Vol. 9; No. 1; 2006
Condition Monitoring in UK Defence – The Challenges Ahead

F. Hawkins
Abstract: Developments in the miniaturization of electronics, processing power and improved distributed communications are creating new possibilities for the health monitoring of military equipment. It is perceived that condition monitoring technology can support the reform of logistic processes, deliver savings and improve equipment availability. Targeting the technology intelligently is a demanding process though; there are many issues to consider. This paper places condition monitoring into perspective within wider reliability issues, discusses some potential problem areas and some of the additional benefits which successful exploitation of the technology can bring.
Keywords: Condition monitoring; HUMS; Defence; Mission reliability. Pages: 2 – 8. Refs: 9.

A Correlation between Acoustic Emission and Asperity Contact of Spur Gears under Partial Elastohydrodynamic Lubrication

C. Keong & D. Mba
Abstract: The application of the Acoustic Emission (AE) technique to condition monitoring of gears is still at a developmental stage. Understanding the source of AE activity at the gear mesh is fundamental if this technique is to be successfully employed for gear diagnostics and prognostics. This paper presents results of an experimental programme that showed a correlation between AE activity and asperity contact under partial elastohydrodynamic lubrication. Based on the observations, it is postulated that AE can monitor the effectiveness of a lubricant in minimising asperity contact.
Keywords: Acoustic emission; Elastohydrodynamic lubrication; Asperity contact; Gearbox diagnosis. Pages: 9 – 14. Refs: 11.

Designing a Human Machine User Interface for a Condition Based Monitoring System

P.A. Higgs, R. Parkin, M. Jackson & F. Zorriassatine
Abstract: The man-machine interface or method of human computer interaction (HCI) represents the most crucial element of a computer system, as its design determines eventual user acceptance and utilisation. Graphic User Interfaces (GUIs) are the most popular form of HCI presently used today. In this paper, practitioner desirable Condition Based Monitoring (CBM) GUI design are discussed.
Keywords: Condition based monitoring; Maintenance; Graphic user interface design. Pages: 15 – 22. Refs: 15.

Artificial Neural Network Prediction and Quantification of Damage in Impeller Shaft using Finite Element Simulation

S. Vijayakumar & S. Muthukumar
Abstract: This paper presents the application of ANNs for damage detection and quantification in impeller shafts, which are used in radial fans at power generation plants. The finite element model is simulated using three dimensional shell elements available in ANSYS. During simulation, the symmetry and other dynamic loads from rotating parts have been taken into account. The ‘taking advantage symmetry’ approach and a wave front method for static and modal analysis considerably reduce the computational time. Using undamped free vibration response, the frequency response functions (FRF) are extracted. For various presumed damages through element stiffness reduction. A multiplayer back propagation ANN, with two hidden layers, is designed using Matlab.
Keywords: Damage detection; Impeller shaft; Finite element modelling; Frequency response functions. Pages: 23 – 29. Refs: 18.

Detection of Induction Motor Stator Winding and Unbalance Faults using Hybrid Methods

W. Li & C.K. Mechefske
Abstract: This paper presents the results of a study aimed at the detection of three phase induction motor stator winding and unbalance faults using experimental techniques. Unlike broken bar and rolling element bearing faults, a stator winding fault does not have well-defined fault frequencies in the vibration, acoustic or stator winding current spectra. On the other hand, an unbalance fault may yield vibration spectral harmonics that are similar to other faults. For these reasons it is sometimes difficult to detect and diagnose these faults by using only one monitoring scheme. This paper employs a combination of vibration, stator current and acoustic methods. Simulated motor fault experiments were conducted in a laboratory environment under different speed and load conditions. Experimental results suggest that particular spectral harmonics from measurements are sensitive to different motor faults. Hybrid methods are highly recommended to obtain accurate monitoring results.
Keywords: Induction motors; Stator winding; Unbalance; Fauts; Hybrid methods. Pages: 30 –36. Refs: 12.

Vol. 9; No. 2; 2006

A Requirements Management Approach Supporting Integrated Health Management Systems Design

P. Soderholm
Abstract: System health management is an approach that is intended to improve the dependability and safety of technical systems, and to decrease the combined cost of operation and support. In order to achieve the potential benefits of Health management, it is necessary to combine stakeholder requirements with a thorough engineering knowledge. The aim of this paper is to present a systemic, systematic, and stakeholder-centred Requirements Management approach that supports the design of Integrated Health Management System (IHMS). A holistic management model intended to increase stakeholder satisfaction with a reduced amount of resources in also presented. This approach is applied to a case-study related to a modern combat aircraft, which is a highly complex and critical technical system
Keywords: Requirements management; Integrated health management system; Case-study. Pages: 2 – 13. Ref: 33.

The Extended Kalman Filter as a Tool for Condition Monitoring in Hydraulic Systems

Y. Chinniah
Abstract: Usually, parameters can be related to the health of the system and as such, deviations from the normal values of the parameters can be linked to developing faults. When the parameters cannot be measured directly using sensors, they need to be estimated. The Extended Kalman Filter (EKF) is a state and parameter estimation algorithm, which has recently been used for early fault detection in fluid power systems. In this paper, the EKF algorithm is described, and to illustrate the technique, it is applied to a simulated model of a simple mass spring damper system and then to a simulated model of a high performance hydrostatic system, the electrohydraulic actuator (EHA). The EKF is used to estimate important parameters such as the spring constant and the viscous damping coefficient in the mass spring damper system and the effective bulk modulus in the hydrostatic system. Using simulation studies, the ability for the filter to detect and estimate changes in the parameters is investigated. The EKF is a viable tool for early fault detection in hydraulic systems.
Keywords: Condition monitoring; Hydraulic systems; Extended Kalman filters. Pages: 14 – 22. Refs: 21.

