The results of 48 component T-stub tests and 10 component clip angle tests are presented and evaluated based on their strength, stiffness and ductility characteristics. The results are used to compare several existing strength models that govern the failure modes of tension bolt fracture including prying forces, net section stem fracture, shear bolt fracture and block shear failure.
A monotonic stiffness model is developed based on the experimental observations and the results of an advanced finite element investigation. The stiffness model which is based on a component spring model similar to that used in the Eurocode, includes stiffness contributions from a flange and tension bolt mechanism, stem deformation, and slip and bearing deformation. The flange/tension bolt stiffness model includes a variable tension bolt stiffness and various stiffnesses of the flange as it passes from a totally elastic state to a plastic mechanism. The slip and bearing deformation model includes the effects of initial shear bolt alignment and lack of fit. Because the stiffness model is based on rational mechanisms, it is able to predict the deformation at failure of a T-stub with reasonable accuracy.
A cyclic model is also developed based on observations made during the experimental program. Th e model, which incorporates a decaying slip resistance, is compared to results of six beam-column tests that were conducted as part of a separate investigation.
Finally, modifications to existing strength models are recommended and a deterministic T-stub design procedure is presented that yields ductile beam-column connections.
It was found that (1) the most desirable behavior was obtained from T-stubs that were proportioned such that a flange mechanism and stem yielding developed simultaneously, (2) little difference in overall behavior was noted between components using A325 and A490 bolts, (3) a modified version of the strength model proposed by Kulak, Fisher, and