Laboratory investigation of pull-out capacity of chemical anchors in individual new and vintage masonry units under quasi-static, cyclic and impact load

Abstract Failure of unreinforced masonry following the 2011 Christchurch earthquake demonstrated that many masonry strengthening solutions were inadequate for the peak ground accelerations that were experienced, and also that many of the failures were associated with underperformance of masonry anchors. Recent in-situ pull-out tests of anchors in vintage masonry structures has identified that in these tests, the failure is predominantly via splitting of the masonry units. This finding is in contrast to current design approaches that only consider failure via the formation of a cone or wedge, or masonry unit extraction. To further examine the potential for unit splitting prior to the failure modes identified in current design approaches, a laboratory campaign investigating masonry unit properties and anchor pull-out capacities, covering quasi-static, cyclic and impact loading of anchors and also incorporating the influence of quality of installation is reported here. The results of this campaign confirm that the unit splitting is an important failure mode, which may explain the observed anchorage underperformance. It is further observed that whilst cyclic and impact loading, as well as poor quality of installation have a detrimental effect on anchor performance, performance nonetheless exceeds published characteristic strengths. Additionally, as part of the laboratory campaign, a simple method for supporting masonry test units has been developed which has demonstrated good replication of the in-situ test results.