Using AI-assisted optimization methods, Voith and the DLR Institute are improving the water turbine design process and enhancing turbine development.
Voith and the DLR Institute of Propulsion Technology are working together using AI-assisted optimization methods through AutoOpti, developed by DLR, to improve the water turbine design process and significantly enhance turbine development. Engineers aim to create a multidisciplinary optimization system that increases turbomachinery efficiency and deepens the understanding of complex flow phenomena.
Modern water turbines face growing innovation demands—from achieving maximum efficiency and long service life to ensuring high operational safety. One of the main challenges is designing complex flow components like runners and guide vanes. At the same time, digital, data-driven processes are becoming more crucial, with increased pressure to cut down development time and costs. To meet these challenges, innovative approaches to turbine design are needed.
AutoOpti
The optimization tool AutoOpti was developed for automated, cross-disciplinary design processes and optimization. It enables the integration of fluid dynamics, structural mechanics, and other simulation models, incorporating smart strategies for design selection and model simplification.
AutoOpti will be used to optimize the flow mechanics of turbine components; for example, rotor blades and guide vanes can be adjusted in a targeted way to reduce hydraulic losses and mechanical stress much faster than traditional development methods.
Real geometries and boundary conditions from Voith’s product development are incorporated into an automated optimization process. AutoOpti:
By using surrogate models (Gaussian Processes), the number of expensive CFD calculations needed is greatly decreased. This approach allows the design space to be explored systematically and efficiently.
When optimizing the Francis turbine, AutoOpti produces new design variations that achieve significant efficiency gains across multiple operating points, while maintaining consistent flow rates. Meanwhile, initial surrogate models trained with high-fidelity CFD data were successfully integrated into the process. Overall, AutoOpti reduces development time considerably and improves the quality of turbine design.
The next phase involves expanding AutoOpti’s use to other turbine components and incorporating additional physical disciplines, possibly applying AutoOpti to other Voith product lines.
The DLR Institute of Propulsion Technology provides close support through technical workshops, specialized training, and ongoing assistance to incorporate these methods into existing development workflows.
Voith also recently
The Hydro Pocket dashboard facilitates seamless local use and underscores the importance of localized, customer-centric digital solutions. Plant owners can monitor, analyze, and optimize their hydropower stations through a simple, cloud-based application.
"Key benefits include faster decision-making, increased performance, and proactive issue resolution," said Dirk Fuchs, Head of HyService Digital & Automation at Voith Hydro. "Voith also provides additional insights and analysis through expert teams."
