Minimizing erosive wear through a CFD multi-objective optimization methodology for different operating points of a Francis turbine. (January 2020)

• Record Type: Journal Article
• Authors: Aponte, R.D.; Teran, L.A.; Grande, J.F.; Coronado, J.J.; Ladino, J.A.; Larrahondo, F.J.; Rodríguez, S.A.
• Is Part Of: Renewable energy. Volume 145(2020)
• Page Start: 2217
• Page End: 2232
• Publication Date: 2020-01
• Subjects: Erosion wear -- Computational fluid dynamics -- Francis turbines -- Optimization -- Genetic algorithms -- Artificial neural networks; Renewable energy sources -- Periodicals; Power resources -- Periodicals; Énergies renouvelables -- Périodiques; Ressources énergétiques -- Périodiques; 333.794
• DOI: 10.1016/j.renene.2019.07.116
• Languages: English
• ISSNs: 0960-1481

Abstract

Abstract: Erosive wear has been a serious concern in mainly run-of-the-river medium and small Francis turbines from both economic and technical perspectives. With the aim of finding ways to mitigate erosive wear, this paper proposes a methodology to obtain, via an optimization approach, geometries that maximize the resistance to erosive wear by hard particles and cavitation of the internal components (runner, guide vanes and cover labyrinths) of a Francis turbine. This improvement was implemented to reduce the costs of corrective maintenance and to maximize the machines' availability and energy generation profits. The methodology used computational fluid dynamics (CFD) and optimization techniques, such as the design of experiments of the factorial type, artificial neural networks and genetic algorithms with a multi-point approach, which includes two operation points, and a multi-objective approach, which simultaneously considers erosive wear by hard particles, cavitation damage and efficiency. It was found that the new geometries of the analysed components of the turbine can allow a decrease of up to 73% in the wear rate, maintaining an efficiency close to the original value throughout the operating range. With the optimized geometry, a mechanical check was performed using finite element simulations to validate that the optimal geometries had the required strength. Highlights: CFD, optimization and machine learning were combined to improve a Francis turbine. A new improved geometry of the runner blades, guide vanes was obtained. A mechanical verification of the optimized components was performed.