We propose to solve the challenge problem of truly integrated high-fidelity design of next generation aircraft. In particular, we propose to enable high-fidelity design exploration for NASA n+3…

This is a cooperative agreement awarded by NASA to the Regents of the University of Michigan, valued at $745,833.56, to develop a new multidisciplinary design optimization (MDO) approach that can handle large-scale aircraft design problems. The goal is to enable high-fidelity design exploration for NASA's N+3 aircraft concepts, such as the truss-braced wing, which are key to meeting NASA's 2025 emissions and fuel burn goals. The work involves integrating a range of aircraft design disciplines, including mission analysis, aerodynamics, and structures, into an efficient MDO framework. The project will result in two main deliverables: a general MDO approach that can be integrated into NASA's OpenMDAO software, and a study of conventional and truss-braced wing configurations within the context of next-generation airspace. There is no set-aside designation for this contract.

We propose to solve the challenge problem of truly integrated high-fidelity design of next generation aircraft. In particular, we propose to enable high-fidelity design exploration for NASA n+3 concepts, such as the truss-braced wing (TBW) by developing a novel multidisciplinary design optimization (MDO) capability. The NASA n+3 concepts contribute significantly towards NASA's 2025 emissions and fuel burn goals, but these concepts also require an unprecedented level of integration and fidelity in their design. This challenge problem requires new methods for large-scale MDO and integration of a greater range of relevant disciplines with appropriate levels of fidelity. Contemporary aircraft design remains a predominantly sequential process that misses significant opportunities to exploit interdisciplinary synergies. One such overlooked synergy is the influence of real world trajectories on the wing design, and their coupling to the aircraft route allocation. This is a challenging problem, because trajectory analysis requires the evaluation of many flight conditions, while wing-shape design requires high-fidelity aerodynamic and structural models. There are currently no MDO architectures that can handle such a challenging problem. Thus, there is a need for new approaches that can handle a large number of design variables and disciplines, with several disciplines consisting of high-fidelity models. To solve the challenge problem, we developed a new mathematical framework in phase i of this project that efficiently handles highdimensional models with large numbers of disciplines. This framework relies on a new view of the mathematical MDO formulation. This results in a new approach to couple disciplinary models that automatically computes coupled sensitivities, uses these sensitivities to achieve faster convergence of the coupled analysis, and makes this more robust. In addition, the coupled sensitivities enable the use of efficient gradient-based optimization. This new MDO approach has potential, but requires significant development to address the challenge problem. One area for further development is the integration of multifidelity and uncertainty quantification methods that automatically determine the appropriate fidelity required for each discipline. We propose an approach based on reduced modeling that integrates well with the overall MDO approach. Another area requiring a breakthrough is the optimization algorithm: although there has been much progress in parallel computing, there are currently no parallel optimization libraries that can be used on practical problems. Thus, we developed of a parallel optimizer in phase i that naturally fits in the proposed MDO approach and showed excellent scalability. In phase II we will extend this optimizer to handle more general problems. To apply the proposed MDO approach to the TBW, we developed high-fidelity a aerostructural model of the configuration in phase i. In phase II, we will integrate a more extensive set of aircraft design tools available within our team to solve this problem. These models include mission analysis, airline allocation, aircraft sizing, propulsion, aerodynamics, and structures. This project will result in two main deliverables: 1) a general MDO approach integrated into NASA's openmdao that can handle large-scale problems efficiently (a version of which was already delivered in phase i) and 2) a study of n+3 conventional and TBW configurations within the context of next generation airspace. These deliverables will contribute directly to the overarching goal of the NASA SFW program to drastically improve aircraft efficiency and minimize the impact of aviation on the environment.