Search Results

This paper begins with a discussion of the various types of risk behavior found in automobile driving. A model is proposed which describes a perceptual approach to risk acceptance in specific driving manuevers. Risk acceptance in lane merging within freeway construction zones is discussed and exemplified by field observations of merging in construction zones. The short gaps accepted led to an experimental study to examine gap acceptance thresholds and driver gap estimation capabilities. Studies were conducted both staticly and dynamicly at 55 miles per hour. Subjects elect short gaps (under 100′) and gaps with less than 1 3/4 seconds at 55 miles per hour. Shorter rear gaps are accepted than front gaps. Gaps are grossly underestimated in both static and dynamic experiments. No difference was found for left vs. right merges and use of mirror vs. direct vision for rear gap estimation. The experimental data was consistent with field observations.

Assuming rigid body motion, recorded acceleration and recorded roll rates at the center of gravity, equations are used to calculate the local three-dimensional accelerations at hypothetical seating positions' Emergency Locking [seat belt] Retractors (ELR) during a steer induced rollover crash. For a threshold of 0.7 g, results demonstrated that intervals in the vehicle's response that may cause the ELR's inertial sensor to move into a neutral zone were limited to localized high magnitude negative vertical acceleration events during the rollover segment with a median duration of 4 ms, average duration of 4.8 ms and a maximum calculated duration of 31.7 ms. Changing the threshold to 0.35 g reduced the interval count by 70 percent and maximum duration by approximately 50 percent.

Plane strain bending (e.g. bending about a straight line) is a major sheet forming operation and it is practiced as brake bending (air bending, U-die, V-die and wiping-die bending). Precise prediction of springback is the key to the design of the bending dies and to the control of the process and press brake to obtain close tolerances in bent parts. In this paper, reliable mathematical models for press brake bending are presented. These models can predict springback, bendability, strain and stress distributions, and the maximum loads on the punch and die. The elasto-plastic bending model incorporates the true (nonlinear) strain distribution across the sheet thickness, Swift's strain hardening law, Hill's 1979 nonquadratic yield criterion for normal anisotropic materials, and plane strain deformation mode.

The motivation of this paper is inverse estimation of the material properties for sheet metals subjected to cyclic loading. Cyclic three-point bending tests are performed. Bending moments are computed from the measured data, namely, punch stroke, punch load, bending strain and bending angle. Bending moments are also calculated based on the selected material model in which normal anisotropy and combined hardening are considered. Material parameters are identified by minimizing the difference between these two bending moments. Modified Levenberg-Marquardt method is used in the optimization procedure. Stress-strain curves are generated with the optimized material parameters.

Reported tensile and fatigue properties of die cast AZ91D and AM60B magnesium alloys indicate that those values depend on the size and shape of the test samples and their global porosities. This paper reviews the mechanical properties reported in the open literature for these die cast alloys and indicates that section thickness and global porosity are inadequate for predicting the tensile and fatigue properties of die cast AZ91D and AM60B magnesium alloys.

For some vehicle dynamics applications, an estimate of a vehicle's center of gravity (cg) height and mass moments of inertia can suffice. For other applications, such as vehicle models and simulations used for vehicle development, these values should be as accurate as possible. This paper presents several topics related to inertial parameter estimation and measurement. The first is a simple but reliable method of estimating vehicle mass moment of inertia values from data such as the center of gravity height, roof height, track width, and other easily measurable values of any light road vehicle. The second is an error analysis showing the effect, during a simple static cg height test, of vehicle motion (relative to the support system) on the vehicle's calculated cg height. A method of accounting for this motion is presented. Similarly, the effects of vehicle motion are analyzed for subsequent mass moment of inertia tests.

We consider a vehicle-roadway system where the control of vehicle movement is based on the instrumentation located both in the vehicle and the roadway. In addition to the sensors which are used for obtaining the information on the vehicle, a radar based sensor system is used for providing information on the position of the car relative to a vehicle ahead, and with respect to a reflective strip placed on the road. The roadway traffic includes standard vehicles with no automatic control as well as the vehicles with automatic control units. Communication between the vehicles is not considered. For longitudinal control, we consider an Radar Based Cruise Control problem where the main goal is to maintain a desired speed set by the driver. At the same time, the controller will decelerate the vehicle if the distance and/or the relative speed between the controlled vehicle and the vehicles traveling in front are below certain limits.

The handling, stability, and rollover resistance of vehicles is presently being studied by both the automotive industry and the National Highway and Traffic Safety Administration (NHTSA). However, to study the handling and rollover behavior of each vehicle on the road is not feasible. The ability to categorize and compare the rollover and handling behavior of various vehicles is a subject of considerable research interest. This paper examines the possibility of characterizing vehicle classes through the use of a three degree-of-freedom linear model. Initially, segregation is studied by evaluating the eigenvalue location in the complex domain for vehicle sideslip velocity, yaw rate, and roll angle. Then the influence of numerator dynamics on vehicle behavior is studied and vehicle class segregation is attempted through evaluation of the amplitude ratio of the frequency responses for sideslip velocity, yaw rate, and roll angle.