Effect of joint deformability on the experimental and numerical response of dry-joint masonry arches subjected to large support displacements

In the literature, rigid no-tension models were generally observed to not perfectly predict the response of dry-joint masonry arches to large support displacements and, in particular, to overestimate the ultimate displacement capacity obtained in experimental tests. This work investigated the reasons for these discrepancies by analysing the case of a small-scale segmental dry-joint masonry arch, which was tested by the authors under vertical, horizontal, and inclined support displacements. To simulate the experimental tests, a finite element micro-modelling approach was adopted, in which the arch was schematized as an assemblage of very stiff voussoirs interacting at no-tension friction interfaces. The comparison between the numerical and experimental results proved that the discrepancies between the predictions from a rigid no-tension model and the experimental outcomes was due to the imperfections and resulting deformability of the joints of the physical model. A strategy to account for such deformability in the adopted modelling approach was thus proposed and successfully validated by comparison with the experimental results.