Ricles, Peng, and Lu 2004
An experimental study on beam-to-column connections of square CFT systems was presented by (Ricles et al., 1997; Peng et al., 2000). Full-scale cruciform specimens were tested under cyclic loading. The connection detailing and load transfer mechanism in the connection region were the focus of the research. Three types of connection details were utilized and their strength, stiffness, and ductility properties were investigated.
Experimental Study, Results, and Discussions
The test setup consisted of a square CFT column and two W24X62 steel girders. The column was pinned at the bottom and free to rotate and translate in the plane of the girders at the top. The girders were connected to the mid-height of the column. The measured yield strength of the flange and web of the steel girders were 43.2 and 49.6 ksi, respectively. The measured yield strength of the steel tubes was 55.0 ksi. The L/D ratio was 9 and the D/t ratio was 32.5. The measured compressive strength of the concrete was 8.96 ksi. The connection types used for the specimens included interior diaphragm, extended-tee, and split-tee moment connection details. The latter two details were designed for the plastic hinge to occur in the girders. Among the remaining specimens, two of them (C1R2 and C2R) were designed for panel zone yielding while the others were again proportioned for a weak-beam-strong column condition. The specimens were tested under constant axial load and cyclic shear force, both of which were applied at the top of the column. Four specimens (C1, C1R, C2, C1R2 and C2R) were manufactured with the interior diaphragm detail. The specimens with R in their labels were retrofitted from the previously tested equivalent connections. Full-penetration groove welds around the perimeter of the CFT were used to connect the diaphragm to the steel tube. For specimens C2 and C2R, the weld along one of the webs of the tube was omitted. Full penetration groove welds were also used for welding the girder flanges to the steel tube; these welds were then strengthened with a fillet weld. In specimens C1R2 and C2R, tapered plates were attached to the flanges of the steel girders. Short cover-plates on the flanges were used for specimen C1. In the test, an early fracture occurred at the toe of the cover plate weld of specimen C1. At the end of the testing of specimen C1, extensive girder flange yielding and insignificant shear yielding of the panel zone were observed. The location of fracture for specimens C1R and C2 was the toe of the weld connecting the flange of the girder to the steel tube. Specimen C2 was overloaded unintentionally and the loading was kept at that level to provide symmetric cycles. Fracture occurred when the specimen C2 was overloaded in the other direction. Specimens C1R2 and C2R exhibited extensive shear yielding mainly in the panel zone, while the other specimens with interior diaphragm underwent inelastic deformation in the girders. For specimens C1R2 and C2R, the formation of diagonal shear cracks in the concrete in the panel zone region was also observed after the test and shear buckling of the steel tube web occurred in the panel zone as the story drift got larger. However, the load-deformation response maintained its stable trend for specimens C1R2 and C2R. These specimens achieved shear strengths of approximately 1.76 and 1.49 times the nominal shear capacity of the steel tube, respectively. The groove weld that was omitted in specimen C2R caused the shear strength to decrease.
Extended-tee connection details were utilized for the specimens with labels C3 and C3R. These connections had external diaphragms made from ST7.5X12 structural tees and welded to the column at the corner as well as to the girder flange. Specimen C3R was manufactured by retrofitting of specimen C3 after testing. The only difference between the two specimens was the tapered plates of specimen C3R welded to the flanges of the girders. Specimen C3 experienced a fracture in the tension flange of the girder during the test and extensive yielding of the flange took place. Shear yielding of the panel zone also occurred for specimen C3, although it was not significant. While testing specimen C3R, plastic hinges occurred in the girders and yielding of extended-tee flanges took place. In addition, the panel zone of the specimen C3R also yielded as the tension flange force transferred directly to the steel tube through the extended tees.
The split-tee specimens were labeled as C4, C5, C6 and C7. Post-tensioned through-bolts were utilized to connect the split-tees to the CFT columns. The bolts of all specimens were post-tensioned as per the AISC LRFD (1993) Specification except for specimen C7, whose bolts were post-tensioned to the snug-tighten condition of 40% of the pre-tension level specified in the AISC LRFD (1993) provisions. Thus it was possible to investigate how the amount of bolt pre-tensioning force affected the connection behavior. For specimens C4 and C5, the stem of the tee was connected to the flanges of the girders by bolting, and washer plates were used to prevent flange local buckling along the bolt line. Welding was utilized for connecting the stem of the tee to the girders for specimens C6 and C7. All of the split-tee specimens experienced intensive yielding and local buckling in the web and flanges of the steel girders beyond the connection region.
The extended-tee and split-tee specimens exhibited maximum moment values of 1.23 to 1.56 times the plastic moment of the girders. This overstrength was attributed largely to strain hardening of the girders. The moment capacities started to degrade following local flange buckling of the girders. All split-tee specimens had deformation capacities satisfying the FEMA (1997) requirement of 0.03 rad of plastic rotation. For specimen C4, hole elongation was noted, which caused pinching in the moment-rotation response. The specimens with washer plates and welded tee stems had higher resistance to hole elongation. Specimen C7 also experienced an increase in the bolt force due to the moment acting from the connection on to the bolts; however, no increase in bolt force occurred for the other split-tee specimens as a result of prying action.
From the test results of specimens C1, C3, and C4, it was observed that the initial lateral stiffness (EI) of the columns were well predicted by the AISC LRFD (1993) modified CFT column stiffness and uncracked transformed stiffness. However, for the specimen C3R, its lateral stiffness was closer to the lateral stiffness of the steel tube alone. When the lateral deformations increased, it was found that the lateral stiffness of the column degraded due to concrete cracking and loss of bond between steel and concrete. Among the specimens, the ones with internal diaphragms had a smaller lateral elastic stiffness since the external diaphragms and split tees contributed to the stiffness of the tested assemblies more than internal diaphragms achieved.
Peng, S., Ricles, J. M., Lu, L. (2000). “Full-scale Testing of Seismically Resistant Moment Connections For Concrete-filled Tube Column-to-WF Beam Hybrid Systems,” Composite and Hybrid Structures, Proceedings of the Sixth ASCCS International Conference on Steel-Concrete Composite Structures, Xiao, Y. and Mahin, S. A. (eds.), Los Angeles, California, March 22-24, 2000, Association for International Cooperation and Research in Steel-Concrete Composite Structures, Los Angeles, California, pp. 591-598.
Ricles, J. M., Peng, S. W., and Lu, L. W. (2004). “Seismic Behavior of Composite Concrete Filled Steel Tube Column-Wide Flange Beam Moment Connections,” Journal of Structural Engineering, ASCE, Vol. 130, No. 3, February, pp. 223-232.