This report is a summary of local buckling tests of plate elements in square columns built-up by welding. The experiments were conducted to verify theories for the elastic and elastic-plastic buckling of plates with emphasis on the effect of residual stress. This was part of a general study on the strength of welded columns and the influence of residual stress on plate buckling $ Both ASTM A7 and A5l4 steels were used. The square section simulated plates simply supported at the unloaded edges, and the length of the column was chosen so that end conditions had no effect either on the residual stress distribution or on the local buckling strength of the columns e Short columns were tested in the nas-placed" condition in a mechanical-type testing machine. The transverse deflection (local buckling) of the plates was measured at a number of cross sections by a 1/10,000 inch dial gage fixed to a frame held manually. The "top of the knee" method was used to estimate the bifurcation load. The experimental results showed good correlation with theoretical predictions including the effect of residual stress for elastic buckling and for elastic-plastic buckling based on the total strain theoryo The results of experiments indicated that considerable postbuckling strength may be expected for elastic buckling of plates, although not for elastic-plastic bucklinge-1

No abstract

The torsional buckling strength of axially loaded H and cruciform columns is studied with particular attention given to the effect of residual stress. A. numerical approach was used in evaluating the torsional buckling strength so that columns with various patterns of residual stress distribution could be analysed. A series of five welded built-up cruciform columns of constructional alloy steel have been tested and compared with theoretical results.The results of this study indicate that the widththickness ratio of an outstanding flange made of constructional alloy steel should be limited to 7.5 in order to avoid premature torsional failure.

No abstract

INTRODUCTIONAn elastic finite displacement theory of the so-called curved Timoshenko beam with shear deformation has been studied by Reissner, in which the derivation of kinematic field applicable to one dimensional beam theory was a major subject of investigation. Sheinman has also studied the same subjectz. The governing equation derived in his study, howver, is applicable only for a range of moderately small rotation and in addition inconsistency seems to be present in the treatment of higher order terms of small quantities3. Recently, the governing equation of Timoshenko straight beams in finite deformation which is identical to that by: Reissner has been formulated by Iwakuma et al4.For discrete system, a total Lagrangian nonlinear formulation of elastic trusses has been made by Nishino et al5. In their study, the stiffness equation was described as the relation between the overall nodal forces and positions, and was solved numerically by the successive substitution procedure, which was proved to be the same as Newton-Raphson iteration due to the equivalence between the total and tangential stiffness matrices of the same system. This study presents a finite displacement theory of plane elastic beams in the range of small strain problems, and it can be divided into two parts. The first part is concerned with the formulation of a finite displacement theory of arbitrary plane curved Timoshenko beams, in which no limitation is imposed for the magnitude of displacements and rotations. It is emphasized that employing physical stress and strain components rather than tensor components simplifies the expression of constitutive equation and hence the resulting governing equation. The second part covers a total Lagrangian formulation of a plane straight * Student Member of JSCE, M. Eng., Graduate Student,