Cross-laminated timber panel buildings are gaining a growing interest of the scientific community due to significant technical advantages, such as the material sustainability, the high fire resistance and quick constructability. Nevertheless, it is well known that timber panels themselves are not able to dissipate a significant amount of energy during an earthquake. In fact, the design of a Cross-Lam building is carried out in order to dissipate the energy in the steel connections (hold-downs or angle brackets) which govern the seismic performance. The paper here presented proposes to substitute the classical hold-downs, which usually exhibit a limited dissipation capacity, with an innovative type of dissipative angle. The new connection, called XL-stubs, apply the concept usually adopted for designing the hysteretic metallic dampers ADAS (Added Damping and Stiffness). In particular, the hourglass shape allows a better spread of plasticization resulting in a high dissipation capacity. In order to characterize the force-displacement response under cyclic loads of XL-stubs an experimental campaign is carried out comparing the hysteretic behavior of the classical hold-down with that of the proposed dissipative angle.

# seismic design

## How code-based linear static analysis for the seismic design of masonry buildings may fail to be conservative

Linear static analysis is the easiest method for the seismic design of unreinforced masonry buildings. What makes this method particularly convenient to apply is that the fundamental vibration period of the structure may be obtained conventionally from a formula given by the codes of practice, rather than being calculated rigorously. Usually, the code-based formula overestimates – sometimes significantly – the rigorous value of the fundamental period of vibration of masonry structures. By referring to a series of unreinforced masonry buildings, the paper shows that the conventional period may be even 4-9 times greater than the rigorous one. However, overestimating the fundamental period may lead to a reduction of the design loads to be applied in the static method. This occurs, in particular, when the design response spectrum decreases in the short-period range, where natural periods of masonry buildings typically fall. Some codes do not allow design spectra to be negatively sloped in the short-period range. While others, such as the Italian code of practice, do allow it, and in so doing, they make the linear static method non-conservative (in the given examples the shear at the base of masonry buildings may be underestimated by even as much as 35% when compared to the values obtained from the more precise linear dynamic method). The paper gives some hints towards making the code-based static method for seismic design of masonry buildings safe.