A Computational Architecture for Coupling Heterogeneous Numerical Models and Computing Coupled Derivatives

Abstract

One of the challenges in computational modeling is coupling models to solve multidisciplinary problems. Flow-based computational frameworks alleviate part of the challenge through a modular approach, where data flows from component to component. However, existing flow-based frameworks are inefficient when coupled derivatives are needed for optimization. To address this, we develop the modular analysis and unified derivatives (MAUD) architecture. MAUD formulates the multidisciplinary model as a nonlinear system of equations, which leads to a linear equation that unifies all methods for computing derivatives. This enables flow-based frameworks that use the MAUD architecture to provide a common interface for the chain rule, adjoint method, coupled adjoint method, and hybrid methods; MAUD automatically uses the appropriate method for the problem. A hierarchical, matrix-free approach enables modern solution techniques such as Newton-Krylov solvers to be used within this monolithic formulation without computational overhead. Two demonstration problems are solved using a Python implementation of MAUD: a nanosatellite optimization with more than 2 million unknowns and 25,000 design variables, and an aircraft optimization involving over 6,000 design variables and 23,000 constraints. MAUD is now implemented in the open source framework OpenMDAO, which has been used to solve aircraft, satellite, wind turbine, and turbofan engine design problems. CCS Concepts: • Software and its engineering → Object oriented frameworks; Software libraries and repositories; Object oriented development; • Theory of computation → Nonconvex optimization; • Mathematics of computing → Solvers;

Multipoint high-fidelity CFD-based aerodynamic shape optimization of a 10 MW wind turbine

“…In addition, the unsteady interaction between tower and blade has been simulated for the NREL Phase VI RWT with EllipSys3D using overset grid capabilities, and an overall good agreement was found with experimental data (Zahle et al, 2009). EllipSys3D has been used in various rotor applications to perform computations, such as aerodynamic power (Johansen et al, 2009) and fluid-structure interaction (Heinz et al, 2016). The latter work also encompasses a comparison across fidelities between the CFD-based tool, HAWC2CFD, and the BEM-based HAWC2 solvers where a good agreement was found.…”

Optimal Control within the Context of Multidisciplinary Design, Analysis, and Optimization

“…Since these methods are so central to the functionality of Dymos, their derivation is very briefly summarized here. Much more complete and detailed explanations of the analytic methods can be found in the work of Martins and Hwang [20,21].…”

“…Due to the effort required, a given implementation is typically uni-directional, supporting either the direct or adjoint methods, but not both. However, recent theoretical developments by Martins and Hwang [20,21] have provided a more generalized approach that has been implemented in NASA's OpenMDAO framework, eliminating the need for one-off implementations of the analytic derivative methods. As-such the implementation barrier to adopting analytic derivative methods has been significantly lowered and it has also become possible to apply a combination of forward and adjoint methods on a single problem to create a bi-directional analytic derivative approach.…”

How Certain Physical Considerations Impact Aerostructural Wing Optimization

“…Throughout this work, we use OpenMDAO [14] as the underlying optimization framework. OpenMDAO was developed at NASA Glenn and uses the modular analysis and unified derivatives theory to allow for modular construction and execution of complicated models [20]. OpenMDAO has been used to optimize a variety of problems, including wind turbines [21,22], boundary layer ingestion aircraft [23,24], thermodynamic engine cycles [25,26], and coupled thermal-mission problems [6,27].…”