Modeling and evaluation of centrifugal pump performance degradation model based on stochastic process

Xuan Chen, Jia Li, Ping Li, Limin Gao, Xiaolong ChenJournal of Physics Conference Series 2024 · DOI

Abstract

In order to clarify the performance degradation problem of fuel centrifugal pumps, a stochastic process-based performance degradation model modeling method is given with an aero-engine fuel centrifugal pump as the object of study, and the reliability assessment is carried out. First, the degradation modeling method based on stochastic process is given. Secondly, PumpLinx is utilized to obtain the performance degradation data of the centrifugal pump and perform data processing. Gamma and Wiener stochastic process models are given based on MATLAB. Finally, based on the centrifugal pump performance degradation data for centrifugal pump reliability assessment, and compare and analyze the advantages and disadvantages of the two modeling methods, and found that the degradation modeling method based on the Gamma process is more suitable for centrifugal pump reliability assessment.

Introduction

Fuel centrifugal pumps in aircraft engines are high-power, long-life components essential for flight safety. Traditionally, assessing their reliability requires "time-to-failure" testing, which is prohibitively expensive and time-consuming. A single test bench might need to run for years to produce a single data point. Furthermore, these statistics often fail to account for individual component variances or provide early warnings of impending failure.

This research shifts the focus from total failure to performance degradation. By monitoring how key performance indicators, such as outlet pressure, decline over time, engineers can predict the remaining useful life of a pump. The central challenge lies in selecting a mathematical framework—a stochastic process—that accurately reflects the physical reality of mechanical wear and tear.

Methodology: Digital Twin and Stochastic Modeling

The study employs a workflow combining high-fidelity simulation with statistical analysis. Researchers constructed a 3D model of a fuel booster pump using Cfturbo and utilized PumpLinx for Computational Fluid Dynamics (CFD) simulations.

3D model of the fuel centrifugal pump used for simulation.

Degradation was simulated by incrementally decreasing the pump speed by 150 rpm for every 500 hours of virtual operation. The resulting drop in outlet pressure served as the primary proxy for performance degradation. Two primary mathematical models were evaluated:

  • Wiener Process: Often used for "random walks," assuming degradation can fluctuate due to random perturbations.
  • Gamma Process: Designed for strictly monotonic (one-way) accumulation, aligning with the physical reality of irreversible mechanical wear.

Key Results and Discussion

The evaluation focused on which model best represented the simulated degradation data, where "failure" was defined as a 10% decrease in outlet pressure.

Comparison of reliability curves for Wiener, Gamma, and IG stochastic processes.

Model Comparison

While both the Wiener and Gamma models showed similar overall trends, the Gamma process proved superior. Because the Gamma process is strictly non-negative and monotonic, it aligns with the cumulative nature of mechanical wear. In contrast, the Wiener process might underestimate a part's life by allowing for unrealistic performance "recoveries" in the model, potentially leading to premature maintenance and resource waste.

Reliability Benchmarking

Based on the Gamma process model, the study identified a critical reliability benchmark: the centrifugal pump maintains a 0.9 reliability level at 4,200 hours of operation. The simulation reached convergence within 2,000 iterations, ensuring a stable baseline for the monitoring curves used in the stochastic models.

Conclusion

The research demonstrates that the Gamma stochastic process is more suitable for assessing the reliability of centrifugal pumps than the Wiener process. By leveraging CFD simulations as a "digital twin," engineers can calibrate these models without the need for exhaustive physical testing. Future work will focus on validating these findings against real-world field data to bridge the gap between digital simulations and physical engine performance.

Summary updated

Abstract

Abstract In order to clarify the performance degradation problem of fuel centrifugal pumps, a stochastic process-based performance degradation model modeling method is given with an aero-engine fuel centrifugal pump as the object of study, and the reliability assessment is carried out. First, the degradation modeling method based on stochastic process is given. Secondly, PumpLinx is utilized to obtain the performance degradation data of the centrifugal pump and perform data processing. Gamma and Wiener stochastic process models are given based on MATLAB. Finally, based on the centrifugal pump performance degradation data for centrifugal pump reliability assessment, and compare and analyze the advantages and disadvantages of the two modeling methods, and found that the degradation modeling method based on the Gamma process is more suitable for centrifugal pump reliability assessment.
BibTeX
@article{chen2024modeling,
  author = {Xuan Chen and Jia Li and Ping Li and Limin Gao and Xiaolong Chen},
  title = {{Modeling and evaluation of centrifugal pump performance degradation model based on stochastic process}},
  journal = {Journal of Physics Conference Series},
  year = {2024},
  doi = {10.1088/1742-6596/2746/1/012014},
}
Figure