Search Results

This paper presents a methodology for validating vehicle stability and control computer simulations. Validation is defined as showing that, within some specified operating range of the vehicle, a simulation's predictions of a vehicle's responses agree with the actual measured vehicle's responses to within some specified level of accuracy. The method uses repeated experimental runs at each test condition to generate sufficient data for statistical analyses. The acquisition and reduction of experimental data, and the processing path for simulation data, are described. The usefulness of time domain validation for steady state and slowly varying transients is discussed. The importance of frequency domain validation for thoroughly validating a simulation is shown. Both qualitative and quantitative methods for the comparison of the simulation predictions with the actual test measurements are developed.

This paper summarizes the results of test maneuvers devised to measure on-road, untripped, rollover propensity. Complete findings from this research are contained in [1]. Twelve test vehicles, representing a wide range of vehicle types and classes were used. Three vehicles from each of four categories: passenger cars, light trucks, vans, and sport utility vehicles, were tested. The vehicles were tested with vehicle characterization and untripped rollover propensity maneuvers. The vehicle characterization maneuvers were designed to determine fundamental vehicle handling properties while the untripped rollover propensity maneuvers were designed to produce two-wheel lift for vehicles with relatively higher rollover propensity potential. The vehicle characterization maneuvers were Pulse Steer, Sinusoidal Sweep, Slowly Increasing Steer, and Slowly Increasing Speed. The rollover propensity maneuvers were J-Turn, J-Turn with Pulse Braking, Fishhook #1 and #2, and Resonant Steer.

Phase IV of the National Highway Traffic Safety Administration's (NHTSA) rollover research program was performed during the spring through fall of 2001. The objective of this phase was to obtain the data needed to select a limited set of maneuvers capable of assessing light vehicle rollover resistance. Five Characterization maneuvers and eight Rollover Resistance maneuvers were evaluated [1]. This paper is “Volume 2” of a two-paper account of the research used to develop dynamic maneuver tests for rollover resistance ratings. Test procedures and results from four Rollover Resistance maneuvers are presented. The Consumers Union Short Course (CUSC), ISO 3888 Part 2, Ford Path Corrected Limit Lane Change (PCL LC), and Open-Loop Pseudo Double Lane Changes are discussed. Details regarding the NHTSA J-Turn, and the three fishhook maneuvers are available in “Volume 1” [2].

Phase IV of the National Highway Traffic Safety Administration's (NHTSA) rollover research program was performed in 2001, starting in the spring and continuing through the fall. The objective of this phase was to obtain the data needed to select a limited set of maneuvers capable of assessing light vehicle rollover resistance. Five Characterization maneuvers and eight Rollover Resistance maneuvers were evaluated [1]. This paper is “Volume 1” of a two-paper account of the research used to develop dynamic maneuver tests for rollover resistance ratings. Test procedures and results from one Characterization maneuver (the Slowly Increasing Steer maneuver) and four Rollover Resistance maneuvers are discussed (the NHTSA J-Turn, Fishhook 1a, Fishhook 1b, and Nissan Fishhook). Details regarding NHTSA's assessment of the Consumers Union Short Course (CUSC), ISO 3888 Part 2, Ford Path Corrected Limit Lane Change (PCL LC), and Open-Loop Pseudo Double Lane Changes are available in “Volume 2” [2].

This paper discusses the importance of having an adequate model for the dynamic response characteristics of tire lateral force to steering inputs. Computer simulation and comparison with experimental results are used to show the importance of including appropriate tire dynamics in simulation tire models to produce accurate predictions of vehicle dynamics. Improvements made to the tire dynamics model of an existing vehicle stability and control simulation, the Vehicle Dynamics Analysis, Non-Linear (VDANL) simulation, are presented. Specifically, the improvements include changing the simulation's tire dynamics from first-order system tire side force lag dynamics to second-order system tire slip angle dynamics. A second-order system representation is necessary to model underdamped characteristics of tires at high speeds. Lagging slip angle (an input to the tire model) causes all slip angle dependent tire force and moment outputs to be lagged.

The goal of this research was to find vehicle characteristics which may contribute to steering maneuver induced rollover accidents. This work involved studying vehicle handling dynamics using the Vehicle Dynamics Analysis, Non-Linear (VDANL) computer simulation. The simulation was used to predict vehicle responses while performing 28 different steering induced maneuvers for each of 51 vehicles. Various measures of vehicle response (metrics), such as response times, percent overshoots, etc., were computed for each vehicle from simulation predictions. These vehicle directional response metrics were analyzed in an attempt to identify vehicle characteristics that might be good predictor/explanatory variables for vehicle rollover propensity. The metrics were correlated, using the Statistical Analysis System (SAS) software and logistic regression, with single vehicle accident data from the state of Michigan for the years 1986 through 1988.

This paper examines variability of two static rollover metrics, Static Stability Factor (SSF) and Tilt Table Ratio (TTR), due to vehicle loading and vehicle-to-vehicle variation. Variability due to loading was determined by measuring SSF and TTR for 14 vehicles/configurations at multiple loadings. Up to five loadings were used per vehicle/configuration tested. Vehicle-to-vehicle variability was studied by measuring SSF and TTR for ten unmodified vehicles of each of four make/models. Five baseline vehicles, as similar as was feasible, were tested. The other five test vehicles spanned the range of submodels and options available. In general, both SSF and TTR decreased as occupants were added to a vehicle. The change in SSF and TTR per occupant was fairly consistent, with changes in TTR being more consistent. Placing ballast on the floor of the cargo compartment had a mixed effect on SSF, raising it for some vehicles and lowering it for others.

Most fatalities due to school bus accidents involve pedestrians being struck by the bus. All too frequently the school bus strikes a disembarking passenger because the driver was unaware of their presence near the bus. To try to prevent this type of accident, two Doppler microwave radar-based pedestrian detection systems have been developed and are commercially available. These systems supplement regular school bus mirrors. They operate only while the bus is stationary. Both systems detect moving pedestrians either directly in front of or to the right of the bus. The National Highway Traffic Safety Administration has performed a three-part evaluation of these pedestrian detection systems. The first part measured the field of view of each system's sensors. The second part evaluated the effectiveness and appropriateness of each system's driver interface. The third part was a small-scale operational evaluation.

Two computer models of drivers using describing function strategies have been successfully implemented in conjunction with a recently developed, all digital vehicle simulation. The driver models determine control inputs to the vehicle simulation by means of feedback loops. Two feedback loops, an outer one on lateral position and an inner one on heading angle are used to determine the steering commands needed to move the vehicle to the desired path. One feedback loop on forward velocity is used to determine braking and acceleration commands. Full technical details of the method of implementation for each of the models are given. The results of sample simulations of the driver-vehicle system are shown and the results discussed.

Two computer models of drivers using preview predictor strategies have been successfully implemented in conjunction with a recently developed, all digital vehicle simulation. The driver models determine control inputs to the vehicle simulation by first predicting future vehicle position and velocity and then determining the steering and braking commands necessary to move the vehicle from the predicted to the desired path. Full technical details of the method of implementation for each of the models are given. The results of sample simulations of the driver-vehicle system using each driver model are shown. Problems of each model are discussed.