Design of self-diagonostic and prognostic system

MASHAVA D – N0070233C IME 2015/168

MASHAVA D – N0070233C IME 2014/15 1
MASHAVA D – N0070233C IME2014/15 i


Design of a Self-Diagnostic and Prognostic System

national university of science and technology
(think in OTHER TERMS)

FACULTY OF INDUSTRIAL TECHNOLOGY
DEAPARTMENT OF INDUSTRIAL AND MANUFACTURING ENGINEERING

Title: Design of a Self-diagnostic and prognostic system
Case Study of CAFCA Limited Casting Plant

AUTHOR : MASHAVA DESTINE
STUDENT NUMBER : N0070233C
SUPERVISOR : Eng. L. NYANGA
DATE OF SUBMISSION : ----- MAY 2015

DECLARATION
I,………………………………………………………………………………..declare that:
1. The research reported in this dissertation, except where otherwise stated, and is my original work.
2. This dissertation has not been submitted for examination for any degree at this or any other university.
3. This dissertation does not contain other persons' writing, data, pictures or graphs except where acknowledged as having been sourced from the other persons. Where written sources have been used:
i. The words are paraphrased and the information attributed to the source through referencing;
ii. With exact words quoted, the words quoted are placed into quotation marks and referenced.

DECLARATION OF COPYRIGHT
I hereby grant permission to the National University of Science and Technology Library to reproduce copies of this dissertation and to lend or sell such copies for private, scholarly or scientific research purposes only. I reserve other publication rights. No extensive extracts from this dissertation may be printed in any form or otherwise reproduced without my written permission.
SIGNATURE: …………………………………
PERMANENT ADDRESS: …………………………………………….
PLACE NAME: ………………………………………………
DATE: ..............................................................






DEDICATIONS
This project is dedicated to my mother and sister who have been instrumental in my pursuit for an engineering profession.









ACKNOWLEDGEMENTS
The author greatly appreciates the following:
National University of Science Technology, Department of industrial and Manufacturing Engineering for its unequalled support, supervision and encouragement through the author's project supervisors Eng. L Nyanga.
The author's family and friends for the advice, encouragement and support throughout the project research.
Lastly all Glory to the Lord Jesus and His Almighty God by affording me the opportunity to pursue engineering to this far.


.















ABSTRACT
Machine fault diagnostic and prognostic techniques have been the considerable subjects of condition-based maintenance system in the recent time due to the potential advantages that could be gained from reducing downtime, decreasing maintenance costs, and increasing machine availability. Therefore, a failure in industrial equipment results in not only the loss of productivity but also timely services to customers, and may even lead to safety and environmental problems. This emphasizes the need of maintenance in manufacturing operations of organizations. Thus, maintenance is of great importance in keeping availability and reliability levels of production facilities enabling them to sustain the right product quality (Su &Chong, 2007). Therefore, this project discusses the design of a self-diagnostic and prognostic tool that is able to monitor a casting plant and offer maintenance decision support based on the current operating status of the plant. This is going to be achieved by designing a self-diagnostic and prognostic system making use of various sensors, the Siemens S7-1200 PLC and the OPC Server. Multi-agents modeled using the Agent Unified Modeling Language .The tool uses a distributed Multi-Agent System comprising of autonomous agents developed in JADE, which serve the function of monitoring plant parameters, will carry out monitoring of the plant. Prognosis in this project is based on the Remaining Useful Life (RUL) concept and is designed using JESS software. This emerging technology has the benefit of improved plant performance, reduced downtime and low maintenance costs.










TABLE OF CONTENTS

DECLARATION i
DEDICATIONS ii
ACKNOWLEDGEMENTS iii
ABSTRACT iv
TABLE OF CONTENTS v
TABLE OF FIGURES vii
LIST OF TABLES viii
Chapter 1 : INTRODUCTORY CHAPTER 1
1.0 Introduction 1
1.1 Aim 1
1.2 Objectives 1
1.3 Scope of the Project 2
1.4 Definition of Critical Terms 2
1.5 Background 3
1.6 Justification 3
1.7 Benefits 4
1.8 Project Methodology 5
1.10 Time plans 7
1.11 Summary 7
Chapter 2 : LITERATURE REVIEW 8
2.0 Introduction 8
2.1 Channel induction furnace and auxiliary equipment 8
2.1.1 Induction heating 9
2.1.2 Cooling water system 9
2.1.4 Refractory Lining 10
REFERENCES 12























