A peridynamic investigation on crushing of sand particles. Issue 6 (13th August 2018)

• Record Type: Journal Article
• Authors: Zhu, F.; Zhao, J.
• Is Part Of: Géotechnique. Volume 69:Issue 6(2019)
• Page Start: 526
• Page End: 540
• Publication Date: 2018-08-13
• Subjects: failure -- numerical modelling -- particle crushing/crushability -- particle-scale behaviour -- sands; Soil mechanics -- Periodicals; Engineering geology -- Periodicals; Geotechnical engineering -- Periodicals; 624.151
• DOI: 10.1680/jgeot.17.P.274
• Languages: English
• ISSNs: 0016-8505

Abstract

Abstract : Particle crushing underpins important macroscopic behaviour of granular materials such as yielding, deformation, dilatancy, failure, mobility and packing features. The crushing condition and crushing pattern have commonly been examined for particles subjected to uniaxial loadings. In the real engineering context, a sand grain is typically in contact with several surrounding particles and is hence subjected to multi-directional loadings, a critical condition that has not been well accounted for in most crushing criteria and studies of crushing patterns relevant to discrete-based sand modelling. In this study, the crushing of single sand particles under different loading conditions is examined based on peridynamic simulations. The peridynamic method is found capable of realistically capturing the crushing of a sand particle under uniaxial loadings in terms of crushing load and the crushing pattern observed in experiments, and is able to simulate multi-contact particle crushing where experimental data are relatively scarce. By examining existing crushing criteria, it is found that the numerical results on the crushing load under multiple contacts compare favourably with the maximum contact force criterion, which states that particle crushing occurs when the maximum contact force reaches a threshold. It is observed that the number of child particles after the crushing of a sand particle bears no apparent correlation with the coordination number. The volumes of child particles can be statistically described by a normal or gamma distribution. The findings from the study offer insights into the behaviour of sand particle crushing, which can be useful for future discrete modelling of granular sand where crushing is important.