R.C.DE VEKEY, Building Research Establishment
PROFESSOR A. W. PAGE University of Newcastle, NSW
and E.G.HEDSTROM National Concrete Masonry Association, USA

The bond wrench has been in use for several years for laboratory research on bond, and more recently, as a quality control tool for newly built masonry and for in-situ measurement of bond on existing structures. Despite the great versatility of the technique one problem holding back a greater acceptance has been the relatively high variability of bond wrench data. In this paper well over 2000 commercially sponsored bond measurements from both new build and investigations of extant structures have been examined statistically. It is concluded (i) that the overall variability for well bonded new masonry is no worse than laboratory measurements using other less-versatile test methods, (ii) that there are no problems of transporting small stack specimens for laboratory tests and (iii) that the result is not sensitive to the position of the joint in a stack prism. In-situ measurements are not necessarily any more variable for well bonded masonry but old weathered poorly bonded masonry does give high variability partly due to inherent variations and partly a consequence of the low mean values.