| Gusset plate connections are commonly used in bridge trusses and braced steel frames to transfer forces from one structural member to another. Due to the complexity of these connections, it is extremely difficult to evaluate the strength of gusset plate connections. Hence, a research program was initiated to investigate the compressive behavior and ultimate strength of gusset plate connections by testing full-scale diagonal bracing connections. In total, twenty-one tests were conducted on nineteen specimens. The test parameters included gusset plate thickness, size, brace angle (30{dollar}spcirc{dollar} and 45{dollar}spcirc{dollar}) and out-of-plane restraint boundary conditions. In addition, the effects of frame action on the compressive behavior of gusset plate connections were investigated by applying beam and column moment to the specimens. Out-of-plane loading eccentricity, which happens frequently in tubular bracing members with a slotted-in splice plate, was also examined by testing three eccentrically connected specimens.; In general, the gusset plate specimens were failed by sway buckling of the connection. However, for the specimens tested with out-of-plane restraint, local buckling of the free edges was observed. The failure mode of the eccentrically loaded specimens was the yielding of the splice member at the conjunction of gusset-to-splice. In general, except for the eccentrically loaded specimens, the test results indicated that significant yielding of the gusset plate specimens prior to reaching the ultimate load was observed. However, only limited yielding was observed for the specimens with a plate thickness of 6.5 mm. The ultimate load of the specimens increased almost linearly proportional to the gusset plate thickness and decreased with increasing plate size. While the out-of-plane restraint boundary condition had negligible effects on the ultimate load of the compact specimens (500 x 400), the ultimate load of the slender specimens (850 x 700) increased when out-of-plane restraint was applied. A slight decrease in:he ultimate load of the specimens was observed when a 30{dollar}spcirc{dollar} brace was used instead of a 45{dollar}spcirc{dollar}. The beam and column moment had only negligible effects on the ultimate load of the specimens; however, yielding of the specimens was detected at a load level significantly lower than that predicted by the Whitmore method. The out-of-plane loading eccentricity significantly reduced the ultimate load of the specimen.; The finite element analyses of the plastic bifurcation buckling loads of the test specimens carried out using the program ANSYS were in reasonable agreement with the test results. The elastic stress predicted by ANSYS at the gusset plate area beneath the splicing member agreed well with the experimental stress evaluated from the rosette readings. For the eccentrically loaded specimens, the load deflection analysis done by ANSYS also compared well with the experimental ultimate loads. Based on the available test data and the finite element analysis, a modified Thornton method is proposed to estimate the inelastic buckling strength of the gusset plate specimens. The beam-column equation is also recommended for the design of eccentrically loaded specimens. |