TABLE OF FIGURES

Figure 1.1 Hardware monitoring system overview. 6
Figure 1.2 Complete structure of rule base system for the diagnostic and prognostic tool. 6
Figure 2.1 Channel Induction Furnace (courtesy of www .Blogspot, 2008) 9
Figure 2.2 Different approaches to condition monitoring (Bengtsson, 2007) 12
Figure 2.3 Expert system architecture (Bulatovic, 2003) 14
Figure 2.4 Prognostic as a prediction and assessment process 15















LIST OF TABLES

Table 1. Project Gantt chart 7


: INTRODUCTORY CHAPTER
Introduction
Machine sudden failure has greatly contributed to the failure of the Zimbabwean industry to operate to its full capacity as this cause is mainly attributed to the employment of traditional maintenance methods. CAFCA Limited, which is a sole cable manufacturing company in Zimbabwe, employs traditional maintenance methods, which include reactive and preventive maintenance, and these systems are labour intensive and less effective in predicting machine failure. Therefore sudden casting plant equipment failure would definitely results in copper products output culminating in reduced profits and loss of market share failure due to increased competition from cheap cable imports. Hence, this implies the need of CAFCA Limited to be flexible in the segment it operates if it is to survive the never-ending competition for customer satisfaction amongst its players. The environment is now dynamic and technology can now substitute the manual and time-consuming systems with an intelligent system, which gives a real time plant state. Hence, through a self-diagnostic and prognostic tool plant equipment monitoring can be executed without human interference and this is achieved by using a multi-agent system based on condition-monitoring there by enabling to predict and prevent failures based on the current and past behavior of the equipment, thus ensuring its maintenance only when needed and exactly when needed.
Aim
The aim of this project is to;
Design of a self-diagnostic and prognostic system for CAFCA Limited casting plant equipment to improve the plant availability.
Objectives
To enable project accomplishment, the author came up with the following
Objectives;
Design of copper casting furnace-monitoring system using sensors, an OPC Server and Siemens S7-1200 PLC.
Predict plant availability using the Remaining Useful Life (RUF) methodology.
Develop a multi-agent system (MAS) using the Agent Unified Modeling Language (AUML) methodology.
Creating a knowledge base for the agents comprised of a rule base and RUF predictions using JADE and JESS.
Development of a user interface based on the distributed agent using Java programming language.
Scope of the Project
CAFCA Limited Casting plant consist of a large number of components, focusing on the whole system would be tedious and time consuming. Therefore, the project scope is constrained on the components whose failure would greatly affect casting plant production output .Therefore design of a self-diagnostic and prognostic system as a modern day plant equipment maintenance tool, would be based on analysis of the operating parameters of the Induction furnaces. Hence, the parameters, to be monitored, include furnace shell temperature, inductor water flow rate and temperature, generated gas pressure, inductor power consumption, molten copper level.
Induction furnaces and the supporting equipment are the backbone in the manufacture of cables, thereby equipment failure results in higher maintenance costs and reduced production output thus culminating in loss of customers through failure to satisfy demand. Hence, the author has put focus on the furnaces whose failure directly affects the organization's profitability if the signs of failure are not observed and corrected in time.
Definition of Critical Terms
Intelligent Agent – is a hardware and/or software based computer system displaying the properties of autonomy, social adeptness, reactivity, and proactivity (Russell, Stuart J. (2003).
Condition Monitoring – is concerned with the detecting and distinguishing faults occurring in a plant that is being monitored (Mangina, et al., 2000).
Design - To design is either to formulate a plan for the satisfaction of a specified need or to solve a problem (Nisbett, 2006).
Decision Support - the assistance for, substantiation and corroboration of, an act or result of deciding; typically this deciding will be a determination of an optimal or best approach (Gachet, A. 2004)
Diagnostics- Involves a complex reasoning activity, where Artificial Intelligence techniques are applied and these techniques use association rules, reasoning and decision making processes as would the human brain in solving diagnostic problems (Angeli & Atherton,2001).
Prognostics- is an engineering discipline focused on predicting the time at which a system or a component will no longer perform its intended function (Vachtsevanos, et al, 2006).
Background
