Structural constraints investigation in wing aerodynamic/structural optimization

For wing aero-structural optimization problem, a large number of stress constraints have to be considered and numerical difficulties in the optimization is caused. To overcome these difficulties and reduce the computation cost, this paper aims to find an efficient constraint aggregation method, in which the constraints are aggregated to one or few constraints. Four approaches are investigated, i.e., maximum constraint approach and constraint aggregation approaches using KS function (with constant parameter, adaptive parameter and aggregating constraints of different component respectively). An in-house surrogate-based optimization (SBO) toolbox called SurroOpt is used to tackle the problems associated with no-smooth constraint bound. The comparison of different approaches is carried out in both structural and aero-structural optimizations of a transport aircraft wing. The results show that, (1) adaptive constraint aggregation using KS function achieves the maximum weight reduction; (2) for constant aggregation method, more wing weight reduction can be achieved, when larger aggregation parameter is used; (3) the SBO is very efficient and robust, even when the curvature of the KS function where the active constraints intersect is very large.