Analytical Investigation of Saddle Connections for Overhead Sign Trusses with Respect to Strength and Fatigue Performance

Abstract

Bridge-type overhead truss sign structures (OHTSS) are widely used over active highways across the states. An OHTSS is comprised of a 3D truss and two support frames at each end. The structures are usually made of steel or aluminum. Many state DOTs use their own types of connections that are not documented in specifications. Since 2015, the Kansas DOT has used a type of ‘saddle connection’ at the joints of truss chords and support frame pipes. Wind loads are the primary type of load a sign structure resists besides the gravity load. Since wind loads are periodic, fatigue properties are important in the design of OHTSS. As a newly developed connection, the Kansas DOT sought information regarding the mechanical performance of the saddle connection. Studies were needed to verify the safety of the connections, particularly regarding its fatigue susceptibility. This report present a study mainly aimed at evaluating the fatigue susceptibility of the saddle connections using finite element analysis (FEA). The study consisted of the following four parts: Part 1 - Global behavior analysis: an analysis aimed at determining the global behavior of the structures and the location of critical connections. Linear-elastic material properties were used. Part 2 - Structural Hot Spot Stress analysis: an analysis was performed to determine structural Hot Spot Stresses along each weld in the critical connections identified in Part 1. Linear- elastic material properties were used. Part 3 – Effective notch stress analysis: a linear-elastic analysis using the effective notch stress method to evaluate three welds identified to have larger stresses in Part 2. Linear-elastic material properties were used. Part 4 – Extreme loading analysis: An analysis to evaluate the behavior of the saddle connections and the overall structures under extreme loading and provide comments regarding the strength-related safety of the saddle connections. Elastic-perfectly plastic material properties were used. Sign structures of four span lengths, including 60 ft, 83 ft, 110 ft, and 137 ft, were analyzed in Part 1 and Part 2. The 137 ft span structure was analyzed in part three using the effective notch stress method. The 60 ft and 137 ft span structures were analyzed in part four. In Parts 1 and 2, AASHTO fatigue loads, including natural wind gusts and truck-induced gusts, were applied in six load modes. They included: natural wind blowing from the back, front, and side of sign structures; and truck-induced gusts acting on the right, middle, and left 12 ft of sign trusses. In Part 3, the AASHTO fatigue load of the natural wind blowing from behind the sign structure was applied. In Part 4, the overall structures and the saddle connections were loaded until the analysis terminated. The termination of analysis was governed by loss of stiffness due to the yielding of material. The study resulted in conclusions that the natural wind in the direction facing the sign panel almost always governed the fatigue demand. The bottom saddle connections were more susceptible to fatigue damage than the top saddle connections, especially the stiffener-to-pipe weld in the bottom saddle connection. Fatigue failures of the saddle connections are not likely to occur in expected real use, but attention should be paid to the stiffener-to-pipe weld in the bottom saddle connection. The analysis of the structures under extreme loading suggests that the ultimate strength of saddle connections do not govern the strength of the overall structures.