Selective maintenance planning of steel truss bridge based on the Markovian approach

Structures of strategic importance, such as bridges, require careful planning in terms of reliability, durability and safety, qualities which must be guaranteed throughout the entire life cycle of the structure. However, due to the ageing of materials and to aggressive environmental actions which cause deterioration, the response of these structures, just like others, changes over time, resulting in a loss of performance. Yet it is important to maintain a satisfactory level of performance in a bridge throughout its service. To ensure such a performance it is important to apply properly planned maintenance strategies. Appropriate mainte-nance strategies require knowledge of the process of deterioration and the consequent damages to be ex-pected in order to schedule proper maintenance procedures. It would be fundamental to define a selective maintenance plan which may involve only some parts of the structure, thus allowing bridge viability even during the maintenance activity. This paper proposes the study of strategies of selective maintenance for a steel bridge immersed in an aggressive environment, starting from the simulation of each individual member. Simulation of deterioration is obtained through the application of an appropriate damage law implemented with a Monte Carlo methodology, while the time prediction of occurrence of the deterioration is obtained through the application of a Markovian probabilistic approach. The results of the Markovian approach were the starting point for choosing strategies of selective maintenance, as the Markov process allowed the identification, in probabilistic terms, of the structure members with the highest risk of collapse and the timing for achieving levels of damage related to the possible collapse of compromised members. This timing was used to identify possible intervals of maintenance. Proposed scenarios are compared with each other both in terms of associated risk, and in terms of life-cycle cost effectiveness.