Intermediate Web Stiffener Spacing Evaluation for Shear Links

Abstract

Eccentrically braced frames (EBFs) are a widely used steel seismic resisting system due to their combination of high elastic stiffness and high ductility. The key component of this system is the portion of the beam in the frame that is formed by an incoming offset brace or braces. In a seismic event, this link undergoes inelastic rotational deformations. Typically, short elements are used for this link, forcing it to yield primarily in shear. Due to a number of recent developments in the material properties and accepted testing protocols for shear links, this paper reexamined the current intermediate web stiffener spacing requirements. Furthermore, the current web stiffener spacing requirements are based on typical web aspect ratios that were used in practice at the time the requirements were drafted, because collector beams and links were of the same section size. Currently, there is an increased use of bolted links, which decouples this requirement. A full-scale, bolted, D-braced EBF was constructed in order to test link sections of various sizes. Three hot rolled section links and four built-up section links were designed, fabricated, and tested using the latest American Institute of Steel Construction link provisions and loading protocol. The links were grouped by their web aspect ratio, either being slender or stocky; stocky links had a ratio of web height to web thickness below 24. Non-code-compliant, unstiffened stocky links performed to an exceptional level, achieving inelastic rotations of 0.17 and 0.19 rad, compared with the 0.08 rad value required by the seismic standards. Attaching stiffeners to stocky webs was found to be either superfluous or detrimental. In order to correlate shear link performance to stiffener spacing design equations, a database of previously tested shear link performance results and geometries was constructed and incorporated the links tested herein. Design equations were derived for each of the web aspect ratio groups, with the specification for slender links remaining largely unchanged, and with the specification for stocky links taking account of the improved performance when fewer intermediate stiffeners are used. The stocky built-up specimens from this study exhibited higher overstrengths (1.80–2.00) than those typically expected based on the previous link testing literature (1.50), and will require appropriate factors for use in design. Further research is required to ascertain the cause of the overstrength, because currently available data does not provide a definitive answer.