Application of a method suggested by my technical note on equilateral and semi-equilateral triangles has yielded values for the buckling coefficients for isosceles triangles which in some cases differ appreciably from those given on Data Sheet 02.04.06.

The technical note quoted gives exact solutions in the form of a terminating trigonometrical series. When the triangle becomes isosceles, the series no longer terminates, but the infinite series solution, on insertion of the boundary conditions, gives an infinite determinant for the buckling load or eigenvalue.

Evaluation of the determinant, or of a sufficient number of terms, was computed on a Univac 1107, with the following results, which have been converted to the notation of Data Sheet 02.04.06.

In recent years there has been an increasing need for data on the rates of fatigue crack propagation in alloys of interest to the aircraft industry. In general alloys in which cracks grow slowly, under given stress conditions, have an obvious advantage over those in which cracks grow faster; there is more time to detect a crack before failure. It is possible that, in the future, quantitative crack propagation data may be used to give better estimates of the safe life of a structure. There is, however, a need to rationalise the presentation of such data and to study the effect of such parameters as specimen geometry and stress level. Some experiments to investigate the effect of varying some of these parameters in one clad aluminium alloy, DTD 5070A are described. It is a summary of work contained in two RAE Technical Reports.

The problem of the prediction of wall and fluid temperature with a turbulent boundary layer can appear in two forms:

1. (a) the prediction of wall heat transfer with specified wall temperature;

2. (b) the prediction of wall and fluid temperature with specified wall heat flux.

Because the differential equations governing heat flow in the turbulent boundary layer are linear, it is sufficient to be able to predict theoretically either the heat transfer rate consequent upon a step function of temperature or the temperature distribution following a step function of wall heat flow rate. Such predictions will then permit the treatment of the more general situations (a) and (b) by superposition.

There are numerous examples in the aero and power producing industries of components which are designed to sustain creep deformation. An important class of structures in this category lies in pressurised shells of revolution. The few analyses which allow predictions of creep deformations to be made are usually restricted to the secondary creep problem, contain approximate relations between stress resultants and deformations or ignore the important effects of bending. While the neglect ot primary creep is not important for cases of steady loading sustained for long periods, stress redistribution should be calculated if the loading period is less than the time taken to reach a stationary state of stress; such a state of affairs is very likely in two-shift operated power producing plant and also in aircraft components. Approximations involving concepts of “sandwich” shells that are often made to form moment-curvature rate relations can be dispensed with by integrating the stress-strain rate relations through the shell thickness. Although the stresses become non-linear through the shell wall they are slowly varying and the integration introduces no problem. The integration requires the use of a digital computer but creep problems have to be solved numerically, except in a very few cases, so that whatever the assumptions automatic computation is necessary. While some practical shells approximate to the momentless condition the most likely case will include bending which, although likely to be relaxed by a process of stress redistribution, will cause additional creep straining.

The purpose of this paper is to summarise some recent results obtained at Standford University regarding the buckling of thin shells. A detailed review of work done by other investigators will not be given here because this was done recently in two papers written by the author. However, a few important steps in the evolution of our knowledge of the buckling of thin .shells will be mentioned to put the new results in proper perspective.