Design-oriented aeroservoelastic optimization of strain-actuated aircraft

An integrated design-oriented aeroservoelastic optimization capability for strain-actuated aircraft is presented. This capability is called SMART and it encompasses a suite of computer applications created for conceptual and preliminary design of aircraft augmented with "smart" actuation technologies. The SMART suite of applications includes: (1) a dedicated pre-processor for vehicle geometry, material, actuator, mechanism, and sensor layout; (2) a dedicated finite element automesher for conventional and strain-actuated flight vehicles; (3) integration of structural dynamics with a state of the art commercial unsteady aerodynamics code (ZAERO) via automated pre- and post-processors; (4) a database architecture for analyzing multiple designs and flight conditions; and (5) automated open- and closed-loop aeroservoelastic (ASE) model preparation.; The analysis techniques used as the basis for SMART are suitable (within the range of application of linear theory) for modeling real flight vehicles with real large-scale structural, aerodynamic, and control systems. These techniques include: (6) dedicated linear finite element infrastructure for modeling conventional and strain-actuated (temperature and voltage induced) flight structures; (7) dedicated static and dynamic finite element solvers; (8) state space stability analysis for coupled aeroservoelastic systems; (9) computational tools for LQR controller design; and (10) analysis tools for the calculation of random response of linear systems to random inputs.; SMART can compute: (11) static aeroelastic deformations and stresses in trimmed maneuvering elastic airplanes; (12) open-loop aeroelastic poles; (13) closed-loop (control by LQR) aeroservoelastic poles; (14) gust response (random gusts) of the open-loop and closed-loop aeroservoelastic system, and, also, when aerodynamic loads are not present; and (15) deformation and stresses of passive or actively-controlled structures subject to dynamic and static loads.; To tailor the analysis capabilities for integration with computational gradient-based optimization tools, SMART includes: (16) computation of both analytic and finite-difference design sensitivities for all behavior functions with respect to all design variables (where the finite-difference sensitivities are used to check the analytic sensitivities, but where analytic sensitivities are used in the optimization process); (17) direct, reciprocal, and hybrid Taylor Series response function approximations; and (18) integration with a Feasible-Direction optimization code (CONMIN) in a design optimization system based on the Approximation Concepts/Nonlinear Programming (NLP/AC) strategy.