Experimental Investigation of Reduced Beam Section Moment Connections for Jumbo Steel Wide Flange Sections

Currently, reduced beam section (RBS) moment connections prequalified for use in special moment resisting frames (SMRF) are limited by AISC 358-16 Prequalified Connections for Special and Intermediate Steel Moment Frames for Seismic Applications to a beam weight of 302 lb/ft, a nominal beam depth of W36, a beam flange thickness of 1.75 inches, and a nominal column depth of W36. Full-scale cyclic testing was conducted on four jumbo-sized beam-to-column RBS moment connections (SP1 through SP4) with a beam weight as large as 925 lb/ft, nominal beam depth as large as W44, beam flange thickness as large as 4.5 in., and column nominal depth as large as W40. The purpose of the experiments was to 1) evaluate if increasing the RBS prequalification limits is appropriate, 2) investigate the cyclic behavior of moment connections with jumbo beam and column sections, and 3) examine column panel zone inelasticity with jumbo columns. The specimens were subjected to the qualification displacement protocol given in AISC 341-16 Seismic Provisions for Structural Steel Buildings and sufficient instrumentation was included to support the research objectives listed above.\n\nSpecimens SP2 (W44x230 beam and a W14x342 column) and SP4 (W44x408 beam and a W40x503 column) were designed to explore the use of beam sections deeper than nominal W36, demonstrated ductility consistent with the SMRF requirements, and neither experienced fracture. Both specimens underwent local buckling at approximately 3% story drift which was associated with strength degradation and inelastic rotation at the RBS. Specimen SP4, which combined a large W44 beam without lateral bracing at the RBS with a deep W40x503 column, experienced column twisting and beam lateral-torsional buckling.\n\nSpecimens SP1 (W36x652 beam and a W14x873 column) and SP3 (W36x925 beam and a W14x873 column) were designed to explore the use of heavier beams and both fractured at the root of the beam bottom flange weld in a brittle manner. Fracture surfaces were studied, and the material was tested to determine the causes of the fracture. It was concluded that the following factors contributed to the fractures: 1) lack of local buckling of the jumbo beams eliminated the strength degradation that typically reduces the demand at the welds, 2) there were significant inelastic strain demands in the beam connection at the column face, 3) inelastic panel zone behavior caused a greater strain demand on the weld, and 4) a nonlinear strain distribution across the beam depth at the weld indicated plane sections did not remain plane. Material testing also found that the material properties varied significantly across the cross-section of the W14x873 column with higher yield stress and hardness at the surface, that the weld material had adequate strength, and there were small circular indications found on the fracture surface.\n\nBased on the results of Specimens SP2 and SP4, it is reasonable to change the prequalification limits for RBS to a beam weight of 408 lb/ft, a nominal beam depth of W44, a beam flange thickness of 2.17 in., and a column nominal depth of W40. It is recommended that lateral bracing be required at the RBS even in the presence of a composite slab for beams and columns beyond the current beam and column limitations in AISC 358-16. Additional testing is required to further evaluate ways to create ductile RBS connections in beams as large as Specimens SP1 and SP3. The panel zone shear strength equations in Chapter J of AISC 360-16 were found to grossly over-predict the experimentally measured shear strength of the jumbo column panel zones, and it is recommended that there be limits on the size of column for which these equations can apply.