Withdrawal resistance of self-tapping screws in unidirectional and orthogonal layered timber products

Abstract

Self-tapping screws placed in timber show a high load-carrying capacity if stressed in axial direction. They are widely used in contemporary engineered timber structures, e.g. for joining of linear members like solid timber or glued laminated timber (glulam) and in plane elements like cross laminated timber. In doing so, an influence on the withdrawal capacity by the number and orientation of layers penetrated by the screw can be assumed. Current standards regulate the withdrawal resistance of screws placed in timber (products) only as a function of the product’ characteristic density. We propose to treat the withdrawal strength f ax as a function of (i) the density of the base material and (ii) the number of layers penetrated by the screw. Therefore a stochastic model approach is defined, based on the assumption that the layer involved with the highest density governs the withdrawal resistance of the screw. For verification, results of several withdrawal test series, performed on unidirectional and orthogonal layered specimens with a number of layers N = 1, 2, 3, 5, 6, 10 and 20, get used. Congruent to the model an increase of the withdrawal resistance with N was observed on mean as well as on 5 %-quantiles, whereas the orientation of layers showed a clear but insignificant trend. For a number of layers relevant in practice (N = 3–7) an increase of 7–25 % on the 5 %-quantile of f ax observed in single layer is given. Additional to the discussed uncertainties in the model, in fact a deeper understanding of the behaviour of self-tapping screws placed in layered timber products can be provided. The results may give support to further developments and optimisation of screw joints and the possibility to consider quantitatively more economic withdrawal strength of screws if more than one layer of a timber product is penetrated by the thread.