Influence of imperfections on the load capacity and stiffness of glued-in rod connections

Glued-in rods (GiR), which consist of rods glued into structural elements of wood, or engineered wood products, share the manufacturing quality concerns associated to adhesively bonded joints in general. Because of the difficulties associated with the bonding process, it is often argued within the timber engineering community that under the specific conditions prevailing in construction sites, it is not possible to meet the high-quality requirements for bonded joints typically encountered in other industries. The present study presents experimental evidence gathered on 240 individual probes in which the effect of several different types of defects (excessive wood moisture content, wood chips not removed from the borehole, simulated air inclusions by mixing polystyrene spheres, oil contaminated rods, and corroded rod) was investigated with threaded rods glued into beech spruce glued laminated timber (GLT) and beech laminated veneer lumber (LVL) using four adhesives (including 2K-epoxies and a 2K-polyurethane). Results were analysed by means of an analysis of variance (ANOVA). The findings showed that glued-in rods represent a relatively robust class of adhesively bonded joints, as the majority of defects types could not be distinguished on statistical ground from the reference sets, with the notable exception of air voids leading to a significant decrease of joint capacity.