Static Strength Of SHS Joints Reinforced With Doubler Or Collar Plates

Collar reinforced and double reinforced square hollow section(SHS) joints arefirst used in ocean structure to prevent the punching failure. In recent years, collarreinforcement and double reinforcement are widely used in structures, while there arenot a design formula and a large number of experimental data in standards or studieswhich have been published. Thus this paper will deduce practical design formulas ofcollar reinforced and double reinforced SHS joints by experiment studies and finiteelement analysis(FEA).This study first compare the static performance of collar reinforced and doublereinforced SHS joints with unreinforced joints. The details of specimens, loadingdevices design, displacement and strain gauges location are described, then failuremodes, load-deformation curves, load-strain strength curves are analysed, theinfluence of different connection. brace to chord side length ratio β, reinforced platethickness to static performance are discussioned. The result of axial load experimentschemes the side wall of chord protruded outward, and concave deformation occurredat the intersection area of the chord in all specimens. In the collar plate reinforcedjoints and the doubler plate reinforced joints, the collar plate and doubler plate bentrespectively in ultimate state. The initial stiffness and ultimate capacity increasedremarkably as the β value increased, but the plastic behavior decreased. When β valueis small, the ultimate capacity of collar plate reinforced joints is close to that ofdoubler plate reinforced joints, both of which are larger than that of unreinforcedjoints. When β value is large, the ultimate capacity of doubler plate reinforced jointsis the largest, that of collar plate reinforced joints comes second, and that ofunreinforced joints is the smallest. As thickness of collar plate increases, the ultimatecapacity, ultimate deformation, ductility and initial stiffness of collar plate reinforcedjoints increase remarkably. As thickness of doubler plate increases, the ultimatecapacity, ultimate deformation, ductility of doubler reinforced joints decrease, butinitial stiffness of the joints increases remarkably. The result of in-plane bendingexperiment schemes with the same β, the in-plane flexural initial stiffness andultimate capacity of doubler plate reinforced joint, collar plate reinforced joints,unreinforced joints decreases progressively. As β increases, the in-plane flexuralultimate capacity and initial stiffness of joints with the same reinforcement typeincrease, but ductility of joints decreases. The thickness of reinforced plate has noobvious effect on in-plane flexural initial stiffness and ultimate capacity of joints. Theconcave and convex deformations of every specimen have good symmetry; as βincreases, the yield and ultimate deformation of the chord decreases. The result ofout-of-plane bending experiment schemes with the same β, the out-of-plane flexuralinitial stiffness and ultimate capacity of collar plate reinforced joint, doubler platereinforced joints, unreinforced joints decreases progressively. As β increases, theout-of-plane flexural ultimate capacity and initial stiffness of joints with the samereinforcement type increase, but ductility of joints decreases. The thickness of reinforced plate has no obvious effect on out-of-plane flexural initial stiffness ofjoints. The concave and convex deformations of every specimen have good symmetry;as β increases, the yield and ultimate deformation of the chord decreases.Base on test results, use the FEA software Abaqus, Adjust theboundary conditions and mesh method, take the displacement loading as loadingmethod, simulate the collar reinforced and double reinforced joints, then compare theresults with the test results. Comparison results show that: the failure modes,load-deformation curves are similar to test results, it follows that Abaqus has goodadaptability, proves validation of FE method available based on the FE models.Base on good adaptability of Abaqus, this paper investigate the staticperformance of collar and double reinforced SHS joints while brace to chord sidelength ratio β is0.3,0.6,0.9, reinforced plate thickness to chord wall thickness ratio λis1.5,2.0,2.5,3.0, reinforced plate length to brace side length ratio Δ is1.0,1.5,2.0,2.5,simulate96axial load speciments,96in-plane bending speciments,96out-of-planebending speciments, and the simulation results are compared, the results show that:As β increases, the ultimate capacity of joints with the same reinforcement typeincrease; effect of λ and Δ has a stage, only through the proper combination, theeffect can be reflected. Some recommendations are made: in axial load speciments,the Δ should be less than2.5and λ should be larger than2while β value is small,improve ultimate capacity by improving the plate length while β value is large, at thesame time, the influence of plate thickness is small; in in-plane bending speciments,improve ultimate capacity by improve Δ value and λ should be larger than2while βvalue is small, Δ advice from2~2.5while β value is not small; in out-of-planebending speciments, Δ should be larger than2and λ advice from2~2.5while β valueis small, Δ advice from2~2.5while β value is not small, in double reinforced joints, βshould be less than0.6.At last, based on calculation ultimate capacity of unreinforced SHS jointsadopting Eurocode3, reference Chinese specification, the IIW specification,increased Δ influence factors φ. A modified equation is developed for the jointsthrough one-variable nonlinear regression analyses. The modified equations areproved to be more accurate and reliable by comparing with the results of parameteranalysis.