Cross-laminated timber panels (X-LAM) are used more and more as construction material for buildings. Since such a panels are very rigid, the mechanical connections between adjacent panels and with the foundations are the most important component affecting the static and cyclic behaviour of X-LAM buildings. For a reliable seismic analysis, therefore, it is crucial to investigate the cyclic behaviour of the connections between wall panels.

 

Within the SOFIE project, a research project sponsored by the Trento Province (Italy), coordinated and conducted by the CNR-IVALSA (Italian National Research Council – Trees and Timber Institute), an extensive experimental programme on X-LAM buildings was carried out. The research project included racking tests of walls, pseudo-dynamic tests of a full-scale one-storey building and shaking table tests of a 3-storey and a 7-storey residential building. Experimental tests provided excellent outcomes, as the buildings were able to survive strong recorded earthquakes, such as Kobe earthquake (1995), virtually undamaged, while at the same time demonstrating significant energy dissipation. However, further research in this field is still needed, particularly on the cyclic behaviour of single connections, single and coupled wall panels. The outcomes of such tests can then be used to calibrate advanced component models for numerical analyses, as well as to calculate ductility and overstrength factors, which are needed in seismic design and are currently not provided by current codes of practice such as the Eurocode 8 and the Italian regulation.

 

The paper presents results in terms of strength, stiffness, energy dissipation, and ductility from cyclic tests on the lateral resistance of cross-laminated wooden panels with different configurations and mechanical joints. A comparison between test results and estimated calculations according to Eurocodes and current literature is also provided.

 

Several configurations of wall panels were investigated by means of cyclic tests, carried out according to EN 12512 standard. Wall configurations included different anchoring systems, single and coupled walls, and different types of screws to connect the panels. The test results confirmed that the layout and design of the joints is critical for the overall behaviour of the structural system. While the wooden panels behaved almost completely rigid, concentration of forces and deformations occured in a relatively small region, in the connection parts, leading to local failure. The use of step joints with single row of screws or joints with LVL strip with double row of screws as a connection between adjacent walls proved to be an effective solution not only to reduce the wall stiffness and therefore the seismic input load, but also to improve the global structural ductility.

 

An extended experimental programme on typical X-LAM connections was also performed. Shear and axial cyclic tests were carried out on hold-downs and steel angles connectors, which are used to anchor the wall panels. Shear tests were also performed on mechanical screwed connections between adjacent X-LAM panels working in the same plane. Cyclic tests were also carried out on orthogonally connected panels subjected to shear and withdrawal load.