Impact of connections on the system behaviour and reliability of timber - Nonlinear Finite Element analyses of laterally loaded CLT wall

Typ
Examensarbete för masterexamen
Program
Publicerad
2019
Författare
Rosenberg, Anna
Henriksson, Michaela
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Current design of timber structures is usually based on an element-by-element approach, mostly for simplicity and due to lack of knowledge of the exact structural behaviour of the connections. For Cross Laminated Timber (CLT) structures, the behaviour of the applied connections, i.e. stiffness and strength, is of high importance in terms of global structural behaviour. Improved knowledge about the stiffness properties and the load-deformation behaviour of connections is therefore crucial for a more efficient and reliable design of CLT systems. The aim is to develop guidance for a more robust and reliable design of CLT structures, focusing on the stiffness and load-deformation behaviour of connections in laterally loaded CLT wall systems. The reliability of CLT wall structures is evaluated by accounting for the variability of the properties of the applied connections, including; load-carrying capacity, initial stiffness and ductility. Initially, a linear analysis is carried out for two types of connections, hold-downs and angle brackets, where wall systems consisting of two, four and eight connections are studied. When the linear behaviour is verified through comparison of the globalwall stiffness between the FE-analysis and linear-elastic hand calculations, a nonlinear FE-analysis is performed, this time only for wall systems with hold-downs. Here, the global load-carrying capacity and the load distribution in the systems are studied. The analysis is based on; already existing theories regarding load-deformation, test data and coefficients of variability provided in published reports. To take the variability of the connections into consideration to be able to evaluate the structural reliability of an entire structure, Monte Carlo simulations were applied on the nonlinear FE-analyses. Additionally, the ductility ratio in the connections is varied, to make it possible to study the influence of the ductility on the global load-carrying capacity. In general, both the nonlinear analyses and the design approach according to Eurocode 5 show that the first connection in the wall system is the most important one in terms of global load-carrying capacity. The nonlinear analyses show that when the maximum load is reached in the first connection, the load is redistributed to the other connections through parallel action, and the global load-carrying capacity of the wall system is not reached until several of the connections have come across their maximum point. While, in Eurocode, the global load-carrying capacity of the wall system is attained when the load-carrying capacity of the first connection is reached. Further, the effects of the variation of the material properties decrease with an increase of number of elements in the structure, provided that sufficient ductility of the connections can be guaranteed. For insufficient load-redistribution between the elements, however, the variation of the connections can lead to unpredictable global failure. This also needs to be accounted for in Eurocode 5.
Beskrivning
Ämne/nyckelord
CLT wall system,, Cross Laminated Timber structure,, variability of connection properties, structural reliability,, global load-carrying capacity,, load-displacement curve,, Eurocode 5,, ductility
Citation
Arkitekt (konstruktör)
Geografisk plats
Byggnad (typ)
Byggår
Modelltyp
Skala
Teknik / material