections of test subjects viewing similar images in print form and on CRT and projector displays. Further, both between-subject and between-experimenter variability was small.