For some years the author has been interested in subsonic ram-jets as a means of helicopter rotor tip propulsion and has done a considerable amount of experimental work with this end in view. Most of the effort has been concentrated on producing engines to operate at the lower end of the speed range as it was thought that, however simple a ram-jet may be, this advantage is lost when other means of propulsion are needed for starting purposes.

An experimental verification of coles’ postulate of the Law of the Wake in three dimensions has been attempted. This hypothesis, which is strongly supported by experimental evidence in two dimensions, purports to represent the velocity vector in a yawed boundary layer by the sum of two vectors, whose magnitudes are found from the universal law of the wall, and a second universal function, the law of the wake. It has been found that the velocity vector can be represented in this way and the wake function required for this is very similar to Coles’ form. Limitations of the experimental arrangement however prohibited the comparison of vector directions with that proposed by Coles.

I read Lt. Cdr. Whitehead’s article in the April Journal (p. 183) with considerable interest. As a summary of current problems I found it most illuminating. In my opinion however the author has rather over-simplified the mirror calculations and I feel that it may be worthwhile recording in slightly greater detail the basis of mirror calculations and the magnitude of the errors implicit in them.

I would like to add yet another note to those already published on the notions of viscous and hysteretic damping in a simple oscillator. In particular, I would like to hammer a nail into the coffin of the so-called “complex stiffness.“

One of the features which hysteretic damping may be defined to possess is the property that the energy loss per cycle (in sinusoidal motion) is independent of the frequency of oscillation, ω. This property is derived from some experimental evidence. It is shown in Reference 1 that this behaviour is reproduced by the particular integral of the equation

The Matrix Force Method is frequently used in the stress analysis of complicated aircraft structures. It assumes as unknown quantities, or redundancies, groups of internal loads, and yields stresses and displacements in terms of a set of unit loads. Lately, this method has been extended to include effects of uneven temperature distribution. With large structures, the method is most effective with high speed digital techniques which allow considerable computational detail to be stored in the computer memory. Theoretically, the method is based on the Castigliano theorem of least work, equivalent formulations being possible. Hence the method is applicable to elastic problems of the first order only, i.e. strains infinitesimal, structural distortions not affecting equilibrium conditions markedly.

The flow through porous screens has been widely studied from both the theoretical and experimental points of view. The most widely used types of screen are the wire mesh and the perforated plate, and the majority of the literature has been concerned with the former. Several attempts have been made to correlate the parameters governing the flow through such screens, i.e. the pressure drop, the flow conditions and the geometry of the mesh.

Expressions are found and curves given for the bending moment, direct force and shear force around a circular frame, in a cylinder partially filled with fluid. The maximum bending moment is found to occur when the cylinder is half full, and its position to be at the level of the fluid surface.

The analysis is an elementary one, that is, engineer's bending theory is assumed to hold in the cylinder. Hence the resultant pressure of fluid on the frame is balanced by a sinusoidal varying shear flow around the frame.

In a recent article, H. Werlé, has described how the free spiral vortices on delta wings with leading edge separation suddenly expand if the incidence is increased beyond a critical value. His description conforms to a great extent with the results, arrived at during an English investigation of the same phenomenon (called the vortex breakdown), but the interpretations of the observations, suggested by the two sources, are different. Against this background it is felt that some further comments and some pertinent high speed observations, may be of interest.

The problem of flow past a thin wing of circular plan form when the undisturbed flow is along the x-axis, U(z)=Az+V has been solved.

The solutions for potential flow past a wing of circular plan form have been given in terms of spheroidal co-ordinates using the concept of acceleration potential. The solutions can be easily adapted to the case when the undisturbed flow is weakly sheared. In the particular case when the undisturbed velocity U(z) = Az+V, where A and V are constants, the pressure equation can be simplified by a simple substitution and the solution for a thin wing of circular plan form can be easily constructed from the solution for potential flow. If the wings in the two cases develop the same lift, it is found that the wing in shear flow should be slightly cambered (potential wing is un-cambered). The results are shown graphically in Fig. 1.

Some measurements in the R.A.E. Bedford 3 ft. wind tunnel on a typical high tailplane illustrate a Mach number effect on the tailplane rolling moments due to sideslip which has been apparently overlooked previously. Fig. 1(a) shows that the increase in tailplane rolling moment due to sideslip is much larger than the increase in the fin lift curve slope a1F, Fig. 1(b). This extra increase in rolling moment is associated with the increase in the tailplane lift curve slope a1T, Fig. 1(c). In fact the rolling moment on this high tailplane varies roughly as the product of the fin and tailplane lifts, as Fig. 1(d) shows. A qualitative explanation follows.

Consider the idealised high tailplane shown in Fig. 2. At zero geometric incidence and positive sideslip the fin lift will induce positive and negative incidence distributions on the starboard and port sides of the tailplane.

Although many investigators have set out to find a method of determining the buckling and failing stresses for unstiffened shear webs with flanged or unflanged lightening holes, it seems that so far no really satisfactory method has been found.

A number of empirical and semi-empirical expressions have appeared over the years which agree closely with the limited series of tests on which they are based, but which the writer has found not to agree with the general mass of experimental data which is now available on this subject.

Following a recent examination of the available data another attempt has been made to derive some expressions or formulae that would help the designer to estimate buckling and failing stresses of such panels, and as a first step a semi-empirical formula has been derived that appears to fit the test results reasonably well for the initial buckling case. It applies to panels with edges fixed.

If a circular plate has a concentric circular hot area, there is a critical temperature for this area at which the plate buckles. This temperature is calculated in this note for the case of a clamped plate supported in such a way that the radial stress in the cold part obeys the inverse square law.

The Rayleigh-Ritz method is well known as an approximate method of determining the natural frequencies of a conservative system, using a constrained deflection form. On the other hand, if a general deflection form (i.e. an unconstrained form) is used, the method provides a theoretically exact solution. An unconstrained form may be obtained by expressing the deflection as an expansion in terms of a suitable set of orthogonal functions, and in selecting such a set, it is convenient to use the known normal modes of a suitably chosen “ basic system.” The given system, whose vibration properties are to be determined, can then be regarded as a “ modified system,” which is derived from the basic system by a variation of mass and elasticity. A similar procedure has been applied to systems with a finite number of degrees of freedom. In the present note the method is applied to simple non-uniform beams, and to beams with added masses and constraints. A concise general solution is obtained, and an iteration process of obtaining a numerical solution is described.