Failure In-the-Row Model for Bolted Timber Connections

Abstract

Forces in heavy timber members are often transferred at connections into steel gusset plates via laterally loaded bolts, which are arranged in rows parallel to member axes. Elastic analysis predicts an unequal sharing of force between bolts, with the first or last bolt within a row transferring the highest fraction of the member force. The capacity of a connection is taken to be inversely proportional to the force carried by the most heavily loaded bolt. Experiments confirm that there is a reduced effective capacity per bolt with any increase in the number that are placed in a row. This so-called “row effect on strength,” however, is often much less stringent than elastic analysis suggests. The discrepancy is because individual bolts within a row exhibit some ductility in their response, prior to global failure of a connection. Global failure of a connection can appear to be either brittle or ductile, depending upon designer-selected parameters such as bolt slenderness, number of bolts, and bolt spacing. It is demonstrated here that, by allowing for a “limited amount” of nonlinear deformation at bolts, accurate predictions can be made of the row effect on the strength of multiple-bolt connections. A hybrid elasto-plastic analysis reliably predicts the capacity for a row of bolts and whether the global failure at a connection will appear to be brittle or ductile.