Self-tapping screws have been widely used in timber constructions. The current practice considers self-tapping screws perform best in connecting two members when they are fully threaded; however, the drive-in resistance caused by the friction between the wood and the screw can potentially damage the screw and reduce the effectiveness of its applications. The relationship between their thread configuration and the drive-in torque force has not been investigated, and the question as to how would knots in the member affect the drive-in force, remains a unanswered. In this study a series of tests was conducted, aiming to demonstrate the influence of thread configuration on the drive-in torque of screws. Two types of self-tapping screws and three different thread configurations were studied. The drive-in torque for partially threaded screws was found to be significantly slower than that of fully threaded ones. The results showed that knots can significantly influence the positioning of the screw and increase the drive-in torque. The application of a pre-drilled hole was found to be an effective way to minimise the influence of knots. This paper points out that with appropriate consideration of thread configuration, partially threaded self-tapping screws can not only achieve the same efficiency as fully threaded ones, they will also benefit from reduced drive-in torque force.

Large-span roofs can be damaged by uneven snow loads caused by snow drifting. To predict wind-induced snow loads more accurately, this paper proposes a numerical simulation method. The method considers the interaction of wind and snow, the changing characteristics of snow phase boundaries over time, and corrects for snow deposition and erosion flux. The effectiveness of the method was confirmed by simulating wind–snow flow on a three-centered cylindrical shell and comparing the results with wind tunnel measurements. The simulation considered different snowfall conditions and time-varying snow phase boundaries on a large-span shell-shaped roof. The study investigated the influence of snowfall conditions and dynamic changes in snow phase boundaries on snow cover distribution by comparing simulation results.