Effect of roof to wall connection stiffness variations on the load sharing and hold-down forces of Australian timber-framed houses
Introduction
Timber-framed houses are the most common residential building in Australia and the roof of these structures experience high wind loads during windstorms. Thus, the fundamental design objective and task of a house’s timber-framed structural system is to securely transfer the wind loads from roof and walls to the foundations, through their inter-component connections. A better understanding of the load path is essential to evaluate the structural response of a house to windstorms. Uplift or vertical load transmission and lateral load transmission are the two basic types of wind load transmission in a timber-framed house structure. The vertical load transmission typically depends on the roof loads and structural response of the roof structure (i.e. roof cladding, batten, truss, ceiling, ceiling cornice, etc.) and their connections (i.e. batten to cladding, batten to truss, and roof to wall connections). Discontinuity in the load sharing and load transfer will increase loads to structural elements, potentially leading to premature damage to the structural system.
Several research studies associated with load paths have been conducted worldwide [e.g., [1], [2], [3], [4], [5], [6], [7], [8], which have provided some qualitative information of the load sharing of a house structural system. However, there is a lack of quantitative data on load sharing through timber-framed house construction as it is very complex. There are several factors affecting the load sharing in timber-framed construction, such as spacing of members and fasteners (i.e. roof trusses, battens, cladding fixings, etc.), stiffness of members and their connections, fascia beams, roof sheathing type and orientation, etc. [9].
Damage observed after recent windstorms, such as in Cyclone Yasi (2011), Cyclone Marcia (2015), Cyclone Olwyn (2015), Brisbane Severe Storm (2014), and the Tornado in Oklahoma (2015) reported roof structure failures due to combination of high wind speed, high internal pressure and construction defects/human errors, which initiated a cascading structural failure. These failures indicate that timber-framed houses remain vulnerable in cyclonic and non-cyclonic regions. After these events, research and investigations were conducted in order to assess design and structural strength of timber-framed houses [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34]. The windstorm damage investigations show that the loss of roof cladding resulted in extreme damage, leading to a significant loss of strength, which resulted in progressive collapse [10], [14], [16], [17], [23], [30].
Research into full-scale house tests has shown the complexity of actual load transmission was difficult to incorporate in design, without full-scale testing of individual house designs [11], [12], [13], [18], [19], [20], [21]. Recently, several other full-scale house tests have been carried out to evaluate the structural response and load transmission of the North American and Canadian residential houses [15], [25], [26], [27]. The studies by Datin et al. [26], Zisis and Stathopoulos [35], Morrison [15] and Doudak et al. [27] have evaluated the wind load sharing and load path network through a typical timber-framed house in North America and Canada. The results from these studies cannot be used to evaluate the structural stability of Australian timber framed houses as their structural systems, construction methods, RWCs and design wind speeds are different to houses from North America and Canada. These research studies and damage investigations have shown that the roof, roofing components and connections are the most vulnerable structural elements in the timber-framed house structural system [10], [14], [16], [17], [23], [30]. Furthermore, these previous researches have highlighted that the construction defects on the inter-component connections reduce their stiffness and create discontinuity in the load sharing and load transfer to the foundations, which contributes to overloading of connections and cascading house failures to windstorms.
Satheeskumar et al. [32] highlighted that every timber-framed house has construction defects (i.e. missing nails) on the RWCs, reducing the stiffness and uplift capacity up to 40% of the standard specified capacity. The safety factor specified in Australian standards and building codes (AS 1649 [36], NCC [37], AS 1684.3 [38], AS 1720.1 [39], AS 4055 [40], AS/NZS 1170.2 [41]) cover up to 25% of variability. Satheeskumar et al. [33] found that these poorly constructed RWCs, structural and non-structural elements in the roof structure could create variations of the vertical reaction forces at RWCs. This can increase the vulnerability of the timber-framed house structural performance under windstorms. The Australian standards and National Construction Codes [37] have requirements that every structure and structural element should be designed to satisfy the “performance-based design and deemed-to-Satisfy condition”. To satisfy these conditions, the timber framed house and their inter-component connections stiffness, structural response and load sharing must be investigated.
