Investigation of Single and Two Bolt Connections Perpendicular to Grain in Laminated Veneer Lumber

Bolted connection with perpendicular to grain loading has been considered as a high priority research area by Smith and Foliente (2002), for the advancement of the load resistance factor design (LRFD) of connections. The results obtained from the experimental testing of this research will provide information regarding the behavior of connections at conditions of capacity and yield, and a comparison between single and two bolted connections for laminated veneer lumber (LVL) from different manufacturers. Comparison of the experimental results with the predicted results from three models: Technical Report -12 (AF&PA 1999), Van der Put and Leijten (2000) and Jensen et. al. (2003), for single and two bolt connections loaded perpendicular to grain will help in accurately predicting LVL connection behavior. Success in achieving the goals of this research will provide enhancement of knowledge and information for single and two bolted connections loaded in perpendicular to the grain connections for LVL and thereby help in calibrating LRFD parameters on pure reliability basis in future. The variables considered included LVL from two different manufacturers, single and two bolt connections with different bolt sizes and loaded edge distances. The connections were loaded to capacity for all the tests. Tests for the material property input values required for these models were also performed as a part of this research.

Connection testing showed splitting failures combined with crushing of main member material and formation of a single plastic moment. Connection resistance increased with increased loaded edge distance and number of bolts. The allowable shear design value controlled the National Design Specification Allowable Stress Design (NDS ASD) lateral design value to the connection design except for one connection configuration with 7D loaded edge distance for two bolts of ½ inch, where connection design strength values controlled. The displacement limit decided for the dowel bearing strength test had a direct impact on the predicted TR-12 capacity values. The capacity resistance calculated by both fracture models increased with increase in loaded edge distances. The Mode-I fracture energy values directly affected the predicted fracture model values. The tension perpendicular to grain strength values directly affected the Jensen model values.

Statistical comparison of 4D and 7D loaded edge distances and LVL-1 and LVL-2 material revealed that Van der Put model had no difference in the calculated to test (C/T) ratios with respect to different loaded edge distances and materials and the Jensen model predicted the C/T ratios at 4D to be significantly greater than at 7D and for LVL-1 to be significantly greater than LVL-2. Van der Put model over predicts at capacity and the C/T values are consistent with change in loaded edge distance. Jensen model C/T ratios over predicted for single bolt connection and predicts accurate for two bolt connection with respect to loaded edge distances. Comparing the two fracture models with a ductile model TR-12 with respect to different loaded edge distances, material, number and size of bolts, Jensen model best predicted the C/T ratios. The Van der Put model tended to over predict values, while the TR-12 model had no consistent trend in C/T ratios, but seemed to be affected inversely by changes in loaded edge distance.