The main focus of this project was the Condition Monitoring (CM) of Axial Piston Pumps (APPs), which, in contrast to other types of machinery, has received far less attention in both research and industry. Nevertheless, APPs are widely used in several critical industrial applications and represent a significant investment. The harsh operating conditions and heavy loads make these systems highly susceptible to damage. Therefore, the development of highly sensitive and responsive CM algorithms has the potential to generate substantial operational and maintenance savings.
Machine Learning (ML)-driven CM methods offer promising damage diagnosis capabilities; however, their performance depends on the availability of sufficient training data. The primary objective of this project is to develop high-fidelity Finite Element Models (FEMs) to create so-called “Digital Twins” of APPs. These digital twins are used as synthetic data generators for various damage scenarios, supporting the training and validation of ML-driven CM algorithms.
The project is funded by the Innovation Fund Denmark.