The effectiveness of a vision-based condition monitoring framework for determining gear health in gearboxes
Van Maele, D.; Poletto, J.C.; De Geest, R.; Fauconnier, D.; De Baets, P. (2025). The effectiveness of a vision-based condition monitoring framework for determining gear health in gearboxes. Journal of Physics: Conference Series 3025: 012004. https://dx.doi.org/10.1088/1742-6596/3025/1/012004
In: Journal of Physics: Conference Series. IOP Publishing: Bristol. ISSN 1742-6588; e-ISSN 1742-6596, more
|  |
| Authors | | Top |
- Van Maele, D.
- Poletto, J.C.
- De Geest, R.
|
- Fauconnier, D., more
- De Baets, P.
|
|
| Abstract |
Gearbox failures in wind turbines significantly impact operational reliability and maintenance costs, particularly in offshore environments. Traditional condition monitoring (CM) methods, such as vibration and oil analysis, often fail to detect defects early due to limitations caused by the indirect measurement approach. This is further amplified by external influences, such as changing weather conditions, on system dynamics. This study presents a vision-based condition monitoring (VCM) framework for gear health assessment. The proposed system integrates a camera system within a backto-back gear tester equipped and utilizes an automated auxiliary system to mitigate oil splash interference on image quality. The test rig was also equipped with accelerometers to compare visual and vibrational monitoring. The effectiveness of the VCM is evaluated through eight accelerated lifetime tests, inducing gear pitting, scuffing, and root cracks. Results demonstrate that visual monitoring enables early detection, straightforward qualitative analysis of gear defects, and quantification of damage progression by leveraging deep learning. Compared to vibrational monitoring, visual monitoring offers an alternative with improved diagnostic capabilities and reduced susceptibility to external factors, showing its potential for application in condition monitoring. These findings suggest that vision-based monitoring can enhance predictive maintenance strategies, reducing downtime and optimizing gearbox longevity in wind energy applications. |
|
