Experimental analysis and artificial neural networks modelling of retarder effects in motor vehicles

Foundation brakes of heavy-duty vehicles while descending downhill and applying the brakes continuously to control the vehicle are subject to coercion. In this case, temperature of the friction materials in brake system significantly increases. This causes to decrease friction coefficient of disc-pad (or drum-pad). One of the reasons of heavy vehicle accidents is inefficiency of the brakes known as ?fading?. In order to keep the servis brake from overheating and solve fading problem an auxiliary system known as retarder can be used. In this experimental study, functions of an electromagnetic retarder used in transportation vehicles and its effects on a vehicle behaviour and brake efficiency were investigated. Experimental aspect of the study has been divided into two sections. In the first section; stopping distance tests were carried out to evaluate the break efficiency of the same vehicle at different brake disc temperatures. In the modeling section of the study, Artificial Neural Networks (ANN) method was used. Using the data obtained from road test results, 8 vii different high performance ANN models which predict parameters like disc temperature, vehicle mass, vehicle velocity, retarder torque, stopping distance and pedal force have been develeoped. Test results show that the vehicles without retarder have the risk of accident and brake fading. The high performance models developed can also be use to predict vehicle accident reconstruction. Key words : Heavy-duty vehicle accidents, retarders, road tests, artificial neural networks