In full-scale cyclic loading tests conducted by the authors on six one-bay two-story high strength steel moment frames, strains on surfaces of beams, columns and their joints were extensively monitored, and the shear distortions in east-side panel zones at both stories were recorded. The global responses of all specimens at the frame level were prepared and analyzed in the first companion paper. This paper deals with those local responses of members and joints under real boundary conditions in frame structures, which were expected to assist more accurate and reasonable assessment of their seismic behavior than individual tests on extracted specimens. Beams remained essentially elastic throughout all tests, while columns developed significant yielding and maximum moments at column bases that were 30% to 75% higher than the fully plastic moment depending on section slenderness. Compact columns of 460 MPa high strength steels exhibited superior cyclic behavior with stable energy dissipation, so did those panel zones that underwent inelastic deformation. Cover-plate connections finished all loading cycles in all specimens and were effective to reduce the risk of fracture at beam-to-column welds. Continuity plates might not be able to resist the seismic demand specified by current codes, since they fractured in one specimen with a demand-to-capacity ratio of 0.72, which was also the highest among all specimens. Plastic resistances of panel zones predicted by various codes except for the Chinese code, underestimated the experimental results due to the effect of strain hardening, while seismic demands on panel zones specified in various codes differed a lot. It was the Chinese code that both provided the lowest demand and overestimated the resistance, and thus resulted in the weakest panel zone design.
Read full abstract