Structural dynamics is a key discipline in aircraft design and certification. The realm of structural dynamics includes problems in which structure flexibility is important, such as dynamic loads (landing, taxi, gust, turbulence...), but it also includes other areas like environmental vibration or impacts technology. This is closely related with its sister discipline Aeroelasticity.

Linearity has been the cornerstone assumption in aircraft design engineering solutions. In the past, linearity allowed to solve many structural dynamics problems that otherwise would have been unaffordable. Nowadays, the necessity of highly optimised structures, combined to an increase of computer power, has made this assumption to be reconsidered.

This paper presents a wide survey on non-linear topics in structural dynamics and the way they have been solved at EADS-CASA (Airbus military) in the past 25 years. Although the paper has been mainly focused on numerical simulation using the Finite Element Method technique, component tests and full aircraft tests are also presented.

The hot-wire anemometer is one of the few instruments which can be used to make velocity measurements in turbulent and unsteady flows. However, the probe supporting the wire inevitably interferes with the local flow and it has been found that the effect of this interference on the reading of the anemometer varies considerably as the orientation of the probe to the flow direction is changed (the wire itself being maintained in the same direction). This leads to errors in any measurements taken where the instantaneous local flow direction differs significantly at any time from the direction for which the anemometer was calibrated. Such errors are quite separate from, and in addition to, errors due to finite wire length, incidence of the wire to the local stream direction, etc.

Many aspects of parachute behaviour are dominated by the permeability of the canopy cloth. Techniques employing kinematic principles to predict parachute opening behaviour, for example, must first be able to predict the air flow through the canopy material. This paper is concerned with the development of such a predictive capability.

The flow of air through cloth of gauze has been studied by many workers; since the 1850’s, in connection with filters; and since the early twenties, in connection with parachutes. Workers in each of these disciplines seemed unaware of the work being done in the other.

In the case of parachute cloth permeability we find the foundations being laid down by such eminent workers as Glauert in 1932 and Taylor in 1944. Yet most subsequent writers seem not to be aware of this, and since World War II, there have been many papers “rediscovering” basic principles; often with errors which could have been corrected by reference to the earlier authorities.

In this paper we attempt to bring together all the work which has been done in this field, and to rationalize the data by simple mathematical modelling. Individual elements of the model have been proposed before, of course.

We then examine the effect of tension on the geometric porosity of fabric, and hence the pressure loss Ap. A theoretical analysis shows that tension is likely to be a major factor. For cloths which have low porosity when unloaded our simple mathematical model shows that, for a given Ap, a stress equal to 50% of ultimate can increase the air volume flow Q by an order of magnitude. In general, the increase in permeability is greatest at the lower values of Ap, indicating that the viscous term is more influenced by tension than the dynamic term. This theoretical result is confirmed by some rather limited experiments carried out by Lashbrook and Marby. But their data also shows that a cloth made from relatively stiff glass fibers can experience a reduction in permeability when loaded, due to a “closing up” of the weave. The reasons for this are discussed.

Experimental work is described showing that a major cause of changes in the calibrations of hot-wire anemometers consists in the precipitation of dust on the upstream surface of the wire during use. A simple means of cleaning the air in closed-circuit wind tunnels is described. The rate of change of calibration due to dust was reduced in the Manchester University Fluid Motion Laboratory by at least a factor of 15, by adequate cleaning of the air and by minimising the amount of interchange between the tunnel air and the atmosphere. It is concluded that where accurate results are required from hot wires used in a wind tunnel, the tunnel should be designed to be dust-free.

The flow and noise created by sawtooth trailing-edge serrations has been studied experimentally at a low Reynolds number. Experiments have been performed on a flat-plate model with an elliptical leading edge and an asymmetrically bevelled trailing edge at Reynolds numbers of Rec = 1 × 105–1.3 × 105, based on chord. Wide serrations with a wavelength (λs) to amplitude (2h) ratio of λs/h = 0.6 were found to reduce the overall sound pressure level by up to 11dB. In contrast, narrower serrations with λs/h = 0.2 produce tonal noise and increase the overall noise level by up to 4dB. Intense vortices across the span of the trailing edge with narrow serrations are shown to be the source of tonal noise. Wide serrations reduce turbulent velocity fluctuations at low frequencies which explains the lower radiated noise. The narrow serrations that produce low Reynolds number tonal noise were shown previously to be effective at higher Reynolds numbers (Rec > 2 × 105), demonstrating that care is needed to fully understand the flow field over serrations for all intended operating conditions.

