Overstrength and variance analysis of angle brackets for timber structures considering the uncertainty in the nail–timber interaction

Engineered wood products, along with advanced processing and fabrication, are pushing the limits of modern wood construction. However, the capacities of such systems are often determined by the capabilities of the connections. Typical connections for timber beams consist of angle brackets with annular ring nails. The load-bearing capacity of these connections is usually provided by their producers and is based on experimental testing, where the connection is typically subjected to short-term monotonic loading. However, the variability of angle bracket capacity is generally governed by uncertainties on the timber side, specifically the interaction between nails and wood. This paper presents the experimental results and a parametric numerical element model for predicting the load-bearing capacity of connections with nailed angle brackets, considering the effects of uncertainties in the nail–timber constitutive behaviour. Three metal brackets for beam–column connections are analysed, and their performance in various loading arrangements is examined. Detailed finite element models have been developed for each connection, where the response of each nail is defined by a semi-physical constitutive relationship calibrated on separate axial and transversal tests on isolated nails. The authors propagate the uncertainty in the nail–timber interaction on the capacity of the angle brackets following a Monte Carlo-based approach in the three considered loading scenarios. The generated dataset is used to estimate the overstrength factors of the connections and explain the variance of the estimated capacity values based on a multilinear regression surrogate model. • Experimental tests on off-the-shelf angle brackets. • Three angle brackets under three loading scenarios. • Stochastic nail–timber constitutive law. • Uncertainty propagation from the nail–timber to the capacity. • Variance explanation and overstrength factors.