An investigation of bus design parameters affecting rollover

Abstract

Reduction in the number and severity of bus and coach rollover accidents is the broad aim of this thesis. Chapter 1 outlines, from accident data, research needs and priorities regarding bus safety. Rollover accidents, particularly those due to sliding sideways into a "tripping" obstacle, are recognised as a major contributor to bus occupant casualties. Occupant injuries and fatalities are significantly higher if a coach rolls onto its roof or beyond, than if it rolls onto its side. A typical sequence of events leading to overturning accidents and consequently to occupant casualties is identified, and research needs into appropriate countermeasures for each step are suggested. Work towards two of these countermeasures is outlined. Chapter 2 describes the first area. A full-scale coach rollover test for measuring superstructure impact strength against a standard, required a mathematical model to explain and predict the motion of unsymmetric vehicles. A computer model was developed, and reasonable agreement between simulation and actual tests is demonstrated, despite the large unmeasurable energy losses hiring the "kerb-impact" phase of the tests. Research into "tripping" rollover stability, the second area, occupies Chapters 3 to 8. Experimental measurements of centre of gravity height, mass, moment of inertia, roll centre heights, suspension and tyre stiffnesses, and maximum stable tilt angle were undertaken on nine representative buses and coaches. A twelve degree-of-freedom computer simulation of a kerb-impact-initiated rollover for a vehicle with front and rear suspensions has been developed, Checked against other methods, and used to predict rollover threshold velocities (which ranged from 2.8 to 4.2 m/s for the tested vehicles). The effect of varying bus parameters on the rollover threshold was assessed and compared with their influence on tilt test performance as evaluated using a tilt test simulation. Centre of gravity height and track width had the greatest effect on results from both simulations. Correlation between tilt test results and computed rollover thresholds is reasonable. Both methods are recommended above a number of others as alternative ways of demonstrating adequate roll stability. Rollover threshold standards and new tilt platform limits are proposed, based on the current double-decker standard, but without self-levelling valves operating. Prohibition of tilt valves and disabling of self-levelling valves during a tilt test are recommended. These are some of the recommendations made to the U.K. Department of Transport concerning methods and standards of overturning stability testing. The major conclusions are presented in Chapter 9, which gives an overview of the thesis contents and findings.