Posters

Abstract

Reliable predictive maintenance in wind turbine transmission systems is critical for reducing Levelized Cost of Energy (LCOE) and preventing critical gearbox failures. Traditional debris monitoring, specifically particle sensing, often faces trade-offs between measurement latency and data resolution. This body of research details the development of a next-generation particle sensor by Inmox GmbH, utilizing a Digital Twin-driven analytical approach to overcome these limitations and exponentially improve the next generation. The research focuses on the transition from empirical prototyping to an advanced analytical simulation environment. By creating a high-fidelity Digital Twin of the transmission lubricant circuit, we were able to simulate complex interactions between magnetic debris, and electromagnetic sensing fields. This virtual environment allowed for the iterative testing of sensor geometries and signal-processing algorithms under variable flow rates that would be prohibitively expensive or time-consuming to replicate in a physical lab setting. The application of this methodology resulted in the development of a next-generation inductive particle sensor with significant performance breakthroughs. By optimizing the coil architecture and flux density within the digital model, we achieved a 50-fold increase in measurement speed and a 2.5X improvement in data resolution. These enhancements allow for the detection of finer wear particles, such as early-stage pitting and spalling, well before they escalate into critical component failure. Furthermore, the Digital Twin allowed for rigorous thermal stress analysis, leading to a redesigned housing and sensor core capable of sustained operation at 105 degrees Celsius. This capability is essential for modern high-capacity turbines where gearbox oil temperatures frequently exceed standard sensor ratings. This body of research concludes by comparing field-test data with the Digital Twin’s predictive models, validating the accuracy of the simulation. This work demonstrates that the "Digital Twin first" approach not only accelerates the development lifecycle but also unlocks a new standard of sensitivity and reliability for industrial IoT solutions in renewable energy.