Abstract:
It is argued that the formulation of the method of applying a flutter constraint in weight optimisation which allows for all the methods of avoiding flutter is so complicated that it is best not to try to include a flutter constraint in a general optimisation program with other constraints but to apply it only if the structure obtained from optimisation with the other constraints alone is flutter-prone. In this way much unnecessary and possibly inappropriate calculation can be avoided. A method of weight optimisation with only flutter constraints, based oninverse iteration, which can be used in conjunction with any suitable optimisation procedure which requires the values of the objective function and its first derivatives to be calculable is suggested. Its use in association with more than one procedure is described as is the optimisation of a wing of fairly-high aspect ratio. The question of how detailed a representation of the structure is needed is also discussed.