LI, JIE1; STEWART, MARK G2; MASIA, MARK J3; LAWRENCE, STEPHEN J4

1) PhD Student, Centre for Infrastructure Performance and Reliability, The University of Newcastle, NSW, 2308, Australia c3124120@uon.edu.au (corresponding author)

2) Professor and Director, Centre for Infrastructure Performance and Reliability, The University of Newcastle, NSW, 2308, Australia Mark.Stewart@newcastle.edu.au

3) Associate Professor, Centre for Infrastructure Performance and Reliability, The University of Newcastle, NSW, 2308, Australia Mark.Masia@newcastle.edu.au

4) Conjoint Professor, Centre for Infrastructure Performance and Reliability, The University of Newcastle, NSW, 2308, Australia spl@bigpond.net.au

It is well known that the material properties of unreinforced masonry (URM) such as the flexural tensile strength of mortar joints and bricks and torsional shear bond strengths of the mortar joints, to name a few, vary considerably from joint to joint and brick to brick, even within the same wall. Furthermore, it is well known that this spatial variability might significantly affect the structural performance and reliability of URM walls. The paper develops a computational method to predict the strength for URM walls subject to one-way horizontal bending considering unit-to-unit spatial variability of mortar joints and bricks. In this context, the term “unit” is being used to describe the location in the wall associated with a single brick and the adjacent mortar joints. In this way, the material properties are assumed to be uniform along the length, height and thickness of individual bricks but may vary from brick to brick within the wall. Tensile strength, shear bond strength and associated fracture energies of the mortar joints, and tensile strength and fracture energy of the bricks are the main parameters considered herein. We examine how spatial variability in unit strengths (mortar joints and bricks) affect the variability of ultimate strength and damage progression of clay brick URM walls in one-way horizontal bending. Stochastic analysis in the form of Monte Carlo simulations used a 3-D nonlinear Finite Element Analysis. The results were validated from a database of available experimental results on masonry four course beams. It was found that good agreement of peak load exists between the stochastic simulation and the experimental results for the four course beam subject to horizontal bending.

Keywords: stochastic, masonry, tensile strength, torsional strength, spatial variability, structural reliability