CAFCA Limited is a cable a sole cable manufacturing company which February 2015 boosted its output tonnage from 200 tonnes to 450 tonnes per month on copper-based products. Hence, this increase in output was possible through installing molten furnace of 7 tonnes capacity, which enabled to increase the daily output of copper from 6 tonnes to 15 tonnes in 24 hours. Reactive and preventive maintenance are the main maintenance approaches implemented at CAFCA Limited followed by condition-based maintenance, applied, to specific equipment. Therefore application of reactive maintenance, when equipment fails, results in both high production costs and significant service downtime caused by equipment and process breakdowns. Whilst Preventive maintenance is implemented to eliminate machine or process breakdowns and downtimes through, maintenance operations scheduled regardless of the actual state of the machine or process. Therefore, in contemporary markets, it becomes increasingly important to predict and prevent failures based on the current and past behavior of the equipment, thus ensuring its maintenance only when needed and exactly when needed.
Hence, the traditional maintenance methods implemented at CAFCA Limited are a drawback to the elimination of equipment sudden failure. Therefore, the implementation of self-diagnostic and prognostic system, which is based on condition monitoring and data analysis for decision support, would result in increased equipment productivity through the elimination of sudden equipment failure.
Justification
Sudden failure of the 7 tonnes induction Furnace at CAFCA Limited on September 2015 just after 7 months of installing the furnace whose life span is 18 months before decommission resulted in the depletion of the production output by 66%.The failure was as a result of inductor overheating culminating in the inductor blowing up. The furnace blew up because of lack of a proper monitoring tool between the molten copper level and the thermocouple position, which govern the inductor operating temperature. Hence, exposure of the thermocouple during the casting process above the molten copper gave a false signal on the inductor power input system to remain at high power during casting process thereby making the inductor to overheat until it exploded into flames. Inductor overheating instances observed whilst the equipment was in service were mainly attributed to clogged cooling water pipes reducing the cooling water flow rate and sudden failure of cooling tower components such as the air drafting fans and water circulating pumps. Wear of furnace lining also results in increased shell temperatures, which also need to be continuously, monitored so as decision can be quickly before the furnace comes to sudden failure.
Therefore, by Implementing of a self-diagnostic tool, advance fault predictions that are likely to occur on the plant equipment based on its current condition can be computed. This has the advantage of eliminating the scenarios where the induction furnace suddenly fails or overheats due to cooling equipment failure .Overheating of the inductor results in reduced overall furnace life and loss useful energy resulting in increased running costs. Failure of the 7 tonne Induction Furnace resulted in the depletion of production output by almost 66% that is from a monthly target of 450 tonnes to 200 tonnes thereby affecting the company's ability to meet customer's orders. Production output recovery takes almost 2 months as this is constrained by the furnace building and equipment acquisition in which all the required equipment is imported resulting in over $60 000 of maintenance costs.
Another factor as to why a self-diagnostic tool is required at the plant is that the data capturing method used at CAFCA Limited is through of log-sheets, in which data capturing is done manually only once per day. Hence, this data is rarely analyzed to check the behavior of machinery over the period of operation. Therefore, with a self-diagnostic tool, analysis is executed further and the data is compared against the key performance indicators of the parameters so that smart decisions are made.
Benefits
Benefits to be realized through this equipment maintenance tool when adopted include;
Improved safety and quality conditions
Improved transparency of the decision process and permits the effects of uncertainty on the decision to be quantitatively addressed.
Reduced maintenance cost through the reduction of maintenance frequency.
Improved consistency in production output.
Improved equipment life.
Permanent storage of data which will be easily accessible either for analysis or when required by an interested third part.
Project Methodology
The methodology used by the author will take a quantitative approach in designing the self-diagnostic and prognostic tool and this involves;
Data collection and Analysis
Data will be collected and analyzed in order to establish the relationship between monitored parameters. The data used in this project will be collected from the Operator's maintenance checks log-sheets, test results and failure reports. The parameters to be analyzed against time include;
Inductor shell temperature
Furnace power consumption
Furnace shell temperature.
Molten copper level
Molten copper temperature
Cooling water flow rate
Cooling water inlet and outlet temperature
Cooling water purity
i. Designing a hardware-monitoring model by incorporating relevant sensors connected to the Siemens PLC and the OPC Server for transferring data to the Data Source as shown on Figure 1.1 below and on figure, 1.2 is the complete system architecture

