The research summarized in this thesis is comprised of an experimental analysis of the mechanical behavior of a wood composite I-joist with different bracing configurations exposed dynamic walking loads. Three 16 in. deep GPI® 65 I-joists were simply supported and laid parallel to each other, while the bracing was attached to the top flange. Five different brace stiffnesses were used: zero stiffness (control), 1.2 lb/in., 8.5 lb/in., 14.0 lb/in. and infinitely stiff. Two different brace configurations were used: one-quarter of the span length (60 in.) and one third the span length (80 in.). The dynamic walking loads consisted of human test subjects attached to a safety platform walking across the I-joist at a designated pace. Experimental results for this research consisted of the I-joist's lateral accelerations, lateral displacements and twist. An Analysis of Covariance (ANCOVA) was used for the statistical analysis of the results and was performed for each measurement. The statistical analysis determined the effects of different bracing configurations, stiffnesses, measurement locations as well as test subjects' weight and occupation. Test results and observed trends are provided for all test configurations. Lateral displacement and twist experienced the same trend throughout the experiment: as brace stiffness increased, lateral displacement and twist decreased. This correlated with basic beam theory and bracing fundamentals. It should be noted that as the stiffness increased, the effect on lateral displacement and twist response decreased. However, the trend for lateral displacement and twist was not observed for the lateral accelerations. The 1.2 lb/in. brace stiffness had much larger lateral accelerations for the 60 in. brace configuration throughout the span and were also larger at the bracing point for the 80 in. brace configuration. This could have been due to the energy applied from the springs or a natural frequency of the I-joist system could have been reached during testing. However, the other four brace stiffnesses followed the same trend as the lateral displacements and twist. In addition, this research demonstrates a method for the measurement of lateral buckling due to worker loads. The mitigation of lateral buckling can use appropriate bracing systems. The measurements of the change in lateral buckling behavior can be used to develop safety devices and ultimately ensure the protection of construction workers. / Master of Science

Recent research has shown that a significant number of the falls from elevation occur when laborers are working on unfinished structures. Workers walking on wood I-joists on roofs and floors are prone to fall hazards. Wood I-joists have been replacing dimension lumber for many floor systems and a substantial number of roof systems in light-frame construction. Wood I-joists are designed to resist axial stresses on the flanges and shear stresses on the web while minimizing material used. However, wood I-joists have poor resistance to applied lateral and torsional loads and are susceptible to lateral-torsional buckling instability. Workers walking on unbraced or partially braced wood I-joists can induce axial and lateral forces as well as twist. Experimental testing demonstrated that workers cause lateral-torsional buckling instability in wood I-joists. However, no research was found related to the lateral-torsional buckling instability induced by individuals walking on the wood I-joists. Furthermore, no research was found considering the effects of the supported end conditions and partial bracing in the lateral-torsional buckling instability of wood I-joists. The goal of this research was to derive mathematical models to predict the dynamic lateral-torsional buckling instability of wood composite I-joists loaded by individuals walking considering different supported end conditions and bracing system configurations. The dynamic lateral-torsional buckling instability was analyzed by linearly combining the static lateral-torsional buckling instability with the lateral bending motion of the wood I-joists. Mathematical models were derived to calculate the static critical loads for the simply supported end condition and four wood I-joist hanger supported end conditions. Additionally, mathematical models were derived to calculate the dynamic maximum lateral displacements and positions of the individual walking on the wood I-joists for the same five different supported end conditions. Three different lean-on bracing systems were investigated, non-bracing, one-bracing, and two-bracing systems. Mathematical models were derived to calculate the amount of constraint due to the lean-on bracing system. The derived mathematical models were validated by comparison to data from testing for all supported end conditions and bracing systems The predicted critical loads using the static buckling theoretical models for the non-bracing system and the static buckling theoretical models combined with the bracing theoretical models for the simply and hanger supported end conditions agreed well with the critical loads obtained from testing for the two wood I-joist sizes investigated. The predicted maximum lateral displacements and individual positions using the bending motion theoretical models for the simply and hanger supported end conditions agreed well with the corresponding maximum lateral displacements and individual positions obtained from testing for both wood I-joist sizes. Results showed that; a) the supported end condition influenced the critical loads, maximum lateral displacements and individual positions, b) the bracing system increased the critical loads and reduced the maximum lateral displacements, c) the critical load increased as the load position displaced away from the wood I-joist mid-span, d) the critical load reduced as the initial lateral displacement of the wood I-joist increased and e) the wood I-joist mid-span was the critical point in the dynamic lateral-torsional buckling instability. / Ph. D.