Search

The failure of several 15-m wood trusses at the lower chord splice joint is described. These trusses were built in the early 1940s with 100-mm split-ring connectors at the splice. The general failure pattern was initially horizontal splitting at the end of the chord member in two places through the depth of the member. As the split progressed, the center plug would move out from the end of the member, with lateral movement as great as 15 mm. Ultimately, a tension failure occurred across the grain. The exact origin of the truss failures is unknown. The trusses were appropriately designed and had enough reserve capacity in the original design to compensate for the decrease in allowable stresses that has occurred since construction, including the use of No. 2 lumber instead of the specified select structural lumber. The best conjecture of the cause of failure would be that it was initiated by secondary stresses. The continuity of the bottom chord over several panels induced secondary tensile stresses perpendicular to the grain at the splice. Additional tensile stresses perpendicular to the grain from drying shrinkage restrained by the split rings contributed to the initial horizontal splitting. Contributing to the development of the failure were probably the load-duration effects and perhaps higher temperatures. Despite the distress observed, there is no immediate danger of collapse. Lateral in-plane restraint at the truss ends from columns and adjacent portions of the building enable an arching mechanism to form that gives stability to the trusses even in a distressed state.