Estimation of probabilistic scenario-based design load for extreme events

As design lifetimes being extended much longer than in the past, the importance of extreme events needs to be addressed. This paper presents a probabilistic design framework for evaluating very rare human-made hazards such as collisions and explosions. The proposed design against artificial hazards involves the development of an integrated design framework that is intended to simultaneously deal with the occurrence of an accidental event and the resulting resultant structural load using probabilistic scenariobased simulation. The proposed procedure consists of three steps. In the first step, event scenarios are identified and the probability of each occurrence is determined. The second step involves the construction of the conditional Complementary Cumulative Distribution Function (CCDF) of the load on those events. Finally, using the total probability theorem, the CCDF of the load during a period encompasses by an accidental events is calculated, and the design load is determined by a specified criterion which is described as the exceedance probability or the recurrence period of an event. As a specific example of the proposed general framework, detailed procedures for determining the design load in the case of a ship-bridge collision design, with an application to a cable-stayed bridge, are presented. Analysis results show that the proposed method effectively provides reasonable design loads, considering the many uncertainties inherent in the nature of a vessel collision hazard.