This interdisciplinary study investigates impeller failure using a combined digital and experimental approach, establishing a proof of concept for cyber-physical integration. First, a CAD model was developed and its structural integrity is validated using Finite Element Analysis (FEA) to ensure the impeller could withstand operational loads and dynamic stresses, following the methodology outlined in 1. Next, an IoT-enabled digital twin framework was implemented with Arduino-based sensors (temperature, humidity, vibration) to monitor 3D-printed impellers made from 316L stainless steel and AlSi10Mg aluminium. The sensors were integrated with a custom test rig driven by a motor capable of 10,000 rpm, with data acquired via analog/digital interfaces and visualized in Node-RED, streaming in real time to an IoT cloud platform. Impeller experiments ran for over 80 hours and were tested under two corrosive conditions: (i) engine oil (5W-30) and (ii) saltwater. SEM/EDS analysis revealed carbon deposits on oil-exposed samples and aluminium oxide on saltwater-exposed ones, while further SEM imaging showed pitting and corrosion. Alicona surface roughness tests confirmed degradation under dynamic loads. Preliminary real-time monitoring demonstrated the of predictive maintenance alerts, though full-scale validation remains future work. Overall, the developed framework provides a robust basis for physical testing with digital representation, offering strong potential for predictive maintenance.
Sathish et al. (Mon,) studied this question.