Fracture moment and ductility of welded connections

Abstract

Examination of Northridge fractured connections has revealed that the materials and the manufacturing parameters associated with the fracture behaviour of these connections are random. Hence, the fracture resistance and ductility of these connections are also random. In this study, most of the parameters which are believed to contribute to the fracture behaviour of Northridge connections are analysed with respect to previous research in the field. Then one of the pre-tested post-Northridge connections is modelled, using finite-element analysis and the results are used as typical examples to illustrate the effect of each parameter on the behaviour of these connections. Experimental fracture moment and ductility data of a set of previously conducted tests on the pre- and post-Northridge design are used to obtain suitable cumulative distribution functions of the normalised fracture moment and plastic rotation of these types of connections. Statistical results show that, despite all the modifications applied to the pre-Northridge connections, the magnitude of the minimum plastic rotation and the strength are still less than what is required by the seismic codes. Therefore, both pre- and post-Northridge connections do not have enough strength and ductility and consequently they are not reliable and should be further modified when used in seismic regions.