Cross-laminated timber, also known as X-Lam or CLT, is well established in
Europe as a construction material. Recently, implementation of X-Lam products and
systems has begun in countries such as Canada, United States, Australia and New
Zealand. So far, no relevant design codes for X-Lam construction were published in
Europe, therefore an extensive research on the field of cross-laminated timber is being
performed by research groups in Europe and overseas. Experimental test results are
required for development of design methods and for verification of design models
accuracy.
This thesis is part of a large research project on the development of a software
for the modelling of CLT structures, including analysis, calculation, design and
verification of connections and panels. It was born as collaboration between Padua
University and Barcelona"s CIMNE (International Centre for Numerical Methods in
Engineering). The research project started with the thesis “Una procedura numerica per
il progetto di edifici in Xlam” by Massimiliano Zecchetto, which develops a software,
using MATLAB interface, only for 2D linear elastic analysis. Follows the phase started
in March 2015, consisting in extending the 2D software to a 3D one, with the severity
caused by modelling in three dimensions. This phase is developed as a common project
and described in this thesis and in “Pre-process for numerical analysis of Cross
Laminated Timber Structures” by Alessandra Ferrandino.
The final aim of the software is to enable the modelling of an X-Lam structure in
the most efficient and reliable way, taking into account its peculiarities. Modelling of
CLT buildings lies into properly model the connections between panels. Through the
connections modelling, the final aim is to enable the check of preliminarily designed
connections or to find them iteratively, starting from hypothetical or random
connections.
This common project develops the pre-process and analysis phases of the 3D
software that allows the automatic modelling of connections between X-Lam panels. To achieve the goal, a new problem type for GiD interface and a new application for
KRATOS framework have been performed. The problem type enables the user to model
a CLT structure, starting from the creation of the geometry and the assignation of
numeric entities (beam, shell, etc.) to geometric ones, having defined the material, and
assigning loads and boundary conditions. The user does not need to create manually the
connections, as conversely needs for all commercial FEM software currently available;
he just set the connection properties to the different sides of the panels. The creation of
the connections is made automatically, keeping into account different typologies of
connections and assembling of Cross-Lam panels. The problem type is special for XLam structures, meaning that all features are intentionally studied for this kind of
structures and the software architecture is planned for future developments of the postprocess phase.
It can be concluded that sound bases for the pre-process and analysis phases of
the software have been laid. However, future research is required to develop the postprocess and verification phases of the research project.
In response to the global drive towards sustainable construction, CLT has emerged as a competitive alternative to other construction materials. CLT buildings taller than 10-storeys and CLT buildings in regions of moderate to high seismicity would be subject to higher lateral loads due to wind and earthquakes than CLT buildings which have already been completed. The lack of structural design codes and limited literature regarding the performance of CLT buildings under lateral loading are barriers to the adoption of CLT for buildings which could experience high lateral loading. Previous research into the behaviour of CLT buildings under lateral loading has involved testing of building components. These studies have generally been limited to testing wall systems and connections which replicate configurations at ground floor storeys in buildings no taller than three storeys. Consequently, to develop the understanding of the performance of multi-storey CLT buildings under lateral loading, the performance of wall systems and connections which replicate conditions of those in above ground floor storeys in buildings taller than three storeys were experimentally investigated. The testing of typical CLT connections involved testing eighteen configurations under cyclic loading in shear and tension. The results of this experimental investigation highlighted the need for capacity-based design of CLT connections to prevent brittle failure. It was found that both hold down and angle bracket connections have strength and stiffness in shear and tension and by considering the strength of the connections in both directions, more economical design of CLT buildings could be achieved. The testing of CLT wall systems involved testing three CLT wall systems with identical configurations under monotonic lateral load and constant vertical load, with vertical loads replicating gravity loads at storeys within a 10-storey CLT building. The results show that vertical load has a significant influence on wall system behaviour; varying the vertical load was found to vary the contribution of deformation mechanisms to global behaviour within the elastic region, reinforcing the need to consider connection design at each individual storey. As there are still no structural design codes for CLT buildings, the accuracy of analytical methods presented within the literature for predicting the behaviour of CLT connections and wall systems under lateral loading was assessed. It was found that the analytical methods for both connections and wall systems are highly inaccurate and do not reflect experimentally observed behaviour.
Advanced sustainable lateral load resisting systems that combine ductile and recyclable materials offer a viable solution to resist seismic load effects in environmentally responsible ways. This paper presents the seismic response of a post-tensioned timber-steel hybrid braced frame. This hybrid system combines glulam frame with steel braces to improve lateral stiffness while providing self-centreing capability under seismic loads. The proposed system is first presented. A detailed numerical model of the proposed system is then developed with emphasis on the connections and inelastic response of bracing members. Various types of braced frames including diagonal, cross and chevron configurations are numerically examined to assess the viability of the proposed concept and to confirm the efficiency of the system. A summary of initial findings is presented to demonstrate usefulness of the hybrid system. The results demonstrate that the proposed system increases overall lateral stiffness and ductility while still being able to achieve self-centring. Some additional information on connection details are provided for implementation in practical structures. The braced-frame solution is expected to widen options for lateral load resisting systems for mid-to-high-rise buildings.