Incompressible turbulent flow in smooth pipe bends of circular-arc curvature and constant cross section is discussed. The effects of bend angle, radius ratio, duct cross-sectional shape and Reynolds number are considered. Particular attention is paid to the influence of the tangents on the flow and losses in a bend. In addition consideration is given to the definition and measurement of the bend loss.

A correlation of experimental pressure loss data is made. It is shown that large variations in experimental results, which cannot be ascribed to the effects of geometrical parameters or Reynolds number, are accounted for by the effects of the downstream tangent and position at which the pressure measurements were made. The inadequacy of existing experimental information is an obstacle to the presentation of comprehensive data on the effect of geometrical parameters on bend losses.

This paper briefly reviews the progress made with the Imperial College hypersonic gun tunnel since August 1958 when it came into operation. The tunnel and instrumentation are described together with the results of various measurements made both in the barrel and in the working section. More detailed reports will be published later.

The discrete element method is modified so as to simulate the movement of passengers evacuating an aircraft. The results display reasonable agreement with timed trials. However, by focusing on the 1985 Manchester air disaster where only some of the exits could be opened and only then a considerable time after landing, it is argued that the certification procedures are inadequate.

This paper presents an approach to the determination of structural member sizes to provide minimum weight under a number of load conditions consistent with specified limitations on stress and displacement. The procedure amalgamates the disciplines of matrix displacement analysis and operations research into a scheme suitable for automatic digital computation.

A measurement of the pressure distributions on D-section prisms over wide ranges of depth to height ratio and incidence in a smooth and in a turbulent flow was made with particular reference to the onset of transverse galloping. Good correlation between the onset of galloping and the critical depth, established already for rectangular and the related prisms with sharp downstream corners, applied also to D-section prisms both in a smooth and in a turbulent flow. The development of a reattachment type pressure distribution on the side face of the prism as the critical depth is reached is responsible for this correlation. There is very little effect of turbulence on the pressure distribution and the Strouhal number of a D-section prism with a depth to height ratio of 0.5.

Recently attention has been given to the effects of the stream turbulence on the characteristics of the turbulent boundary layers and the turbulent shear flows. For the boundary layers considerable information on the effects of the stream turbulence is available but for the shear flows it is limited. This paper describes an experimental investigation for the effects of stream turbulence on the development of an asymmetric jet and a plane mixing layer. It is shown that the characteristics of these flows are considerably affected by the stream turbulence but for the measurements in the plane mixing layers at large downstream distances the influence of the stream turbulence is negligible provided the intensities are less than 0.6%. For the plane mixing layer the results show that the flow characteristics are not affected significantly by the initial boundary layers, the acoustic disturbances and the boundary conditions imposed by the end plates.

This paper is one of a series addressing the need for simple, yet accurate, methods for the estimation of cruise fuel burn and other important aircraft performance parameters. Here, a previously published, constant Reynolds number model for turbofan-powered, civil transport aircraft is extended to include Reynolds number effects. Provided the variation of temperature with pressure is known, the method is applicable to flight in any atmospheric conditions. For a given aircraft, cruising in a given atmosphere, there is a single Mach number and Flight Level pair, at which the fuel burn per unit distance travelled through the air has an absolute minimum value. Both these quantities depend upon the Reynolds number, which, in turn, depends upon the aircraft weight and the atmospheric vertical temperature profile. Simple, explicit expressions are developed for all parameters at the optimum condition. These are shown to be in close agreement with numerical solutions of the governing equations. It is found that typical operational mass and temperature profile variations can change cruise fuel burn rate by several percent. In the International Standard Atmosphere, when the speed and altitude deviate from their optimum values, the fuel burn penalty is reduced slightly relative to the constant Reynolds number case. By way of example, the method is used to estimate the minimum fuel, speed-versus-height trajectory for cruise in a realistic atmosphere.

For each aircraft, cruise fuel burn is found to be governed by six independent parameters. All are constants. Two are simple, involving only size and weight, whereas four are complex and must be determined by either theoretical, or empirical, means. The estimation of these quantities will be considered in Part 2.

In an axisymmetric turbulent boundary layer along a circular cylinder at constant pressure, measurements have been made of mean velocity profile and turbulence characteristics: longitudinal velocity fluctuations, Reynolds shear stress, transverse correlation and spectrum. It has been found that the qualitative behaviour of an axisymmetric turbulent boundary layer is similar to that of a two-dimensional boundary layer in the wall region, where as in the outer region the effects of transverse curvature are observed.

