Modelling embedment strength of palm wood and cross-laminated timber with mixed palm species

This study addresses the critical need to evaluate and model the embedment behaviour of palm wood for structural applications. While the embedment strength of coconut wood has been investigated and prediction equations have already been proposed, the embedment behaviour of oil palm wood remains underexplored. In this study, the embedment strength of oil palm wood over its entire density range under both parallel- and perpendicular-to-grain loading was measured by implementing half-hole tests based on ASTM D5764–97a. These experimental results revealed that wood density and dowel diameter significantly influence the embedment strength of these palm species, particularly under perpendicular-to-grain loading. Regression analyses and 5-fold cross-validation methods were conducted to develop empirical equations for oil palm wood, and palm wood embedment strength equations that can generalise embedment strength predictions across both coconut and oil palm wood species. The applicability and prediction of these models were validated using experimental results from mixed species cross-laminated timber (CLT) made of coconut and oil palm wood. The results indicated that the palm wood embedment strength predictive equations yield the root mean squared error of 5.23 and 4.31, and R² values of 0.89 and 0.90 under parallel- and perpendicular-to-grain loading, respectively. The model predicts embedment strength for both species and is also applicable to CLT connection design. The Monte Carlo simulation results confirmed that when calculating the embedment strength of mixed species CLT, the density ratio between the outer and inner layers should be considered.