Keynote Lecture
| Keynote Lecture 1 | Challenges in Determining Rock Mass Properties for Reliability-Based Design |
| Date / Time | 12 December 2019, Thursday / 09:40 - 10:10 hrs |
| Venue | Room IB-101 |
| Speaker | John Harrison University of Toronto, Canada |
Biography
John Harrison’s research and professional interests lie principally in the field of rock engineering design, most recently contributions to the development of limit state design principles for engineering in fractured rock masses. His most recent research has concentrated on Bayesian data analysis applied to the assessment of rock strength and in situ stress, development of tensorial statistics for quantifying variability of in situ stress, and characterisation of uncertainty models for rock engineering. From 2011 to 2018 he was convenor of a specialist pan-European rock mechanics and rock engineering expert group working under the auspices of CEN to revise Eurocode 7, and currently chairs the ISRM Commission on the Evolution of Eurocode 7. He holds the W.M. Keck Chair of Engineering Rock Mechanics at the University of Toronto, Canada, a position he has had since 2010.
Abstract
In the context of RBD for rock engineering, the principal source of the structural resistance is provided by the rock mass, and in accordance with the principles of RBD this needs to be characterised statistically, on the basis of aleatory variability. Only data that have been obtained by objective quantitative measurement can be used for this, and so rock mass properties currently can only be determined by a synthesis of component properties. Variability appears to exist in all component geometrical and mechanical properties, and the similarity with in and between rock types is so slight that reference values are unsuitable, indicating that variability will need to be determined project-by-project. A major challenge in this work is the perennial problem of limited data, but Bayesian approaches may ameliorate this. Modern investigation techniques allow large quantities of discontinuity geometry data to be obtained, and our understanding of discontinuity geometry is advanced; together these bode well for application of RBD. The use of numerical modelling to combine component properties and obtain large scale rock mass properties is proving efficacious, and has confirmed stress and scale dependent variability. Significant variability at the engineering scale has been revealed, which renders large-scale in situ testing unfeasible for characterising rock mass variability. Overall, significant challenges remain in determining the behaviour of rock masses for use in RBD. Appropriate techniques that will allow determination of the necessary properties have largely been developed, but much work remains to be done before routine and simplified design methods appear.
Full Text (PDF)