Predicting and optimizing lifetime performance of existing steel bridges under ...

• Autor(es): Neves, Luís A. C. ; Frangopol, Dan M. ; Cruz, Paulo J. S.
• Data: 2005
• Identificador Persistente: http://hdl.handle.net/1822/5001
• Origem: RepositóriUM - Universidade do Minho
• Assunto(s): Deterioration; Existing structures; Genetic algorithms; Maintenance; Management; Optimization; Preventive maintenance
• Tipo de Documento: Documento de conferência
• Idioma: Inglês

Descrição

In this paper, the lifetime performance of existing steel bridges under preventive and essential maintenance actions is analyzed. Performance is measured by the reliability index (resulting from structural analysis) and the condition index (resulting from visual inspections). The analysis of performance using these two indicators allows the incorporation of the results of visual inspections in the assessment of the safety of a structure, resulting in a more accurate prediction of the need to maintenance. Due to the uncertainty in the deterioration process and time of application and effect of maintenance actions, a probabilistic framework is used. The reduction in structural performance due to corrosion of the steel profiles is simulated using an extension of the model proposed in 1998 by Frangopol [1], considering all parameters defining the profiles the condition and reliability profiles as random variables. The effect of the maintenance actions on the condition and reliability profiles is defined in terms of the improvement in performance and the reduction or suppression of deterioration for a period time after application of maintenance. Monte-Carlo simulation is used to compute the probabilistic properties of the condition, reliability, and life-cycle cost profiles associated with each maintenance strategy. The model described is applied to steel bridges located in Colorado. The results show the importance, in terms of life-cycle cost, of considering both preventive and essential maintenance in a rational lifetime maintenance strategy.