High-fidelity Hydrostructural Optimization of Composite Hydrodynamic Lifting Surfaces

Abstract

To improve efficiency, functionality, and reliability of hydrodynamic lifting surfaces, such as propeller blades, control surfaces, and hydrofoils, composite materials are increasingly used because of their high strength-to-weight ratio, corrosion resistance, and desired fatigue properties. Despite these superior properties, their design remains challenging for maritime applications due to the need to operate over a broad range of flow conditions, and associated complex physics, including separation, cavitation, high loading, and complex material failure mechanisms. Additionally, considering both the coupling between fluid forces and structural responses, and the interplay effects between the many geometry and material design variables, the optimal design is not intuitive.

Multidisciplinary design optimization (MDO) is a powerful tool that can tackle these design challenges by considering various design requirements simultaneously and effectively evaluating the tradeoffs. With advances in computing, it is possible to use coupled high-fidelity hydrostructural simulations to better capture the flow physics and predict the structural failure onset. However, using high-fidelity simulations with MDO is still limited due to the high computational cost, especially when considering a large number of design variables needed for composite hydrodynamic lifting surfaces with complex geometries and material configurations. The objective of this dissertation is to use an efficient high-fidelity MDO framework to explore the design of composite hydrodynamic lifting surfaces and examine relevant design and research questions that are important but still unresolved so far. To address the issue of high computational cost, this dissertation uses a gradient-based optimization approach and leverages the adjoint method to compute the gradient efficiently. The contributions of this dissertation are the development of methods that optimize composite hydrodynamic lifting surface designs. Optimizations are performed to yield novel findings on the tradeoffs and coupling effects between design conditions and design variables.

First, a more effective cavitation constraint, a solid composite element for the structural solver, and the corresponding failure initiation criteria are implemented in the framework. Second, this dissertation pioneers the use of a displacement constraint as a surrogate for dynamic loading consideration to yield a safer and more reliable design. With these developments, this dissertation presents an optimized composite hydrofoil with significantly delayed cavitation inception. A series of optimization studies are conducted to investigate how planform variables, material configurations, and failure initiation model uncertainties affect composite lifting surface designs.

This dissertation also advances the methodology to consider a more complex detailed geometry problem -- optimization of a structure with junction shape, which commonly exists and is critical to the overall performance. Specifically, this is demonstrated and investigated with hydrodynamic optimization of a hydrofoil-strut system. These optimization studies show the framework can adjust the junction shape to avoid junction cavitation and flow separation, all while improving efficiency.

The framework and presented optimization studies in this dissertation demonstrate the usefulness of the developed methods for hydrodynamic lifting surface designs. The discussions also provide valuable insights for designers and future research.