Condition Diagnostics of Marine Centrifugal Pumps Based on Blade-Passing Frequency Harmonics with Analytical DFT Leakage Compensation
Abstract
The reliable operation of high-power marine centrifugal pumps is critical for the safety and efficiency of ballast and fire-fighting systems, as well as for the functioning of a wide range of power and auxiliary installations on large ocean-going vessels. This paper presents a vibrodiagnostic method for the condition monitoring of a vertical in-line marine centrifugal pump manufactured by DESMI, based on the frequency-domain analysis of vibration signals measured in the impeller zone. The paper focuses on the study of discrete amplitude spectra obtained under steady-state operating conditions at a rotational speed close to 900 rpm. Particular attention is paid to the distortion of diagnostic features caused by spectral leakage of the fundamental rotational harmonic, which significantly affects the identification of key diagnostic parameters. An analytical procedure for compensating the power leakage of the central harmonic is proposed, enabling accurate reconstruction of the true spectral energy distribution without increasing the measurement record length or applying complex windowing techniques. The proposed approach improves the resolution and interpretability of vibration spectra within the operating frequency range of the pump. The method enhances the reliability of vibration-based condition monitoring for marine centrifugal pumps operating under real shipboard conditions and can be implemented in existing condition-based maintenance systems without additional hardware modification. Moreover, a procedure for a continuous vibration monitoring system for a high-power centrifugal pump, suitable for integration into modern controllers with FFT functionality, is developed. The results demonstrate that the proposed analytical compensation of spectral leakage is an effective tool for improving the vibrodiagnostic reliability of high-power shipboard pumps and supports the early detection of hydraulic and mechanical degradation in critical auxiliary marine machinery.
© 2026 Roman Varbanets, Vladyslav Kyrnats, Volodymyr Kholdenko, Dmytro Minchev, Ievgen Bilousov, Vladyslav Maulevych, Kucherenko Volodymyr, Aleksandrovska Nadiia, published by Gdansk University of Technology
This work is licensed under the Creative Commons Attribution 4.0 License.