Two different stochastic approaches modelling the decay process of masonry surfaces over time

Masonry structures subjected to aggressive environment may suffer degradation during their service life; the decay strongly depends on type of the component materials and the technique of construction. This can lead the structure to high states of damage even if not failure. The great randomness connected with the occurrence of critical attacks suggests approaching the deterioration process of masonry under a probabilistic point of view. Following this way, the deterioration process L(t) of stone masonries has been carried out. It has been approached as: 1) L(t) time dependent stochastic process of the random variable (r.v.) l. 2) L(t) time dependent stochastic process of the r.v. l. The approaches 1) and 2) are able to model the reliability of masonry materials over time and predict, in probabilistic terms, the occurrence time of the expected damage. The procedures have been applied to full-scale models built in Milan in 1990. To measure the masonry decay in time a non-destructive technique has been adopted and the data collected have been elaborated using the approaches 1) and 2). The results obtained allow the good convergence of both the procedures as well as their different possible applications. On the base of these results a discussion on the possible use of these procedures in the maintenance strategies planning is introduced.