Search results

IN R. & M. 1679 and AIRCRAFT ENGINEERING, December, 1935, the author gave certain expressions for the stable loads which a stiffened‐skin panel, flat or curved, was capable of carrying without collapse by buckling. As regards the curved panel, it was there implicitly assumed that waving over relatively long lengths was impossible, because of the presence of rigid bulkheads or similar supports at intervals. While this will normally be the case in practice, it may happen that in some part of the structure a long length of panel remains without support other than that given to it by the ordinary frames and stringers. As this will tend to reduce the buckling load, it seems desirable to complete the theory to include the case of buckling in long waves.

HELICOPTER rotor blades, as is well known, are subjected to alternating bending stresses at various multiples of the rotor r.p.m. when the aircraft is flying forward. One of the factors determining the magnitude of the fluctuating stresses at the various frequencies is the approach to resonance between the harmonic components of the periodically varying aerodynamic loads and the natural blade frequencies. Most rotor blades are flexible enough to allow the frequencies of at any rate the lower modes, which are probably the most important, to be estimated by applying a correction for blade bending stiffness to the natural frequency of the blade if it were perfectly flexible, and subject only to inertia and centrifugal forces. It is on this latter aspect of the problem that the present article is concentrated.

IT is fairly standard practice to correct wind‐tunnel results for a model or aerofoil by taking measurements of lift, etc., with and without dummy supports fitted, the correction for support interference being deduced accordingly. Such experimental correction figures are liable to be rather scattered owing to various causes, and it is useful to have at least some independent check by direct calculation on the amount of interference to be expected.

ORDINARILY, for the purpose of strength calculations, as well as in estimating bending and torsional rigidity with a view to deriving blade deflections and investigating flutter characteristics, the blade is regarded as flat (i.e. as if designed for zero pitch); any effects of the twisted form of the blade in causing departures from the classical bending and torsion theories being regarded as secondary. It has never, however, been proved that they are actually negligible, and an approximate analysis indicates that they may in some cases become appreciable.

IT is fairly well known that for any given wing arrangement—monoplane, biplane, monoplane with end‐plates corresponding with an increasingly popular type of tail unit, etc.—there is a definite lift distribution associated with the minimum value of induced drag of which a particular span is capable. The theory giving the appropriate lift distribution and, what is perhaps of more general practical interest, the resulting induced drag coefficient, is perhaps not so well known and while numerical results for arrangements in common use have been published, novel arrangements require a knowledge of the underlying theory for their evaluation, and this is in any case useful as giving a clear grasp and understanding of the problem.

ONE of the outstanding problems in design of stressed‐skin structures is that of ensuring adequate rigidity to guard against instability failure or buckling of the parts of the structure which have to carry compressive loads. Such structures consist usually of a scries of longitudinal members or stringers with intersecting transverses or frames. Apart from the waving of the skin itself in the panels thus formed, which must usually be tolerated and which remains modcrate in extent so long as the stiffening members themselves do not deflect, and docs not impair the ability of the structure as a whole to continue to take increasing load, there are two possible forms of buckling. The stringers alone may bow between adjacent frames, which remain fixed, or longer waves involving both stringers and frames may occur. The strength to resist the first type of failure can be estimated with reasonable accuracy by treating the stringers as struts of length equal to the frame spacing, and gives a design criterion for the stringers. It is fairly evident that for this rather elementary treatment of the stability problem to be adequate it is necessary that the frames themselves should have a certain minimum rigidity. It is in determining this minimum rigidity that the question arises of the possibility of the second type of buckling, which will be specially considered here. Whether this will, in any practical case, be sufficient to fix the size of frame required, cannot of course be stated in general terms, as this will depend on the local loads which the various frames may have to carry. Fig. 1 shows the type of deformation involved, for the special case of the side of a monocoque fuselage, which may be subjected to compression due to lateral bending under the action of rudder and inertia loads. The mode of waving is specified by the longitudinal wave‐length (two frame spaces in the illustration) and by the transverse wave length, which may be the full width of the panel or something less than this.

IT has been noted by Falkner and others that the ordinary aspect‐ratio corrections based on lifting‐line theory are not as accurate as had always been assumed even for normal aspect ratios of the order of 6. Lifting‐surface theory involves heavy calculation, and the present work is intended to give at least an indication of the magnitude of the finite‐chord correction on a fairly simple basis.

THE simplified method of stressing which follows is an extension of one first developed as a means of attacking the shear‐lag prob‐lem for a ship‐beam (Trans. N.E. Coast Inst, of Engrs., 1924–5). It is now extended to cover also axial stresses due to torsion and departure from the Batho torsion shear, a related subject. Shear‐lag and diffusion are closely connected and a theory which covers one should have at any rate a general application to the other; it is not, how‐ever, claimed for the present method that it is very suitable for use where the rates of diffusion are locally high, and its applicability to diffusion problems in general is a matter of judgment.This of course applies also to any corresponding method. The method is considered first for a girder of uniform transverse dimensions (no taper of depth or width) but with variation in scantlings along its length. Later the first‐order effect of tapered dimensions is considered. There may be longitudinal stiffeners at any point in the section, e.g. spar booms or stringers. The transverse sections of the girder are assumed to retain their dimensions with absolute rigidity.

IN these days of metal construction, the aircraft designer frequently uses the so‐called Batho theory of the torsion of thin shells. This simple and well‐known theory is a ecial case, applicable to tubes and boxes of any cross‐section, of the general St. Venant torsion theory. The same theory, applied to thin “ open ” sections, of which a channel is a mplc example, gives the result that the torsional strength and rigidity are very small. While it is true that such a section tends always to be weak in torsion in comparison with a closed section of similar dimensions, subject to certain conditions as regards fixing of the ends it is capable of transmitting an appreciable torque. The method of transmission can be described as differential bending of the two langes, in the case of the channel. Calculation of differential bending stresses in a two‐spar wing under torsion is a simple procedure, but alien the two members in differential bend form part of a continuous section, the conditions are somewhat altered. The general theory applicable to such cases may be called the theory of torsion‐bending. The results of this theory will be summarised, and the proofs given in the Appendix.

REPORTS from America speak of flight tests of two types of screw‐lift aircraft, the Sikorsky multi‐rotor machine, and a twin‐rotor type. From this it would seem that the day of the practical helicopter, so Ions foreshadowed, is not far distant. The following general notes on design considerations may therefore be sufficiently topical to be of interest.