ON THE STATIC STRENGTH OF TUBULAR T-JOINT AND Y-JOINT UNDER COMBINED BENDING AND AXIAL LOADING

Sixty-three tin-lead alloy models of T and Y joints were tested under simple or combined loading. Combined loadings consisted of compression or tension with pure bending in three different planes (IPB, OOPB and an intermediate direction) applied to the brace. They are equivalent to eccentric or offset axial forces. Simple loadings were each of thes e loads applied separately, All models had the same shape (beta = 0.5, gamma = 18.4, tau = 0.96 and theta = 45 degrees for the Y joints). Pr oportional loading was used, and good repeatability of the non-dimensi onal axial and bending strengths of nominally identical models was ach ieved. For the particular geometry used for the tests, it was found th at all models under tension were about 2.3 times as strong as under co rresponding compression; the strengths of the models were based on the peak load obtained in the tests. In bending, Y joints may be slightly stronger when the brace is bent towards the chord than when it is ben t away from it. The use of only the component of load perpendicular to the chord axis as a basis for comparing data leads to inconsistent re sults for Y joints. If the greater strength of Y joints is attributed only to the longer footprint of the brace on the chord, the prediction of the strength of Y joints based on T joint results is safe. The par ametric equations in the Department of Energy's Guidance Notes predict the strength of T joints very well but the strength of Y joints only approximately. There is good agreement with relevant steel model test results. Dividing the failure load of a joint by the maximum strength of the brace (based on the brace cross-sectional dimensions and the ul timate tensile strength of the tin-lead alloy) under the same loading mode gives a ratio, called USRR, which is independent of arbitrary cho ices of normalizing dimensions and allows true comparisons of the stre ngth under different loading modes. For the particular geometry used f or the tests, this gave a simple expression for the minimum strength u nder combined loading. A comparison of the failure loads of Y joints w ith the nominal tensile strength of the braces showed why some of the Y joints failed remote from the tube intersection.