Condition Monitoring and Fault Diagnosis of Engineering and Manufacturing Systems. Part I: Vibro-Acoustics Monitoring

C.K. Mukhopadhyay, T. Jayakumar, Baldev Raj & B.K.N. Rao
Abstract: Machinery vibration and noise play a major part in industrial plants and machinery. There are many reported instances where the performance, reliability, safety and health of both physical and human assets have significantly deteriorated, thereby shortening not only the useful life-cycle but also the availability, maintainability and productivity of these valuable assets. Intelligent sensing, monitoring and diagnosing the root cause failure of these industrial assets is now urgent than ever before. In this state-of-the-art review, the authors have attempted to highlight some of the current trends and progress in the world of vibration and acoustic emission technology for condition monitoring applications.
Keywords: Condition monitoring: Failure diagnosis; Noise and Vibration; Sensor technology; Acoustic emission. Pages: 23 – 40. Refs: 88.

Vol. 9; No. 3; 2006
Intelligent Systems

S.J. Wilcox
Abstract: In this paper, the nature of intelligence is explored briefly and artificial intelligence in introduced. Three techniques that can allow a computer to achieve limited intelligence are discussed and some examples are presented.
Keywords: Intelligent systems; Expert systems; Fuzzy systems; ANNs; Genetic Algorithms. Pages: 2 – 6. Refs: 6.

Optimisation of Heat Treatment Parameters using Artificial Intelligence Techniques

Z. S. Chong, S.J. Wilcox and J. Ward
Abstract: This paper describes the work undertaken to apply ANNs to model the cold alloy-steel bars and the heat treatment parameters with their end-product quality characteristics. Standard multi-layered feed forward ANNs were employed to represent the functional mapping of inputs such as physical dimension, material composition and the parameters of the heat treatment cycles to the Brinell Hardness (HB)and Ultimate Tensile Strength (UTS). The ANNs were then integrated into a Genetic Algorithm (GA) search strategy to identify the best material characteristics and furnace operating parameters in order that both HB and UTS are maximised. The results demonstrated that such a hybrid strategy can deliver sensible results
Keywords: Heat treatment; Cold-alloy steel bars; ANNs; GA; Hybrid strategy. Pages: 7 – 14. Refs: 18.

Induction Motor Fault Detection and Diagnosis using Artificial Neural Networks

L. Li and C.K. Mechefske
Abstract: This paper investigates induction motor fault detection and diagnosis using ANN. These techniques include feed-forward back-propagation networks (FFBPN) and self organizing maps (SOM), used individually and in combination. Common motor faults such as bearing faults, stator winding fault, unbalanced rotor and broken bars are considered. The ANNs were trained and tested using dynamic measurements of stator currents and mechanical vibration signals. The effects of different network structures and the training set sizes on the performance of the ANNs are discussed. In addition, incipient motor fault detection has been investigated.
Keywords: Induction motors; Fault detection and diagnosis; Feed-forward networks; Self-organizing maps networks. Pages: 15 – 23. Refs: 17.

How to Diagnose the Wear of Rolling Element Bearings based on Indirect Condition Monitoring Methods

A. Jantunen
Abstract: Detection of wear of the components in rotating machinery is usually based on indirect methods. The prediction of how much time there is left to maintain these components before they break down is not easy. This paper attempts to tackle this issue related to rotating machinery and in particular to rolling element bearings. The developed approach starts from the idea of modelling the wear of the component. A limited number of condition monitoring parameters are used for diagnosis of the fault. These parameters are used as input in the form of higher order polynomial regression functions with a limited number of terms. The regression functions can give prognosis of the development of the fault. The severity of the situation is analysed using simplified fuzzy logic.
Keywords: Wear; Rolling element bearings; Indirect condition monitoring; Higher polynomial regression. Pages: 24 – 38. Refs: 40.

Constrained Optimisation of Pulverised Coal Fired Boilers using Genetic Algorithms and Artificial Neural Networks

C.K. Tan, S. Kakietek, S.J. Wilcox and J. Ward
Abstract: This paper examines an approach to optimise the combustion air distribution to a 200MW pulverised coal fired boilers using ANNs and Genetic Algorithm (GA). The nitrogen oxide (Nox) and carbon monoxide (CO) emissions as well as the carbon burnout characteristics of the boiler were investigated through a series of field experiments. On the basis of experimental results, three ANNs were used to model the NOx and CO emissions and the unburned carbon (UBC) in the fly ash of the boiler respectively. Once trained, the ANNs were coupled with a GA to determine the optimum solution of the ANN models. Minimum human efforts and little insight into the details of the combustion mechanisms within the boiler are required to generate the optimal settings of the combustion air distribution. The combinations of ANNs and a GA represents a technically plausible approach to determine the boiler air settings for improved combustion behaviour.
Keywords: Genetic algorithm; ANNs; Optimisation; Boiler; Coal combustion. Pages: 39 – 46. Refs: 9.