On the failure modes and strength of steel-wood-steel bolted timber connections loaded parallel-to-grain

Citation
Jhp. Quenneville et M. Mohammad, On the failure modes and strength of steel-wood-steel bolted timber connections loaded parallel-to-grain, CAN J CIV E, 27(4), 2000, pp. 761-773
Citations number
16
Categorie Soggetti
Civil Engineering
Journal title
CANADIAN JOURNAL OF CIVIL ENGINEERING
ISSN journal
03151468 → ACNP
Volume
27
Issue
4
Year of publication
2000
Pages
761 - 773
Database
ISI
SICI code
0315-1468(200008)27:4<761:OTFMAS>2.0.ZU;2-M
Abstract
The current Canadian code provisions for the design of timber bolted connec tions were essentially developed based on connections showing a ductile beh avior and then further modified to account for situations where connections fail in a brittle way. An experimental study was undertaken to evaluate th e strength of bolted connections specifically experiencing a brittle mode o f failure. Specimens consisting of steel-wood-steel connections with either 19.1 mm or 12.7 mm bolts were tested in tension. Test variables included e nd distance, bolt spacing, row spacing, number of bolts per row, number of rows, thickness and species of wood member, glulam or sawn lumber members. Connections were tested to the ultimate to observe possible modes of failur e as variables were changed. Results show that the current Canadian standar d approach to evaluate the resistance of timber bolted connections is not o ptimal although conservative. Brittle modes of failure such as row shear-ou t, group tear-out, and splitting were observed. The resistances calculated using the O86.1 design provisions are as little as a third as compared to t ested results. Also, the design equations do not allow the designer to take advantage of the increases in strength as a result of increases in row spa cing, as observed in tests. Analysis of the results show that the longitudi nal shear stress at failure is related to a parameter which is a function o f the smaller distance (end distance or bolt spacing) and the specimen thic kness. This relation was used to formulate design equations to predict the row shear-out and group tear-out strengths of glulam specimens using the sp ecified strength values listed in O86.1. As well, it was found that Mode I of the European yield model is the only ductile ultimate failure mode and t hat other equations for bearing failure can be neglected. In this paper, th e research program is described, results are presented, and an alternate de sign approach is proposed to predict the failure mode and the ultimate stre ngth of steel-wood-steel bolted connection groups.