LAUSR

Abstract Investigations of damage sustained following the 2010–2011 Canterbury earthquake sequence highlighted premature failure of anchorages in previously strengthened masonry. These failures suggest that a lack of understanding surrounding anchorage design is limiting the ability to seismically retrofit masonry structures. In order to provide experimental observations for use in developing and calibrating anchorage models a series of in-situ tests have been undertaken on three masonry structures to quantify in-situ mechanical properties and corresponding anchor pull-out loads and failure modes. Importantly, the outcomes of this field testing show poor correlation with the outcomes predicted by current design standards and approaches – that is, that the failure of anchorages by splitting of the masonry units rather than cone/wedge type failure or masonry unit extraction was the predominant observation. Additionally, due to difficulties with common test procedures, interpretation of standardised tests such as the “shove” test and the bond wrench test have been difficult. Alternative material tests and statistical distributions are proposed and a new anchorage failure model is suggested.