In the design of plating subject to lateral loading, the principal load effect to be considered is the amount of permanent set, that is, the maximum permanent deflection in the center of each panel of plating bounded by the stiffeners and the crossbeams. The present paper is complementary to a previous paper [1]2 which dealt with uniform pressure loads. It first shows that for design purposes there are two types of concentrated loads, depending on the number of different locations in which they can occur; single location or multiple location. The hypothesis is then made that for multiple-location loads the eventual and stationary pattern of plasticity which is developed in the plating is very similar to that for uniform pressure loads, and hence the value of permanent set may be obtained by using the same formula as for uniform pressure loads, with a load parameter Q that is some multiple r of the load parameter for the concentrated load: 0 = rQP. The value of r is a function of the degree of concentration of the load and is almost independent of plate slenderness and aspect ratio. The general mathematical character of this function is established from first principles and from an analysis of the permanent set caused by a multiple-location point load acting on a long plate. The results of this theoretical analysis provide good support for the hypothesis, as do also the relatively limited experimental data which are available. The theory and the experimental data are combined to obtain a simple mathematical expression for r. A more precise expression can be obtained after further experiments have been performed with more highly concentrated loads. Single-location loads produce a different pattern of plasticity and require a different approach. A suitable design formula is developed herein by performing regression analysis on the data from a set of experiments performed with such loads. Both methods presented herein, one for multiple-location loads and the other for single-location loads, are valid for small and moderate values of permanent set and can be used for all static and quasistatic loads. Dynamic loads and applications involving large amounts of permanent set require formulas based on rigid-plastic theory. Such formulas are available for uniform pressure loads and were quoted in reference [1]. A formula for single-location loads has recently been derived by Kling [4] and is quoted herein.

Buckling experiments with spherical shell segments subjected to three different axisymmetric loading conditions are described. Initial shape imperfections were measured so that the experimental results could be compared with theoretical predictions for imperfect shells. Relevant theoretical solutions including classical buckling and initial postbuckling are therefore reviewed. Due to the inherent irregular nature of the imperfections, a comparison is not straightforward. It was necessary to define the amplitude of an"equivalent" regular imperfection pattern. Because there is no rigorous justification for this approach, the experiments can neither verify nor invalidate the initial postbuckling theory, but there is no doubt that very important characteristics of the shells were predicted qualitatively correctly. According to classical linearized theory, the buckling stress should be approximately Independent of loading condition, but experiments as well as initial postbuckling analyses revealed that the critical principal stress for imperfect shellsmay well vary by a factor of 2 to 3 depending on the ratio between the principal stresses. Experimental results for a spherical shell subjected to an inward point-load were found to be in very good agreement with theoretical large-deflection analysis. Bifurcation from the axisymmetric condition associated with small loads was observed. As predicted by initial postbuckling theory, the bifurcations were found to be stable.