In standard design practice, the hold down forces derived from a single truss analysis does not account for the loads shared to and from adjacent trusses and the connections stiffness variation. This standard design practice could underestimate or overestimate the hold down forces, where underestimation of hold-down forces my overload connections, contributing to the whole-roof failure under windstorm. Thus, this paper aims to investigates the effect of RWC stiffness variation and how its influences the load sharing through the structure and the RWC vertical reaction force of a timber-framed house. This paper further investigates the hold-down forces at the RWCs from realistic wind loads across the roof during a windstorm by combining the wind loads from wind tunnel tests with the load sharing through the structural system with a finite element model (FEM) analysis. Results are compared with RWC hold down forces derived from standard design practice (i.e. basic tributary area analysis using wind loads derived from the wind loading standard AS/NZS1170.2 [41]).
Section snippets
Contemporary house structural system
A field survey of contemporary houses under construction in non-cyclonic regions of Australia, near Brisbane and Melbourne, was conducted by a team from the Cyclone Testing Station (CTS), to determine representative houses and their structural systems [34]. Based on the analysed survey data, the houses in Melbourne are of similar size, shape and construction type to houses in the Brisbane region. The houses’ construction types in both Melbourne and Brisbane are brick veneer with timber trusses
Influence function
The can be determined from the single truss analysis (i.e. traditional method) and full or part of house structure analysis (i.e. load sharing method).
Results and discussion
Truss hold-down forces were derived for Trusses A, B, C, D and E using the time varying pressure distributions for all 36 wind angles obtained from the wind tunnel study and the values obtained from the traditional and load sharing methods. For the discussion in this section, a dimensionless hold-down force coefficient (CN) is introduced below,where AN is the total tributary area of single truss (i.e. 5.6 m2).
Conclusions
This study investigated the effects of stiffness variation in the RWCs on load sharing in the roof structure and hold-down forces of roof trusses by modelling the structure of the timber-framed house. Wind loads on the RWCs were obtained by using vertical reaction influence functions obtained from the single truss analysis (i.e. traditional design method) and FEM analysis of part of a contemporary timber-framed house (i.e. load sharing method). Realistic dynamic wind loads were obtained via a
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This study is part of the research conducted by Climate Adaptation Engineering for Extreme Events Cluster funded by CSIRO Land and Water Business Unit. The authors gratefully acknowledge the funding support of CSIRO and the assistance provided by Dr David Henderson and Prof. John Ginger from Cyclone Testing Station, James Cook University, Townsville, Australia.
References (49)
- et al.
Storm duration effects on roof-to-wall-connection failures a residential, wood-frame, gable roof
Wind Engineering and Industrial Aerodynamics
(2014) - et al.
Failure mechanisms of roof sheathing under fluctuating wind loads
Wind Engineering and Industrial Aerodynamics
(2013) - et al.
Investigation of wood shear walls subjected to lateral load
Structures
(2018) Wind loading of structures: framework, phenomena, tools and codification
Structures
(2017)- et al.
Response of toe-nailed, roof-to-wall connections to extreme wind loads in a full-scale, timber-framed, hip roof
Eng. Struct.
(2013) - et al.
Load sharing and structural response of timber-framed house
Eng. Struct.
(2016) - et al.
Wind load effects and equivalent pressures on low-rise house roofs
Eng. Struct.
(2000) - et al.
Finite element modelling of the structural response of roof to wall framing connections in timber-framed houses
Eng. Struct.
(2017) - et al.
System reliability and load sharing effects in light frame wood construction
J. Struct. Eng.
(1991) - et al.
Rate-of-load and duration-of-load effects for wood fasteners
J. Struct. Eng.
(1999)