An experimental study of the bucking of cylinders under axial compression

The serious discrepancy between theoretical and experimental results for the buckling load of axially compressed cylindrical shells is well known. This paper presents results of a test program specifically designed to throw some light on the causes of the discrepancy. Ten nominally identical nickel cylinders were fabricated by electrodeposition and tested under both dead-weight and controlled end-shortening conditions. The test results indicate the following: the buckling load under usual laboratory conditions depends on the properties of the test specimen itself rather than on the testing environment, the buckling loads under dead-weight and controlled end-shortening conditions are the same, a sharp puff of air during loading can cause buckling at a considerably reduced load level, minimization of initial imperfections in the test specimens greatly increases the buckling load, and the magnitude of the minimum postbuckling equilibrium load is relatively insensitive to initial imperfections.