Effect of Imperfections on Buckling of Thin Cylinders Under External Pressure

1956 ◽  
Vol 23 (4) ◽  
pp. 569-575
Author(s):  
L. H. Donnell

Abstract The buckling of thin cylinders under external pressure is studied by finite-deflection theory, assuming the same type of imperfections as had been assumed in a previous study of buckling under axial compression (3). Unlike the axial compression case only a small elastic postbuckling reduction in resistance is found, and this only for long cylinders. However, if reasonable imperfections are assumed, failure loads initiated by yielding are found to be of the same order as those indicated by experiments.

2012 ◽  
Vol 135 (1) ◽  
Author(s):  
J. Błachut ◽  
A. Muc ◽  
J. Ryś

The paper provides details about tests on six steel cones. Test models were machined from 250 mm diameter billet. All cones had substantial and integral top and bottom flanges in order to secure well defined boundary conditions. Experimental data were obtained for: (i) two cones subjected to axial compression, (ii) two cones subjected to external pressure, and (iii) the remaining two models subjected to combined action of external pressure and axial compression. Apart from axisymmetric modeling of tested cones, true geometry with true wall thickness was also used in calculations. Theoretical failure loads were obtained for: (i) elastic perfectly plastic, (ii) engineering stress–strain, and (iii) true stress–true strain modeling of steel. The latter approach coupled with measured geometry and wall thickness secured safe predictions of the collapse loads in all cases. Comparisons of experimental collapse loads with estimates given by ASME and ECCS design codes are included. It is seen here that the ASME and ECCS rules provide a safety margin of about 100% against the collapse (except 50% for axial compression in the case of the ECCS).


Author(s):  
J. Blachut ◽  
A. Muc ◽  
J. Rys´

The paper provides details about buckling tests on six steel cones and the corresponding numerical estimates of failure load (asymmetric bifurcation and/or collapse). Test models were machined from 250 mm billet. The wall thickness was 2 mm, small-end radius was 74.0 mm and the large radius end was 100 mm. The semi-cone angle was 14 deg. Cones had substantial, and integral top and bottom flanges. Experimental failure loads were obtained for: (i) the first two cones subjected to axial compression, (ii) subsequent two cones subjected to external pressure, and (iii) the remaining two models subjected to combined action of external pressure and axial compression. The magnitude of test pressure was about 5 MPa, and the axial failure load was approximately 230 kN. Good repeatability of experimental failure loads was obtained. Numerical estimates of failure loads were obtained for elastic perfectly plastic, engineering stress-strain, and true stress–true strain modelling of steel. Apart from axisymmetric modelling of shells, true geometry with true wall thickness distribution was adopted in calculations. Some of the numerical estimates of buckling loads are close to test data but other are not. The reasons for these discrepancies are highlighted in the paper.


1974 ◽  
Vol 96 (4) ◽  
pp. 1322-1327
Author(s):  
Shun Cheng ◽  
C. K. Chang

The buckling problem of circular cylindrical shells under axial compression, external pressure, and torsion is investigated using a displacement function φ. A governing differential equation for the stability of thin cylindrical shells under combined loading of axial compression, external pressure, and torsion is derived. A method for the solutions of this equation is also presented. The advantage in using the present equation over the customary three differential equations for displacements is that only one trial solution is needed in solving the buckling problems as shown in the paper. Four possible combinations of boundary conditions for a simply supported edge are treated. The case of a cylinder under axial compression is carried out in detail. For two types of simple supported boundary conditions, SS1 and SS2, the minimum critical axial buckling stress is found to be 43.5 percent of the well-known classical value Eh/R3(1−ν2) against the 50 percent of the classical value presently known.


Sign in / Sign up

Export Citation Format

Share Document