Abstract
A vast majority of utility-scale wind turbine generators in the United States depend on foreign-sourced rare-earth permanent magnets that are vulnerable to supply chain uncertainties. Many small wind original equipment manufacturers are motivated to pursue continuous improvements to the generator design to lower the material and production costs and improve performance by lowering cogging torque and increasing efficiency. Traditional design and manufacturing offer limited opportunities. In this work, we demonstrate advanced design approaches for a 15-kW baseline wind turbine generator by making use of recent progress in three-dimensional (3D) printing of polymer-bonded magnets and electrical and structural steel. We explore three methods of magnet parametrization using Bezier curves resulting in symmetric , asymmetric and multimaterial magnet designs. We employ a multiphysics approach combining parametric computer-aided-design modeling, finite-element analysis, and targeted sampling to identify novel designs with more opportunities for reducing rare-earth material, improving efficiency and minimizing cogging torque. The results show that asymmetric-pole design and multimaterial-pole designs offer a greater opportunity to minimize rare-earth magnet materials by up to 35% with similar performance as the baseline generator, suggesting newer opportunities with design freedom beyond traditional limits of symmetry and as allowed by 3D printing.
Original language | American English |
---|---|
Number of pages | 8 |
DOIs | |
State | Published - 2023 |
Event | 12th International Conference on Power Electronics, Machines and Drives (PEMD 2023) - Brussels, Belgium Duration: 23 Oct 2023 → 24 Oct 2023 |
Conference
Conference | 12th International Conference on Power Electronics, Machines and Drives (PEMD 2023) |
---|---|
City | Brussels, Belgium |
Period | 23/10/23 → 24/10/23 |
Bibliographical note
See NREL/CP-5000-86580 for preprintNREL Publication Number
- NREL/CP-5000-88635
Keywords
- Bezier curves
- multiphysics design
- polymer-bonded magnets
- shape optimization
- targeted sampling