Search Results

The purpose of this study was to reconstruct the accident situation of a Golf-Cart with three persons in it. The cart was designed to carry two persons and their golf bags, someone had fixed an add-on cab. Tilt-table testing was carried out to estimate the rollover threshold and to measure the center-of-gravity location of the vehicle, which was used to do preliminary stability analysis for different loading conditions. Tests were carried out with and without the add-on cab. The add-on cab deteriorated the rollover characteristics by raising the center-of-gravity height, while the presence of an additional rider on the rear-left wheel arch, biased the load distribution. These two effects resulted in reducing the rollover threshold. Generally, tilt-table test results overestimates the rollover threshold, so there is a good possibility in real world situations the acceleration experienced by the vehicle during steady-turn maneuvers on an even surface could lead to rollover.

A hybrid approach has been developed to determine the dynamic vehicle rollover threshold. The approach uses a roll-plane vehicle model and a simulation based on the testing that was carried out at low speeds, where lateral acceleration are measured. Then, curve fitting techniques are used to extrapolate for the threshold values of lateral acceleration and forward velocity. This approach was validated using lane-change maneuver test data and a good correlation was obtained. Further, this approach is well suited as an objective test procedure to determine the rollover threshold as only lateral accelerations need to be measured.

In 1976 the author presented an SAE paper on ride quality in which a two-degree-of-freedom ride quality simulation was developed for measuring vertical motions. The present paper reports on the work done since that time. Side-to-side motion has been added to the simulator developed for the Association of American Railroads as well as the studies needed to find the impedance of humans when subjected to side-to-side and fore-and-aft motion. These studies found that rough rides are better endured during warmer weather and that women do not resonate as readily as men at the two critical and annoying side-to-side frequencies. Last, the paper deals with the use of a rigid mass to measure seat acceleration. Through a transfer function, a computer simulation has been developed to find the vertical seat acceleration.

Extensive testing was performed on both the ring-gear and hypoid-gear churning used to move the oil in the axle bowl to lubricate tandem axles. The gear pump test results were used to establish the pump characteristics, as well as the volumetric and mechanical efficiencies. The churning test results were used to formulate an empirical model used in an analysis to predict the energy loss. All testing was done with SAE 90W and 140W lubricants. The results from these experiments, along with some analytical formulations, were used and incorporated into a computer program. The results of the component testing showed that the hypoid-gear churning loss is a large energy loss, amounting to as much as 1.5 horsepower per axle. Since the energy losses in the gear pump amounted to only 0.2 horsepower, around a one horsepower savings per axle is achievable by lubricating the entire axle with the gear pump.

This paper describes the use of a measurement and data reduction system for objective ride quality assessment which can be used in all types of transportation systems. The basis of the data handling is the use of the absorbed power criteria developed by Pradko and Lee, modified by using the Amplitude Frequency Distribution (AFD) method, and incorporating a human simulator to obtain the objective parameters used in the ride calculations. The AFD method was originally developed by the authors as a better description of road roughness than commonly used random data descriptions. The method is a joint probability density of amplitude and frequency and therefore maintains both spectrum data and amplitude distribution. The human simulator was developed to eliminate the need for extensive subjective testing. The study showed that in the frequency range of interest to ride quality work, a two-degree-of-freedom simulator adequately modeled human response.