抄録
Past test results indicated that tubular X and T-joints under tensile brace loading sustained greater deformation than the same joints under compression, without losing the overall stability. It has been proposed in the past by several investigators to calculate the ultimate strengths of these joints under tension only by multiplying the capacities of the same joints under compression by a constant value greater than 1.0. Additional test results recently obtained, however, evidenced that the capacity of the joints in tension did not vary propotionally with the capacity in compression. Further, the observed ultimate strengths in tension were subjected to greater than those in compression, because the ultimate strengths were governed by separation of a tension brace from the chord and, therefore, were influenced by welding conditions, weld defects, material propeties, and so forth. This paper presents presents results of analysis of all the test results available for tubular X and T-joints under tension. Comparison is also made with the capacity of simple model rings, on which the ultimate strength formulae are based. The following conclusions are drawn : (1) The nonlinear finite element analysis is found to interpret well responses of model rings to concentrated brace forces. Many similarities are found between the behaviors of tubular X-joints and model rings. (2) The ultimate and the yield strengths of tubular joints are defined basing them on the load-deformation curves. It is assumed that the yield strength can be used as a measure of a serviceability limit state. (3) A multiple regression analysis is carried out to obtain new prediction equations for both the ultimate and the yield strengths of the X and T-joints under tension. Greater scatter is observed in these equations as compared with the joints under compression. (4) The resistance factors for serviceability and ultimate limit states are calculated using simple LRFD formats. The serviceability limit state can be more critical than the ultimate limit state, although this statement varies with the probability distributions of loads.
