Abstract

Earthquake induced steel structural failures are prevalent due to the high lateral forces applied to the structures resulting in excessive sway and more importantly ultra-low cycle fatigue of structural members. Even though lateral load resisting systems such as concentrically braced frames are installed to the moment resisting frames, ultra-low cycle fatigue induced brace failures play a dominant role in the context of local buckling of the members due to the amplified strains at the vicinity of the brace mid length followed by high strain concentrations. However, the concentric braced frames are designed and detailed deliberately to buckle out of plane in order dissipate the input seismic energy given to the structure by an earthquake. This global buckling of the braces is favorable to a structure in terms of the hysteretic response where the local buckling of the braces drives instant failure of the brace due to ultra-low cycle fatigue fracture. Cyclic void growth models are used to predict this fracture initiation due to ultra-low cycle fatigue in small scale steel components where the method requires a detailed stress and strain history obtained from a meticulous finite element model. If these models are to apply to a case where a multi storey building with braced frames in which an earthquake is applied, analysis of the finite element model to obtained the stress and strain histories until the failure, is quite impossible due to the high computational expense. In this regard, the present study first proposes a new method to model the structural behavior of concentric braces under cyclic loading using fiber beam column elements used in the OpenSees framework and then a new modified cyclic void growth model is proposed. To do that, experimental tests of concentric braces were modeled using OpenSees to observe the applicability and the limitations of fiber beam column element in terms of hysteretic response and strain variation in predicting the ultra-low cycle fatigue fracture. These tests were modeled using OpenSees in such a way, the numerical models fail at the same circumstances as the experimental test and thus validated the hysteretic response asserting the good prediction of global behavior by the numerical model. Penultimately, the resultant stress, strain histories having very low values due to fiber beam column element’s inability of predicting local buckling were used to calibrate a cyclic void growth parameter called modified damageability index. Finally, all the calibrated modified damageability values (λ′) for several experiments were initially found and in future studies, a regression analysis to build a relationship for λ′ in terms of brace geometric parameters such as slenderness ratio and width to thickness ratio to come up with a virgin equation in supporting the new simplified cyclic void growth model will be formulated.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call