Utilization of Response Surface Methodologies in the Multi-Discipline Design Optimization of an Aircraft Wing

A multi-disciplinary optimization analysis is a highly iterative process that requires a large number of function evaluations for computing the objective functions and the constraints. Metamodels (i.e. response surface methodologies) can be constructed before starting the optimization for each one of the objective functions and the constraint functions. The metamodels can be employed in the multi-discipline optimization instead of high fidelity simulations resulting in significant computational savings. A multi-discipline design optimization of an aircraft wing under aerodynamic and structural analysis considerations is performed in this manner. Design variables associated with the shape of the wing are considered in the CFD simulations, while sizing structural design variables are considered in the structural discipline. At the top system level, a cost type metric is defined for driving the overall design optimization process. The validity of employing the metamodels instead of the actual solvers during the optimization is demonstrated by ensuring that when actual solvers are used for computations at the optimal point all constraints are satisfied and all objective functions are improved at the expected levels.