Modeling load-slip behavior of nailed joints

Abstract

A commonly used mathematical model to predict the lateral load-slip (P-∆) behavior of nailed joints in wood is expanded in this study to account for several important characteristics of the connection. Nonlinear curve-fitting procedures are used to develop prediction relationships for parameters A and B in equation (P = A log<inf>10</inf>[l + B∆]) for a variety of joint material and geometric characteristics. P-∆ curves are experimentally obtained for 292 single-shear nailed joints made with main and side members consisting of seven hardwoods indigenous to the Amazonian region of Brazil, four diameters of common wire nails (2.87, 3.33, 3.76, and 4.07 mm), three nominal moisture contents (6, 12, and 18%), and six nominal side-member thicknesses (12,19, 25, 31, 38, and 50 mm). The specific gravity (oven-dry volume basis) of individual specimens ranges from 0.36 to 0.85. This range is larger than would be found among temperate zone species customarily used in construction. A contemporary statistical software package is used to perform regression analyses to predict A and B from 33 functions of various combinations of the nail diameter, side-member thickness, and main- and side-member specific gravity. Results indicate that the best prediction equation models the shape of the experimental curves with reasonable accuracy. When the estimated parameters are used to predict the P-∆ curves for the 292 test joints, an average mean deviation of about —11% was found. The negative value suggests that the model tended to overpredict joint stiffness. © ASCE.