CUMULATIVE DAMAGE IN COMPONENTS OF STEEL STRUCTURES UNDER CYCLIC INELASTIC LOADING

The objective of the work presented in this dissertation is to quantify cumulative damage in structural components subjected to cyclic inelastic loading histories. The cumulative damage models developed in this research can serve as a basis for assessing the safety against failure of steel structures in severe earthquakes.; As part of this study two series of experiments were performed in order to evaluate damage accumulation and resistance deterioration due to local buckling in beam flanges and crack propagation at weldments. The experiments have shown that the response of components can be separated into a deterioration threshold phase and a deterioration phase. In the local buckling mode, the deterioration threshold is relatively small but deterioration occurs at a slow rate, indicating that some deterioration should be accepted in the failure definition. In the crack propagation mode, localized crack growth does not cause noticeable deterioration of the component strength and stiffness until the crack has grown considerably. Once deterioration takes place, it occurs at a very fast rate leading soon to failure. Thus it appears to be justified to ignore deterioration in this case and to associate failure with the onset of noticeable deterioration.; Cumulative damage models are developed for the deterioration threshold range as well as the deterioration range, using constant amplitude tests to quantify model parameters. These tests have shown that the number of cycles to failure for constant amplitude cycling can be described by a Coffin-Manson fatigue model. Utilizing the linear damage accumulation hypothesis, suitable damage models are derived for arbitrary loading histories. Adequate predictions of the experimental response have been obtained using these models.; As an alternative to the low-cycle fatigue approach, plastic fracture mechanics concepts can be employed to predict crack propagation and fracture at weldments. A crack growth rate model together with an initial crack size are used for this purpose. The life predictions based on this fracture mechanics approach are close to the predictions obtained using a Coffin-Manson model together with the linear damage hypothesis.; The results of this study show that simple models can be developed for damage assessment and resistance deterioration in components of steel structures. These models can be used for performance evaluation of deteriorating structures subjected to severe earthquake excitations.