Effect of sub-zero temperature, moisture content and loading rate on the bending strength of beech and birch wooden dowels

Wooden dowels are emerging as a promising alternative or complement to adhesives in adhesive-free timber products and pure-wood structural joints. The performance of these connections is critically governed by the mechanical properties of the dowels, which can be influenced by environmental and loading conditions. This study experimentally evaluated the bending strength of approximately 100 beech and birch dowels (diameter 16 mm) under two temperatures (20 (Formula presented) and (Formula presented) (Formula presented) ), two moisture states (dry and water-soaked), and five loading rates (0.04 to 25 mm/min). Four-point bending tests were conducted in accordance with EN 408 using a constant-rate displacement protocol, and moisture content was determined gravimetrically for each conditioning state. The results were statistically assessed using a linear mixed-effects (LME) model to quantify main effects and interactions among species, temperature, moisture state, and loading rate. Birch dowels consistently outperformed beech, with an average strength margin of 19%. Moisture content emerged as the dominant factor, with soaking causing a mean 47% strength reduction, while cooling to (Formula presented) (Formula presented) provided only modest gains (5%-8%), particularly in wet specimens. Rate sensitivity was notable in dry dowels but became negligible once saturated. An empirical design model was proposed to estimate a modification factor for predicting bending strength, incorporating temperature-moisture interactions and rate effects, and calibrated for 95% reliability. SEM (Scanning Electron Microscopy) observations indicated that moisture promotes more ductile fracture through fibre pull-out, whereas freezing partially restores matrix stiffness, resulting in cleaner and more brittle fracture surfaces. The results provide practical modification factors for dowel design under coupled environmental and rate effects, supporting more robust and sustainable dowel-based connection systems. Future work should extend the experimental matrix to intermediate temperatures and moisture levels and assess long-term and cyclic actions (e.g., fatigue and freeze-thaw exposure) to further support standardization and code implementation.