Finite Element Analysis of a Weight Optimized Pre-Swirl Stator for Marine Applications

Sammanfattning

This thesis investigates the structural feasibility of replacing a conventional solid cast pre-swirl stator (PSS) hydrofoil with a lightweight lattice-filled design intended for additive manufacturing. The work was carried out in collaboration with Kongsberg Maritime and focused on evaluating whether significant weight reduction could be achieved without compromising structural performance under operational loading conditions. The proposed design consisted of an outer hydrofoil shell combined with an internal lattice filled volume intended to replace the solid interior of the reference component. The study combined finite element analysis using Ansys, lattice generation and homogenization using nTop, and surrogate-based optimization using Python. A solid reference wing was first analyzed to establish baseline values for stress, deformation, mass, and eigenfrequencies. Different lattice topologies, hydrofoil shell thicknesses, and lattice cell sizes were then evaluated in order to identify which of these were most suitable for further optimization. A polynomial response surface model was thereafter used to support the optimization of shell thickness and relative lattice density under predefined stress and deflection constraints. The results indicated that the Kelvin lattice topology showed the most favorable overall structural performance among the evaluated candidates. The optimized design achieved a mass reduction of approximately 65% compared to the reference wing, while still showing acceptable structural response under the studied loading conditions. Additional static structural and linear buckling analyses of a representative Kelvin lattice indicated sufficient local load carrying capacity and a large margin against global linear buckling. A simplified manufacturing lead time assessment also suggested that laser powder directed energy deposition (LP-DED) may offer shorter lead times than conventional casting under optimal manufacturing conditions. Overall, the study indicates that a lattice filled PSS design manufactured using LP-DED has promising potential as a lightweight alternative to a solid cast design. At the same time, the results should be interpreted in light of the limitations associated with homogenization, idealized material assumptions, and the absence of experimental validation.