|
|
||
|
|
|
|
This and the next several images demonstrate the effect of welding residual stresses and residual displacements on buckling of certain structures.
This image is from:
Tests performed at the David Taylor Model Basin many years ago, described briefly in the 1981 AIAA "Pitfalls" paper by D. Bushnell. This is Fig. 5.10 from that paper.
Fig. 5.10 Buckling patterns in ring-stiffened cylindrical shells. Left side = buckled cold-bent and welded specimen BR-4, for which the critical pressure was 390 psi. Right side = the machined specimen BR-4A with the same dimensions and material properties, for which the critical pressure was 540 psi.
This figure is from Wenk, E., Jr., "Pressure vessel analysis of submarine hulls", Welding Research Supplement, June 1961, p. 277
The difference in critical pressures in the two specimens is caused by:
1. residual stresses caused by cold bending, springback, and subsequent welding of the rings to the cold-bent shell.
2. Axisymmetric initial imperfections caused by welding the external rings to the cold-bent shell.
FROM:
M.E. Lunchick and J.A. Overby, “Yield strength of machined ring-stiffened cylindrical shell under hydrostatic pressure”, Experimental Mechanics, Vol. 1, No. 6, pp 178-185, June 1961
ABSTRACT: The effects of initial imperfections and residual welding and rolling stresses on the yield strength of a stiffened cylinder were investigated by tests of a machined and stress-relieved model, Model BR-7M, identical in geometry and of the same mateerial as a previously tested fabricated model, Model BR-7. The experimental collapse pressure of 1502 psi agreed well with collapse pressures computed from theories which account for the plastic reserve strength. The pressure at which yielding began agreed more closely with that calculated by the maximum-shear-stress or the Hencky-Von Mises criterion than with that calculated by the maximum principal-stress criterion. The data also indicated that the mathematical form of the deflection function of the shell did not change appreciably in the elastic-plastic range. Comparison of the collapse pressures of the machined and welded models indicated that, for the geometry tested, the residual welding and rolling stresses do not adversely affect the collapse pressure.
Page 185 / 227