Modelling injury severity of bicyclists in bicycle-car crashes at intersections

Bicyclists are vulnerable road users as they are not protected during a road collision. Although numerous studies have been conducted to understand the parameters contributing to bicyclist's injury severity, most of these studies have focused on the relationship between crash severity and road, environmental, vehicle and human demographic parameters. No study has been found that investigated the relationship of bicyclist's injury severity with speed and mass of both vehicles, as well as other crash dynamics aspects. This study developed a modelling framework to investigate the effect of variables such as speed, mass and crash angle on bicyclist's injury severity in bicycle-car crashes at intersections. A combination of Newtonian Mechanics and statistical analysis was utilised to develop this theory. This modelling process followed a two-step approach. In the first step, Newtonian Mechanics was used to develop numerical models to estimate the impact force applied to the bicyclist. Variables affecting the associated impact forces were then identified. In the second step, a mixed binary logistic regression model was developed to estimate injury severity of a bicycle-vehicle crash as a function of mass of both vehicles, speed of both vehicles before and after the crash, restraint use and age of bicyclist. Transport Accident Commission (TAC) validated crash data was used to develop the model. The results of the numerical models showed that kinetic energy of the car before crash and kinetic energy of the bicycle after crash are important parameters affecting the injury severity of the cyclist in bicycle-vehicle crashes. The results of the mixed binary logistic regression model confirmed that the addition of kinetic energy of the car before crash and the kinetic energy of the bicycle post-crash had a statistically significant effect on injury severity of bicyclist. The results further showed that older bicyclists were involved in higher severity crashes and helmet-wearing reduced the injury severity of the bicyclist.

A kinetic energy model of two-vehicle crash injury severity

An important part of any model of vehicle crashes is the development of a procedure to estimate crash injury severity. After reviewing existing models of crash severity, this paper outlines the development of a modelling approach aimed at measuring the injury severity of people in two-vehicle road crashes. This model can be incorporated into a discrete event traffic simulation model, using simulation model outputs as its input. The model can then serve as an integral part of a simulation model estimating the crash potential of components of the traffic system. The model is developed using Newtonian Mechanics and Generalised Linear Regression. The factors contributing to the speed change (ΔV(s)) of a subject vehicle are identified using the law of conservation of momentum. A Log-Gamma regression model is fitted to measure speed change (ΔV(s)) of the subject vehicle based on the identified crash characteristics. The kinetic energy applied to the subject vehicle is calculated by the model, which in turn uses a Log-Gamma Regression Model to estimate the Injury Severity Score of the crash from the calculated kinetic energy, crash impact type, presence of airbag and/or seat belt and occupant age.