Fatigue damage analysis of steel components subjected to earthquake loadings

Examination of steel moment resisting frames after the 1994 Northridge earthquake showed fatigue cracks presented in the beam–column connections of the frames. These observations indicate that fatigue failure may occur in the steel components of building structures in an earthquake event. To apply the fatigue design approach using the Palmgren–Miner’s rule for steel components of the moment resisting frames requires the knowledge regarding the damage index value at fatigue failure. The purpose of this paper is to perform fatigue tests to give the first damage values of steel components subjected to real earthquake-induced loadings.

The added-damping-and-stiffness steel plates which are used in building structures for earthquake mitigation were fabricated and tested by constant amplitude, SAC block and earthquake-induced loadings to failure. The earthquake loadings were obtained from the dynamic analysis of a steel frame with the mentioned plates. The load cycles of the SAC block and the calculated earthquake loadings were counted using the rainflow-counting method, and the damage index value of each specimen were calculated using the Palmgren–Miner’s rule.

Reverse stiffness obtained from cyclic load-displacement loops is a robust and consistent parameter that can be used for determining fatigue failure of tested components. The Palmgren–Miner’s damage values at failure, caused by earthquake loadings, are smaller than 1, and in addition, are also smaller than those obtained from the tests of the SAC block loading. The large-amplitude cycles in the earthquake loading produce large damage on the specimens, and intermediate range cycles also produce damage that should not be neglected in the fatigue analysis.

Today’s building design code allows large plastic deformation to occur in steel frames during an earthquake. However, the pre-Northridge earthquake steel frames showed fatigue cracks without the expected substantial plastic deformation at beam flanges. Proposed solutions to this problem were the reduced beam section neglecting the existence of the cracks at beam–column connections. This study considered the fatigue phenomenon in steel frames and provided the first set of tested fatigue damage values for steel components subjected to realistic earthquake loadings, which offered a possible method of dealing with fatigue cracks in the steel components of a building structure.