There are many examples in his two Alice books where Lewis Carroll, perhaps unwittingly, provides analogies with the world of reality. The lines quoted above are often used to portray the rapidly changing scenario in which we struggle to maintain the status quo of our social and political lives. The same is true in the world of structures: the scenario continually shifts as we discover new hazards which threaten our products. Aircraft speeds have increased, operators have demanded longer service lives, new materials have been introduced with their own susceptibilities, and so we must develop more comprehensive design requirements and better design techniques to ensure that each new aircraft maintains the existing standards of airworthiness—or at least the standards we imagine to exist.

Twenty years ago, the electronic valve gave way to the transistor; now the transistor as a discrete component has been replaced by the integrated circuit, and micro-processors are emerging as cheap, reliable and basic elements of electronic architecture.

Each step has yielded a major reduction in the size of equipment needed to achieve a particular function or to improve performance—but in spite of miniaturisation, an increasing volume of avionics has been incorporated into succeeding generations of aircraft as requirements for greater precision and safety have arisen.

Application has already provided striking improvements in the quality and ease of communication and navigation, which in turn have brought about the demise of the dedicated radio operator and navigator. The trend will continue with more automation of flight control and system management, leading to two-crew operation of most short and medium range aircraft—providing, of course, that the cost of ownership of additional equipment is less than the costs of the displaced crew member, and that flight safety standards continue to improve.

Although hundreds of moving base simulators exist in the world today, there is a paucity of objective and quantitative methods of assessing the necessity for, and the effectiveness of the motion provided. This paper describes a quantitative method of assessing the effects of motion on an aircraft simulation when there is short-term interactive control of the aircraft by the pilot.

The integration of a system—or a group of systems — into a modern fixed-wing aircraft can only be satisfactorily achieved if the fuel penalty incurred by the aircraft is minimised, i.e. if the amount of fuel needed to restore the initial aircraft performance is kept to a minimum.

Over the years, aircraft systems have had to meet more and more stringent requirements which reflect not only on the increasing performance of aircraft, but also on the greater demands made by systems on the aircraft, a somewhat extreme example being that of the Mach 2 + Rockwell International B-l bomber on which the cooling requirements for the avionic equipment were to exceed several hundred kilowatts.

As such trends are likely to continue in the foreseeable future, the current deteriorating situation regarding fuel supplies and the ever rising costs involved, stresses the need for ensuring that systems on the next generation aircraft are designed for the least penalty and most efficient use of the available fuel.

The honeycomb sandwich type of construction in plate and shell shapes is playing an increasing role in aerospace structures where structural efficiency is important. In some applications, such as wing or control surface trailing edges, the core is of non-constant thickness, and in others, such as engine thrust reverser buckets, the shell shape is doubly curved. Faceplates are commonly manufactured from aluminium, stainless steel or other metals, but the use of anisotropic materials, such as carbon fibre reinforced plastics (CFRP) can be expected to increase in the future.

In finite element analysis, as in analytical work, sandwich plates and shells have received far less attention than thin plates and shells. However, elements have been developed specifically for sandwich plates by Barnard, Cook and Bartelds and Ottens. Of these, only the latter allows a non-constant core thickness. Evidently, such elements could be used in a facet shell representation. Only two general (as opposed to axisymmetric) curved elements specifically formulated to deal with sandwich shells seem to have appeared in the literature. Both are subject to deficiencies as noted for thin shells by Morris, and Webber discusses certain geometrical errors in the doubly curved element of Reference 5. Furthermore, both are of limited practical application. It may be concluded, therefore, that there is a need for a honeycomb sandwich shell element which is easy to use and accurate while being capable of modelling the full range of materials and shapes found in practical structures. To this end, a family of elements for use in the linear elastic finite element method is formulated. They are of general doubly curved shape and can be used to approximate thick and thin shells of positive, zero and negative Gaussian curvature in a simple and straightforward manner. They also include anisotropic faces and non-constant core thickness. Degrees of freedom are limited to three displacements at each node, and this allows mixing with other element types.