Experimental Method To Estimate Remaining Useful Life Of Gear Pumps

Pumps are critical components in hydraulic systems, particularly in off-road machinery, where unplanned downtime can lead to significant operational costs. Developing predictive maintenance models can mitigate these disruptions. However, for developing accurate predictive maintenance models, reliable data is needed. To address this, we designed and implemented an accelerated life test rig that simulates real-world operating conditions for gear pumps. This test rig allows the collection of degradation data under controlled conditions, laying the groundwork for remaining useful life (RUL) estimation models for predictive maintenance.Baseline testing demonstrated that both gear pumps, under clean conditions, performed similarly with stable flow rates. To accelerate degradation, 20 µm aluminum oxide particles were introduced into circulation of one pump, while the other operated without contamination. The particle-contaminated pump showed significant wear, with its flow rate decreasing rapidly in subsequent tests. In contrast, the uncontaminated pump maintained consistent performance over a longer period. A linear degradation model based on volumetric efficiency and flow rate reduction was applied to estimate the RUL. The results indicated that degradation can be expedited by over twofold using the ALTR. Future research will focus on expanding the dataset and employing more advanced data-driven and hybrid models to enhance the accuracy of RUL predictions for hydraulic systems.