Bolted connections between open section beams and box columns

This thesis reports on the full scale testing of jOints comprised of open section beams, bolted to square closed formed columns (SHS) using the novel Flowdrill blind bolting connector. Twenty-six joints were constructed using Flowdrill connectors. In addition six stand alone tests were conducted on endplates. Three joint tests were also conducted using a rival Hollo-bolt connector, in which the joint geometry was nominally identical to those of the Flowdrill joints. Thus a total of thirty-five cantilevered joint tests of member configurations typically found in building frames were conducted. Only welded end plates were examined in the programme of joint tests which included partial depth (flexible), flush and extended endplate details. These particular details enabled the full range of both connection stiffness and strength in typical simple and rigid construction to be examined. The aim of the project was to investigate the joint's semi-rigid behaviour, and the acceptability of the flowdrill connector as a method to site bolt the beam to the column rather than using the welded solution commonly associated with closed formed columns. From the experimental moment-rotation data collected during the tests, a proposed joint model for both the flush and extended end plate details was developed to predict its full non linear behaviour. The model provides guidance on the joint's rotation limit and ultimate design capacity. It assumes that the endplate of the joint is rigid and provides only a limited contribution to the joints overall rotation for its effect to be ignored in the model; the principal failure of all the joints being severe column face deformation. Although the capacity of the connector was not in the original aim of the investigation, an assessment on the published capacities has been conducted for completeness. As well as providing design guidance for these type of joints, the thesis also presents the results of a parametric study into sub-frame behaviour using the actual joint responses from the test data. The findings suggest that the restraint afforded by the connection on stocky columns in sub-frame arrangements do not outweigh the disadvantages of the moment transfer. A further study on the economics of using tubular columns has also been included. This shows only a marginal increase in cost to that using equivalent open sections, when the advantages of reduced fire protection are included in the pricing.