Mohd Salman Leong

Condition Monitoring of Blade in Turbomachinery: A Review

Blade faults and blade failures are ranked among the most frequent causes of failures in turbomachinery. This paper provides a review on the condition monitoring techniques and the most suitable signal analysis methods to detect and diagnose the health condition of blades in turbomachinery. In this paper, blade faults are categorised into five types in accordance with their nature and characteristics, namely, blade rubbing, blade fatigue failure, blade deformations (twisting, creeping, corrosion, and erosion), blade fouling, and loose blade. Reviews on characteristics and the specific diagnostic methods to detect each type of blade faults are also presented. This paper also aims to provide a reference in selecting the most suitable approaches to monitor the health condition of blades in turbomachinery.

Dempster-shafer evidence theory for multi-bearing faults diagnosis

Support vector machines (SVMs) are frequently used in automated machinery faults diagnosis to classify multiple machinery faults by handling a high number of input features with low sampling data sets. SVMs are well known for fault detection that involves binary fault classifications only (i.e., healthy vs. faulty). However, when SVMs are used for multi-faults diagnostics and classification, they result in a drop in classification accuracy; this is because the adaptation of SVMs for multi-faults classifications requires the reduction of the multiple classification problem into multiple subsets of binary classification problems that result in many contradictory results from each individual SVM model. To overcome this problem, a novel SVM-DS (Dempster-Shafer evidence theory) model is proposed to resolve conflicting results generated from each SVM model and thus increase the classification accuracy. The analysis of results shows that the proposed SVM-DS model increased the accuracy of the fault diagnosis model from 76% to 94%, as SVM-DS continuously refines and eliminates all conflicting results from the original SVM model. The proposed SVM-DS model is found to be more accurate and effective in handling multi-faults diagnostic and classification problems commonly faced in the industries, as compared to the original SVM method.

Reconstruction of Vital Blade Signal from Unsteady Casing Vibration

Some important information pertaining to blade fault is thought to be concealed in highly unsteady casing vibration. This paper explores suitable methods to best reconstruct blade related signals from raw casing vibration, which could be used for diagnosis of blade fault. The feasibility of translation invariant wavelet transform and cycle spinning (TIWT-CS) technique in reconstruction of these signals is investigated in this paper. Subsequently, a new parameter for blade fault diagnosis, namely, the energy profile of blade signal (EPBS), is formulated. Experimental results show that TIWT-CS method effectively retained blade related signals, while other unwanted signals such as system noises and aerodynamic induced vibration are reasonably suppressed. EPBS provides an indication of the condition of blade faults in rotor system, whereby the exact position and the quantity of faulty blades, as well as the root cause of blade fault, can be identified. In comparison, the energy profile plots using unfiltered casing vibration were found to be highly unstable and therefore provides inconsistent results for diagnosis of blade fault.

An improved wrapper-based feature selection method for machinery fault diagnosis

A major issue of machinery fault diagnosis using vibration signals is that it is over-reliant on personnel knowledge and experience in interpreting the signal. Thus, machine learning has been adapted for machinery fault diagnosis. The quantity and quality of the input features, however, influence the fault classification performance. Feature selection plays a vital role in selecting the most representative feature subset for the machine learning algorithm. In contrast, the trade-off relationship between capability when selecting the best feature subset and computational effort is inevitable in the wrapper-based feature selection (WFS) method. This paper proposes an improved WFS technique before integration with a support vector machine (SVM) model classifier as a complete fault diagnosis system for a rolling element bearing case study. The bearing vibration dataset made available by the Case Western Reserve University Bearing Data Centre was executed using the proposed WFS and its performance has been analysed and discussed. The results reveal that the proposed WFS secures the best feature subset with a lower computational effort by eliminating the redundancy of re-evaluation. The proposed WFS has therefore been found to be capable and efficient to carry out feature selection tasks.

Condition Monitoring of Industrial Gas Turbine Critical Operating Parameters Using Statistical Process Control Charts

This paper presents condition monitoring of industrial gas turbine by monitoring its critical operating parameters using statistical process control. This will consequently enables the detection of any degradation of gas turbine operating parameters and thus to better prepare for any forward actions that required. Basically performance of gas turbine and its critical operating parameters degrades over time. These parameters however degrades and eventually reach the OEM recomended limits without even triggereing any earlier alerts. Therefore, corrective maintenance actions are required to bring the parameters back to an acceptable operating condition which causing downtime in operation and accounts for large maintenance together with operating costs. Hence by identifying any degradation and deviation in gas turbine parameters in advance before it reaches its OEM limit will help to improve maintenance scheduling and practices and thus enhanced the reliability of the machine. It also able to identify false alarms and shutdowns which can cause unnecessary maintenance and non profitable stops. SFC method is also found to be able to estimate the progression of component/ performance degradation and thereby generating a continuously updated prediction of the remaining useful life of machine components. SPC based machine condition monitoring uses statistical process control charts such as individual and moving range methods to create the operating threshold of the machine. These thresholds were showed to be capable to determine and identify performance degradation in advance or earlier before it reaches the OEM limits for each individual parameters.

The effect of non-uniform aerodynamic loading on the blade responses

Turbine blades are always subjected to severe aerodynamic loading. The aerodynamic loading is uniform and Of harmonic nature. The harmonic nature depends on the rotor speed and number of nozzles (vanes counts). This harmonic loading is the main sources responsible for blade excitation. In some circumstances, the aerodynamic loading is not uniform and varies circumferentially. This paper discussed the effect of the non-uniform aerodynamic loading on the blade vibrational responses. The work involved the experimental study of forced response amplitude of model blades due to inlet flow distortion in the presence of airflow. This controlled inlet flow distortion therefore represents a nearly realistic environment involving rotating blades in the presence of airflow. A test rig was fabricated consisting of a rotating bladed disk assembly, an inlet flow section (where flow could be controlled or distorted in an incremental manner), flow conditioning module and an aerodynamic flow generator (air suction module with an intake fan) for investigations under laboratory conditions. Tests were undertaken for a combination of different air-flow velocities and blade rotational speeds. The experimental results showed that when the blades were subjected to unsteady aerodynamic loading, the responses of the blades increased and new frequencies were excited. The magnitude of the responses and the responses that corresponding to these new excited frequencies increased with the increase in the airflow velocity. Moreover, as the flow velocity increased the number of the newly excited frequency increased.Copyright