Ricerc@Sapienza

This research is mainly aimed at the mechanical study of masonry constructions of historical and archaeological interest, with particular regard to the assessment and adjustment with respect to the seismic actions, taking into account the peculiar non-linear behaviour of masonry.

Several existing researches are oriented to model the constitutive aspects of the masonry materials as well as to the analysis of masonry structures. Among these studies, a remarkable research line focuses on the description of brick/block masonry directly made at the discrete level (Distinct Element Method, Limit Analysis). This approach is suitable for a wide class of ancient masonry, predominant in the Mediterranean historical landscape, built up according to the arrangement of distinct elements of various size and shape, disposed with a periodic rule or in layers of roughly equal height ("regola d'arte"). The several formulations of these approaches, extensively adopted for many years (Nappi,TinLoi 2001; Baggio, Trovalusci 1999, 2016; Orduña, Lourenço 2005; Caporale et al 2006; Milani, Lourenço 2010; Smoljanovic et al 2013; Portioli, Cascini 2016), prove their recognized effectiveness. The discrete approach, however, often turns out to be impractical for computational difficulties due to the large number of degrees of freedom and the running time.
Alternatively, some recent and valuable efforts to obtain continuous models of masonry behaviour have been produced within the context of "multiscale"/"homogenization" theories which, by adopting various equivalence criteria and techniques, allow one to transfer the information of detailed descriptions at finer scales (microscale) to less defined description at the macroscopic scale (Anthoine 1995; Luciano, Sacco 1997; Zucchini, Lourenço 2002; Massart et al 2007, Cecchi, Sab 2009). These models exploit the advantages of the continuous field description preserving a memory not only of the mechanical properties of the constituents, the mortar and the bricks, but also of the shape and arrangement of the bricks themselves. Unfortunately, most of these models are continuous of the classical type (Cauchy) which are inadequate for the study of problems conditioned by the element size, as it occurs in the presence of geometric and load singularities (Trovalusci, Masiani 1999). The main difficulties arise when the characteristic dimension of the structure considered is comparable with the size of the constituents. The problem becomes more complicated when non-linear and non-monotonic stress-strain relationships are taken into account, because ill conditioning in the field equations and numerical solutions dependent on the discretization arise (Sluys-etal93). Non-classical continuous models (Capriz 1989, Eringen 1999), able to retain information of the size of constituents, besides of their shape and texture, circumvent this problem. Among these models, the Micropolar Continuum (Cosserat) holds the field descriptors appropriate to take into account the bricks size and orientation, and it has been proved to be an effective model for treating the problems of masonry mechanics, also within the non-linear constitutive frame (Trovalusci, Masiani 2003; De Bellis, Addessi 2011, Trovalusci, Pau 2015). Another very challenging research trend combines the advantage of discrete and continuum descriptions using "multi-domain" techniques (Greco et al, 2017).
A mention finally goes to the methods directly oriented to the structural analysis which have received, especially in Italy, a particular attention for some years. These studies focuses on the definition of specifically conceived simplified models to adopt case-by-case for which we return to the wide specific literature and to the directives proposed in the relevant legislation mentioned above.

In the case of buildings of significant architectural value, as the designer who needs to sharpen his sensitivity toward the choice of a particular method of intervention, we here propose, as analysts of structures, targeted approaches that often imply the renunciation to the certainties acquired during the last two centuries of the theory of elasticity, in favour of specifically refined mechanical models. Alternatively to the computational instrumentation currently offered by the wide range of commercial software available, more reliable for materials and structures different from masonry, we pay attention to the definition of specific constitutive models to implement and to the development of advanced (`ad hoc') computational codes. To this aim, we also propose to revisit old ideas, which date before the theory of elasticity, which reveals very useful for developing targeted advanced modelling tools oriented to the monument to analyse. Despite of the resulting computational burden, the use of numerical experimentations allows us to effectively treat the various problems of mechanics of ancient masonry without losing sight of their specificity.