Probabilistic modelling of the damage induced by salt crystallization in fiber reinforced clay brick masonry

Externally Bonded Reinforcement (EBR) systems applied with Fiber Reinforced Cementitious Matrix (FRCM) are increasingly used to strengthen existing masonry constructions. They consist of dry fibers composing open meshes embedded in layers of inorganic-based matrix. Composite systems are preferred to other more conventional systems like the reinforced concrete (r.c.) or steel, since they are light weight and able to adapt themselves to every shape. Moreover, with the use of properly selected matrices, FRCMs represent an alternative to Fiber Reinforced Polymer (FRP) systems, in terms of higher compatibility and removability. Masonry buildings are also affected by time-dependent attacks, particularly severe in case of Cultural Heritage constructions, which may weaken the durability of the masonry, Salt crystallization is one of the highest causes of damage in masonry connected with an aggressive environment: mortar joints as brick and stone can be a vehicle of the water and salts coming from capillary rise or sea spray. This phenomenon can also affect FRCM applied as EBR to masonry. To investigate durability against sulphates, salt crystallization tests were carried out according to a RILEM pre-standard procedure on small masonry assemblages reinforced with FRCM, to investigate . The specimen includes both soldier course and running bond configurations, to simulate various possible applications in masonry components of buildings. The loss of surface material measured was assumed as parameter of damage so as the bulging under the fiber. The loss is referred to the vertical section of the specimen and quantified through a computer code and successive readings of the surface decay using a laser profilometer. The randomness affecting this kind of damage and the consequent loss of material under fibers suggests studying the deterioration process under a probabilistic view, where the continuous deterioration of specimens can be assumed to be a stochastic process.