Monitoring of Rotating Blades Using Shaft Vibration Measurement and Tracking in a Frequency Domain

Monitoring of vibrations is an important task to secure proper operation of rotating machines. Early detection of an anomaly behaviour can help to prevent operation in non-design conditions or even from a fault of a machine. This is emphasized especially in case of a rotating blade, that is one of the most stressed parts of the steam turbine. Standard approach for monitoring of the blade vibrations is the use of sensors detecting the time of arrival of each blade, that is so-called the tip timing method. The sensors need to be drilled into the stator or the instrumentation must be taken into the account in early phase of the turbine designing. To avoid these challenges and be competitive, the approach using the relative shaft vibration sensors was developed by the authors of the paper. The methodology uses advanced signal processing techniques to analyse the vibration signals acquired by employing spectral analysis and frequency domain tracking algorithm. The proposed approach enables the identification of specific blade frequencies associated with the longest impeller blades. Next, the paper discusses the design and implementation of a prototype monitoring system, based on this approach. Case study demonstrate the proposed methodology in detecting the blade characteristics under different operating conditions of 250 MW turbogenerator.