Experimental and empirical study on the embedment behaviour of wooden and steel dowels in hardwood and softwood

Existing design standards predominantly address metal fastener-based connections, with limited guidance available for wooden dowel connections. Wooden dowels products, however, offer several advantages, including ease of demounting, recyclability, and compatibility with computer numerical control (CNC) fabrication due to the absence of metal components. These benefits have contributed to the growing interest in dowel-laminated timber (DLT) products. In DLT systems, the embedment behaviour of individual dowels and their composite interaction with the base timber are critical to the performance of timber-to-timber connections. Specifically, embedment strength and embedment stiffness serve as key input parameters for determining the load-carrying capacity and stiffness of wooden dowel connections. Understanding embedment behaviour is therefore essential for accurate load transfer prediction, connection design, and structural optimisation. This study investigates the embedment behaviour of wooden and steel dowels in Australian softwood and hardwood panels. Tasmanian Oak (TO) was used as the wooden dowel material, while Spotted Gum (SG) and Radiata Pine (RP) were selected as the base timbers. SG and TO are high-density Australian hardwoods valued for their strength and durability, whereas RP is a widely used softwood known for its ease of processing. Tests were conducted with and without dowel overhangs to examine their influence on embedment strength and failure mechanisms. In addition, TO dowels were tested in isolation under varying boundary conditions to evaluate the effects of confinement and material geometry. The results indicate that wooden dowel overhangs significantly enhance both embedment stiffness and strength. In contrast, steel dowel diameter has no significant effect on embedment strength, and the base timber density has little influence when wooden dowels replace steel. Embedment strength increased from 14.87 MPa (SG) and 8.75 MPa (RP) with wooden dowels to 72.48 MPa (SG) and 37.69 MPa (RP) when steel dowels were used. A refined spring-in-series model was proposed to estimate composite stiffness, and new empirical equations were developed to predict embedment strength for both steel and wooden dowels, including high-density species such as SG. The models provided predictions with R² values of up to 0.80.