Multidisciplinary optimization — Application to aerospace systems design

The design of aerospace systems (airplanes, helicopters, launchers, satellites, missiles, etc.) involves complex processes in which the search for the best performance at the lowest cost, while ensuring maximum reliability, is decisive. These design processes call on numerous disciplines (aerodynamics, propulsion, trajectory, structure, etc.) which must be mastered in order to guarantee the optimality of the designed system and reduce design cycle times. These disciplines, often with conflicting objectives, require appropriate design tools to integrate the constraints inherent in each of them and facilitate the search for compromise. Taking into account couplings between disciplines significantly increases the complexity of the problem to be solved (new constraints to be satisfied, increase in problem dimension, computation time).

Multidisciplinary Design Analysis and Optimization ( – MDAO) is a set of methods derived from applied mathematics (mathematical formulation of the optimization problem, optimization algorithms, substitution models, sensitivity analysis, quantification of uncertainties, etc.) with the aim of rationalizing and managing the entire design process. The aim is to achieve greater exploration of the search space, as well as gains in efficiency and computing time. as well as better control of the uncertainties inherent in the disciplines modeled and the operational context. Unlike conventional design processes, in which disciplines are considered sequentially, interactions between disciplines are directly taken into account in CADD approaches. , enabling effective management of interdisciplinary couplings and trade-offs. However, the complexity of the design process (problem size, computation time, number of constraints, non-linearities) is significantly increased by the simultaneous management of all disciplines within a single optimization problem. The aim of this article is to provide an overview of the tools available in the CADD field.