Investigation of the uplift load path within the gable roof of a wood-frame residential building using a full-scale wind tunnel and non-linear finite element modelling

This paper examines the vertical load path within the roof of a light-frame wood structure from full-scale wind tunnel experiments. The global roof uplift and overturning moments obtained from integrating pressure data are in good agreement with the those captured by load cells located at the roof-to-wall connections (RTWC). Examination of the uplift measured in each roof truss at the critical wind angles indicates that a large proportion of roof uplift loads are transferred to the gable end truss – more than would be estimated based on tributary-area calculations using the aerodynamic measurements. The amount of load transfer to the gable end is influenced most by the number of points of connection between the bottom of the gable end truss and the top of the wall below. The archetype roof indicates little dynamic amplification or attenuation in the range of frequencies in which typical wind fluctuations occur. Therefore, a quasi-static modelling approach provides a good indication of load transfer behaviour while being more computationally efficient than time-history analysis. A three-dimensional finite element model of the test roof is developed to further investigate the load distribution behaviour observed in the tests. Toe-nailed RTWCs are modelled with non-linear withdrawal and shear behaviour, allowing for wind speeds causing roof loss to be estimated. The results from the non-linear modelling indicate that the gust wind speeds causing failure of a typical light-frame wood roof can be increased from about 165 km/h to 210 km/h by providing additional nailed connections along the gable end truss roof-to-wall interface. • Examining the vertical load path within the roof of a light-frame wood structure using full-scale wind tunnel experiments. • Most of the total roof uplift goes into the gable end truss, influenced by the number of connections to the wall below. • Minimal dynamic effects in the range of typical wind frequencies make quasi-static modelling efficient and effective. • Non-linear modeling of roof-to-wall connections found that additional gable end connections increases failure wind speeds.