## Info

Moment [kNm]

â€”Diagonal strut and king post -Diagonal strut and rafter

Figure 14. Rotational response of the connection between the rafters and the chord.

Newton-Raphson iteration method was used iterative algorithm.

Long-term behaviour and moisture content of wood elements were taken into account in a simplified manner. In fact, modification factors were used to reduce the mean elastic moduli to account for long-term behaviour and the effects of moisture content of wood elements, in accordance with technical literature (Giordano, 1999) and relevant building codes (EC5, 1995). Joint between the main chord and the rafters are characterized by a double tenon and mortise, see Figure 6; the other joints have a single tenon and mortise (Figure 4 and 5). Values of elastic stiffness of the interface elements were calculated according to Equation (1).

The loading scheme consisted of an initial axial compressive force applied at the upper edge of the main rafter, which was derived from the elastic analysis of the whole truss structure. In addition, a linear distribution of pressure load was applied over the upper edge of both rafters which generated a positive bending moment in the joint. The behaviour of the connections is reported in Figure 14 in terms of moment vs rotations curves.

As the rafters-chord connection is considered, the global stiffness of the joint increased after a first part of the response. This circumstance can be related to the loading sequence, that basically reflects the installation process. In fact, during the application of the axial force in the rafter, the contact elements at the joint are firstly engaged and give an initial stress and strain distribution. The subsequent bending moment applied at the edge of the rafter, forcing the skew angle to decrease, interacts with contact elements and generates the observed non linear behaviour.

This trend was also detected for the joints between the diagonal strut and the king post, and between the main rafter and the king post. With reference to the initial linear branch, an estimation of the initial stiffness can be carried out: 1700 kNm/rad for rafters-chord joint; 5000kNm/rad for main rafter-king post

Figure 15. Contour of the compressive stress along the grain (values in N/mm2).

joint; 2900 kNm/rad for diagonal strut-rafter joint; 2800 kNm/rad for diagonal strut-king post joint.

Therefore the elastic behaviour of the joint between the rafters and the chord was influenced significantly by the stiffness of the metal stirrup. In fact, the modelling of the metal stirrup ensured the engagement of the contact elements between the joint elements in the