When a flow with a non-uniform velocity passes through a cascade, the blades impose on each layer of the flow local pressure changes which are roughly proportional to the dynamic head in the layer. This results in local pressure differences which, within the blade passages, produce rotational secondary flows and which also cause overall spanwise accelerations and displacements. These effects are analysed for frictionless flow by treating the blade row as an actuator plane whose outlet angle is determined by the secondary velocities in the blade passages.

At low outlet angles or large deflections the secondary flows effectively control the downstream flow angles. When the outlet angle is large the displacement effect can result in downstream flow angles of opposite sign to those induced by the secondary flow alone.

Measurements on cascades with small spanwise variations induced in the approaching stream confirm the main predictions of the first order theory.

The correct design of intermediate vertical stiffeners on web plates subjected to shear becomes very important when the web plates are designed to operate at loads close to their buckling loads. This paper presents details of an extensive series of tests conducted on stiffened web plates subjected to shear. From the analysis of the results obtained from these tests, new empirical relationships between the flexural rigidity and spacing of the intermediate stiffeners and the buckling stress of the stiffened web plate have been obtained.

One interesting and important feature of these new relationships is that they define more clearly than hitherto the difference in the behaviour of single-and double-sided stiffeners.

The introduction of an interference-fit pin in a circular hole in a flat plate is found to result in a reduction in the stress concentration factor at the hole boundary under tension applied to the plate. The S.C.F. decreases rapidly with increase in the ratio of interference stress to applied tension, and decreases also with increase in the ratio of hole diameter to width of plate. Very little difference in the S.C.F. is found for different ratios of Young's modulus of pin and plate.

The climb performance of a turbo-jet transport aircraft is considered, using the concept of the energy height. The analytical formulae used in evaluating the time needed, distance covered and fuel consumed during climb under various conditions are developed. It is concluded that the application of this energy approach method to the climb performance estimation of a turbojet transport aircraft will prove satisfactory from the points of view both of the accuracy of the results and of saving some of the considerable time needed for the climb calculation by the conventional method. A numerical example is given to illustrate the usefulness, applicability and flexibility of this method.

Using Rayleigh's principle, the natural frequencies of thin isotropic rhombic plates, with three possible combinations of boundary conditions obtained by combining clamped-clamped and clamped-supported edge conditions, are determined in Part I of this paper. To introduce constant limits of integration, non-orthogonal co-ordinate systems are used and the wave shape for the vibrating plate is approximated by using normal functions representing mode shapes of corresponding bars. To estimate the accuracy of these eigenvalues, Kato's theorem is used and the lower bounds for the natural frequencies are determined in Part II of the paper. It is also shown that normal beam functions are not generally suitable for the determination of eigenfrequencies of skew plates with large skew angles.

In this paper (published in The Aeronautical Quarterly, Vol. VII, Part 2, p. 156, May 1956), a set of equations is derived giving the responses at the generalised coordinates q1, q2, … qn which are caused by a harmonic generalised force Φseiωt corresponding to qs. They are equations (26), namely

This note is concerned with a sentence which follows these equations which says (of them) that “ … there is no longer any assurance that the ratios between the constants mBrs in any column of partial fractions are the same for all co-ordinates qs at which excitation is applied, since the previous limiting condition cannot now be applied. Thus the shapes of the ‘ modes ’ which are associated with the various columns depend upon the nature of the applied harmonic excitation.”