The wings of flying animals undergo large, periodic deformations during flight due to aerodynamic and inertial loads. Insect wings express a sparse array of mechanosensors sensitive to strains and airflows, allowing rapid monitoring of local loads and wing aerodynamic state. Our goal is to discover how insects perceive mechanosensory information and use it to control flight. We use dragonflies as a model because they have excellent control in both gliding and flapping flight. Here, we present numerical simulations of a gliding dragonfly with high-fidelity wing geometries that flex under aerodynamic loads. To solve for the wings’ aeroelastic responses, computational structural dynamics (CSD) and computational fluid dynamics (CFD) solvers are loosely coupled. We also present CFD results of a flapping dragonfly in free flight, using our high-fidelity wing models, but with prescribed kinematic deformations. There are three advances: i) a high-fidelity wing model containing detail of micro-structural ridges, valleys and vein cross-sectional geometries, acquired using a hybrid approach of micro-computed tomography and stereo photogrammetry; ii) a demonstrated capacity for fluid-structure interaction (FSI) modelling using commercial code; iii) acquisition of accurate, deforming-wing, flapping kinematics, including wing twist, from freely flying dragonflies using nine, synchronised high-speed cameras. The resulting flow fields and strain fields on the wing surfaces are examined to see how the distributed sensors would observe discrete data for flight control.