On accident analysis and slip-resistance measurement

Abstract

Four accident analysis models are discussed in relation to a specimen slipping accident. One classification, or ‘single-type’ model, commonly used in accident statistics, lumps all events and agencies from each accident in one ‘cause’ group. This may lead to serious underestimates of particular hazards. In past occupational injury statistics for Sweden, slipping appeared in less than 5 % of all accidents during 1975. With an entirely different sequential model of the present statistics system (‘ISA’, recording about 100 descriptors for each of more than 100000 annual accidents), slipping has been registered in more than 77% of all accidents during 1979. Fault-trees and dynamic work analysis models are considered for case studies and hazard descriptions.

Having identified a major hazard, appropriate techniques and criteria must be developed for safety performance tests. Reports of more than 60 different slip-resistance meters have been found in the literature. Many of the meters are designed to gauge the static friction coefficient, even though tribology research concludes that any frictional force requires motion, at least between rubberlike materials. Thus, the actual forces and motions from human slip-ups must be known, if real slip-resistance is to be measured. Such biomechanical data have been recorded in the laboratory with a computerized gait analysis system. Practical conclusions are presented, based on more than 100 experiments with subjects walking on the level.