The results of measurements in the flow along a rectangular corner are presented in the form of velocity profiles. The profiles form two sets: one for flow in a slightly favourable streamwise pressure gradient and one for flow when the pressure gradient is ‘practically zero’. The appearance of the profiles is quite different from that of previously reported experimental work and it is suggested here that the primary cause of this is the different type of corner leading edges used in each case. The flow appears to be only marginally stable and even very slightly changed entry conditions caused by alterations in the leading edge geometry may have a marked effect on the profiles further downstream. The new results are consistent with the notion of corner layer similarity (an implication in every attempted theoretical solution) but there remains a significant difference between the experimental results and the more reliable theoretical solutions available.

The ability of microramps to control shock - boundary layer interaction at the vicinity of an axisymmetric compression corner was investigated computationally in a Mach 4 flow. A cylinder/flare model with a flare angle of 25° was chosen for this study. Height (h) of the microramp device was 22% of the undisturbed boundary layer thickness (δ) obtained at the compression corner location. A single array of these microramps with an inter-device spacing of 7.5h was placed at three different streamwise locations viz. 5δ, 10δ and 15δ (22.7h, 45.41h and 68.12h in terms of the device height) upstream of the corner and the variations in the flowfield characteristics were observed. These devices modified the separation bubble structure noticeably by producing alternate upwash and downwash regions in the boundary layer. Variations in the separation bubble’s length and height were observed along the spanwise (circumferential) direction due to these devices.

The local buckling of structural sections composed of isotropic plate elements is well documented. It is the purpose of this paper to extend this work to include sections composed of orthotropic plates. It has been shown that the buckling behaviour of single orthotropic plates with certain boundary conditions can be related to that of an equivalent isotropic plate in terms of a single parameter involving the principal plate rigidities. It is proposed to investigate whether this principle may also be applied to the structural buckling problem.

It has been shown that simple approximations to the deflected form of buckled plates can be used to give fast and reasonably accurate results for the plate buckling problem using an energy method. If these simple deflection functions are used in an energy analysis for structural sections then large errors may be introduced.

In 1973, automatic flight controls is one of the most exacting fields of technology, requiring a balanced combination of art, science and human understanding, and the efforts of large project management teams backed by adequate resources to bring any new concept to fruition.

Today's systems derive almost entirely from the technology of the past 25 years, and hence it is little known that there was some activity in automatic flight control for fixed wing aircraft as long ago as 1873, some 30 years before the world-changing events at Kittyhawk and when mechanical flight was still confined to balloons and a few rudimentary gliders.

The aerodynamic effectiveness of a cylinder-flare body at zero incidence under laminar and turbulent boundary-layer conditions has been studied experimentally. Two nose geometries, namely a 10° half-angle sharp cone and a hemisphere, were used. The study has been carried out in a hypersonic gun tunnel at a Mach number of 8·2 and a Reynolds number of 158,100, based on the cylinder diameter, at flare angles 0°, 10°, 20°, 30° and 45°. The surface flow was studied using oil-dot and liquid crystal techniques. Some information regarding the shock layer was obtained from schlieren pictures. The effects of entropy layer and boundary-layer state on flare effectiveness were deduced from pressure measurements over the cylinder and the flare.

This paper presents a comparison of linear and non-linear methods for the analysis of aeroelastic behaviour and flutter boundary prediction. The methods in question include NASTRAN and ZAERO, based on linear aerodynamics, and the non-linear coupled CFD-CSD methods RANSMB and PMB, developed at the Universities of Bristol and Glasgow respectively. The test cases used for this comparison are the MDO and AGARD 445.6 weakened wing. In general, it was found that the non-linear methods demonstrate excellent agreement with respect to pressure distributions, deflections, dynamic behaviour, and flutter boundary locations for both cases. This is in contrast to previous studies involving similar methods, where notable differences across the MDO span were found, and is taken to imply good performance of the CFD-CSD interpolation schemes employed here. While the linear methods produce similar flutter boundaries to the coupled codes for the aerodynamically simple AGARD 445.6 wing, results for the transonic ‘rooftop’ MDO wing design did not agree as well.

The “second stage” two-dimensional jet flap experiments are briefly described and some of the results presented with comments. The relative simplicity of these experiments and the measure of their agreement with the theory provides, it is hoped, yet another illustration of the value of comparatively simple experiments in the early stages of work on a novel idea.