Experimental and numerical investigations on barriers for rockfall hazard mitigation

Managing the rockfall hazard is a complex task, which involves several phases from land planning to the design of protective structures. Despite the increasing interest in rockfall-related research over the last few decades, some issues remain only partially addressed to date. In particular, existing predictive tools for the simulation of block trajectories could be improved and the criteria for the design of flexible metallic barriers are potentially flawed by block size effects. In the Australian context, the lack of research into rockfall phenomena has prevented the development of a comprehensive hazard characterisation so far; moreover, some types of barriers currently in use present cost-effectiveness issues. In this thesis, formed by six peer-reviewed publications, the aforementioned issues are investigated, with the aim of improving the knowledge regarding rockfall, especially for the Australian environment. The first paper presents the results of extensive in situ tests performed in different geological environments in NSW. Rockfall motion parameters are obtained for the first time in the Australian context and uncommonly high values of the normal coefficient of restitution are highlighted. An investigation of these results is undertaken in the second paper through extensive laboratory testing: low impacting angles, rotational energy and block shape are correlated to the high values of kn obtained in the experiments. In the third paper, the results from in situ tests are applied to real profiles taken from a database of Australian slopes: a basic hazard characterisation is performed using a 2D lumped mass model and low impact energy values are found for most of the cases. The fourth paper shows a comparison between four different flexible barriers, carried out by full-scale laboratory testing. The results provide an insight on the estimation of the barriers’ performance: in particular, stiffness and load transmission are evaluated, and modifications to improve the performance of one of the systems are suggested. Experimental and numerical evidence of the bullet effect are presented in the fifth and sixth papers. A Finite Element model of a mesh panel, calibrated by means of laboratory testing, is used to validate an innovative dimensional approach.