Buckling capacity model for timber screws loaded in compression: Experimental, analytical and FE investigations

This paper investigates the buckling of screws loaded in compression inserted into timber members. Screws are often used as a reinforcement in timber structures. However, under compression forces, they are prone to axial buckling. The current model for the screw buckling, enclosed in the EC5 proposal, is based on the general framework of EC3 for the instability of compressed steel members. The main shortcomings of the current formulation for the buckling of screws are the following. (1) The analytical expression for calculating the theoretical buckling load does not follow the observed modes. (2) Due to the need for dedicated studies, the value of the imperfection coefficient is arbitrarily chosen. This paper fills the above gaps. Firstly, a simple analytical expression for predicting the buckling of screws is proposed and validated against experimental and finite element (FE) findings. Furthermore, the formulation adopts a more accurate expression for lateral deformation based on experimental observation. Secondly, a FE model calibrated on experimental tests is used to estimate the defect coefficients of the instability curves as a function of the amplitude of the geometric defects of the screw, expressed as a fraction of its length. Finally, a Markov chain Monte Carlo analysis is carried out to simulate the capacity of screws with different sizes, assuming the uncertainty of all input parameters sampled from suitable probability distributions. The results are used to validate the proposed deterministic capacity model and estimate the uncertainty factors of the design equation.

Automated summary

This paper investigates the buckling of screws loaded in compression inserted into timber members. The current model for screw buckling based on EC5 proposal has shortcomings in terms of analytical expression and imperfection coefficient. This paper proposes a simple analytical expression for screw buckling and adopts a more accurate expression for lateral deformation based on experimental observation. Furthermore, a FE model and Markov chain Monte Carlo analysis are used to estimate the defect coefficients and validate the proposed deterministic capacity model.