Abstract The development of a digital twin to simulate the vibrational behaviour of a Pelton hydropower unit during startups offers significant potential for enhanced maintenance by detecting early warning signs of anomalies. The proposed methodology, combining high-frequency vibrational data acquisition with an offline digital twin to improve data interpretation, is applied in this study to a 15MW horizontal dual Pelton unit located in the Swiss Alps. The digital twin models fluid-structure interactions during startup with a combination of CFD simulations and transient FEM analysis, predicting the vibrational response of the unit. This numerical model allows investigating the effect of specific anomalous cases, such as jet misalignment, on the excited frequencies of the runner, enabling the creation of a library of frequency responses for predictive maintenance. Experimental modal analysis and in-situ vibration data acquisition validate the numerical model. Preliminary results show the effectiveness of this approach in understanding and predicting the vibrational behaviour of a Pelton unit during startup. Further refinement is needed, especially in conducting transient modal analysis of the full runner and shaft line, which will be conducted as part of the Horizon ReHydro project.
Read full abstract