桥(图论)
地铁列车时刻表
可靠性工程
概率逻辑
计算机科学
桥梁维护
预防性维护
工程类
运筹学
甲板
结构工程
人工智能
医学
内科学
操作系统
作者
F. B. P. de Jong,P D Boersma
摘要
This paper consists of two parts. The first part describes a new kind of maintenance philosophy. Conventional maintenance strategies are formed by time based inspection intervals and replacement of parts of the structure at the end of their lifetime. This strategy and the insufficiency of this strategy to handle fatigue problems are described. Subsequently the properties of a new risk based maintenance strategy, based on probabilistic calculations models, are described. Successful implementation of this strategy is only possible if the necessary tools in this strategy are available. The paper describes the tools in the strategy. These are lifetime calculation models, local reparation techniques, renovation techniques for complete bridge decks and inspection techniques. This strategy and the developed tools have resulted in practical solutions for the 80 fixed and movable bridges in the Netherlands, which enables the bridge owners to maintain their bridges, minimizing the risks and maintenance costs. The second part of the paper describes more in detail the lifetime calculation model, which is of vital importance to minimize risks and to schedule inspections and renovations. Reliable lifetime calculations need accurate models of the number and amplitude of the stress cycles and the fatigue behaviour. A formal model to calculate the fatigue crack growth of deck plate cracks is presented. Based on this formal model a computer program has been developed for bridges in the Netherlands. Aspects related to the number and amplitude of the cycles are investigated. These are traffic properties, axle load distributions, the fatigue classification and the calculation of the stresses at the crack location. Special attention has been paid to the effect of the asphalt wearing courses, which are applied at fixed bridges. Asphalt has a reducing effect on the stress in the steel deck plae , but
科研通智能强力驱动
Strongly Powered by AbleSci AI