Fast masonry collapse and debris analysis using a novel physics engine-powered hybrid numerical framework: first exploratory validations

Discontinuum modelling strategies such as the Distinct Element Method (DEM) are regarded as the leading methods for collapse and debris analysis of unreinforced masonry (URM) structures. Despite their exceptional capabilities for explicit representation of failure phenomena, DEM suffers from impractical computational times and inaccessibility for practicing engineers. An underexplored discontinuum modelling alternative is the use of physics engines, software packages primarily developed for visually plausible, real-time simulations in videogame and animation applications. While physics engines present unmatched numerical speed and stability for simulations comprising of high quantities of dynamic rigid body interactions, they have limited validations for URM structural analysis. This study presents a novel hybrid continuum-discontinuum framework (C2D) for complete collapse analysis of URM structures that leverages the computational capabilities of the Bullet Physics engine (operated using the Python module, PyBullet). Firstly, the quasi-static pushover response is modelled using the Finite Element Method (FEM) up to collapse – the point at which blocks are expected to separate and start interacting dynamically (a transition from quasi-static to dynamic response). A rigid registration algorithm is used to “tessellate” the deformed continuum into discrete URM units. The tessellated block displacements and velocities are imported into PyBullet, where the simulation is resumed until complete collapse. The collapse of a URM obelisk is investigated herein as an example application – mechanical outputs are compared with those obtained using DEM. The proposed framework presents a novel numerical strategy for collapse analysis that has the capability to predict debris distribution from a simplified pushover analysis at exceptional computational speeds, e.g. less than 5 minutes for 37 s of dynamic analysis for the case study presented.