Figure 1.1 Hardware monitoring system overview.
ii. Predicting plant availability using the Remaining Useful Life (RUF) methodology
iii. Designing the software model of the decision support by using JADE for creating the multi agent system and incorporating the rule base required for decision support using JESS.

Figure 1.2 Complete structure of rule base system for the diagnostic and prognostic tool.
iv. Providing results of the running system.
v. Recommendations for stabilizing the system
vi. Suggestions required for optimizing the decision support tool.









Time plans
Table 1 below outlines the project Gant chat as planned for project implementation.
Table 1. Project Gantt chart

Project Time line

Project Task
October
November
December
January
February
March

April
May
Project proposal








Chapter 1: Introduction








Chapter 2: Literature review








Chapter 3: System development








Chapter 4: System and Hardware designing
Chapter 5:Results








Chapter 6: Costing








Chapter 7: Recommendations and Conclusions








Project Documentation









Summary
The chapter has introduced the reader to the project and how the author seeks to carry out the project. Therefore, the project aim and objectives were also mentioned in this chapter thereby familiarizing the reader on how the project is to be accomplished. After stating the problem, justification of the solution was given along with the benefits of implementing the project. With this information, the reader can then prepare for the review of the relevant literature that follows this chap
: LITERATURE REVIEW
Introduction
All the relevant literature used by the author for the project is included in this chapter. The author gathered literature on the operation of a channel induction furnace and the effect of operational parameters variation on the overall equipment life. Included in this chapter are the appropriate systems on self-diagnostics used to monitor induction furnaces parameters. The selection of sensors is also reviewed and so is the design of self-diagnostic and prognostic tool. The PLC, Java development framework used, the algorithms used for decision making and the standards used by the author are explained in this chapter. The model used throughout this entire project will be that of a channel induction furnace manufactured by Outokoump Finland, which is used at CAFCA Limited for copper melting operations.
Channel induction furnace and auxiliary equipment
The channel induction furnace consists of a refractory lined steel shell, which contains the molten metal when it is attached to the steel shell and connected by a throat is an induction unit, which forms the melting component of the furnace. The induction unit consists of an iron core in the form of a ring around which a primary induction coil is wound. This assembly forms a simple transformer in which the molten metal loops comprises the secondary component. The heat generated within the loop causes the metal to circulate into the main well of the furnace. The circulation of the molten metal effects a useful stirring action in the melt. Channel induction furnaces are commonly used for melting low melting point alloys and or as holding and superheating unit for higher melting point alloys such as cast iron. These furnaces consist of a vessel to which one or more inductors are attached. The water is fed by means of cooling water hoses, and ducted away by the water-cooled cables used for the energy supply. The individual cooling water circuits are monitored with regard to volume flow rate and temperature. (Su & Chong 2007).

Figure 2.1 Channel Induction Furnace (courtesy of www .Blogspot, 2008)
Induction heating
An induction heating or melting system consists of highly conductive, heavy-wall copper tubing that is specially wound into a coil. This coil is either free standing for heating applications or placed inside of a furnace that acts like a melting pot. The design and construction of the coil is critical for achieving maximum electrical efficiency and optimum performance for each specific heating or melting application. The induction power supply generates a high electrical current that travels to the coil through flexible water-cooled power cables and/or copper buss bar. The electrical current supplied is at the proper output frequency and voltage to match the design of the melting furnace or heating coil. An electromagnetic field is then created within the center area of the copper coil and reacts with the metal that is placed in the center of the coil, thus heating it to the desired temperature. The output wattage and frequency play a factor in how hot and how fast the metal is heated as well as the depth of heat penetration. Therefore, the power supply is also custom-designed according to the heating or melting application (Rudnev et al, 2003).
Cooling water system
The operation of an induction furnace system requires a cooling water system due to the high amounts of electricity that generate a tremendous amount of heat. Therefore, water is recirculated throughout the induction system by a water pumping system and is cooled by a water-cooling tower. In the converter, including the capacitors and the smoothing choke, the water circulating in the circuit is heated up from approx. 34 °C to 38 °C, and must be cooled down again to 34°C by a cooling system activator. Thus, approximately 215 l/h must be pumped through the electrical equipment per kW of performance loss. In the furnace coil, the water is heated up from approx. 35 °C to 62 °C, and must be cooled back down to 35 °C by a separate cooling system. Hence, approximately 32 l/h have to be pumped through the system per kW of performance loss. In the event of a power failure or cooling system failure, an emergency water supply must be installed for the furnace circuit (Netzel H, 2009).
Water Purification System
Calcium ions Ca2+ and hydro carbonate ions HCO3- are important components of distilled water and are present in the latter because of the dissociation of calcium carbonate by the reaction:
H2O + CO2 + CaCO3 Ca2+ (HCO3-)2
The greater the number of Ńŕ2+, ions in water, the harder the water. In accordance with this equation, hard water can be made softer through the precipitation of CaC03. Under the influence of changes in temperature or pressure, calcium in the water forms solid deposits on the walls of pipes. The formation of such deposits increases pressure and decreases heat transfer, which in turn makes the cooling operation less efficient. In addition to Ca2+ and HCO3-, the coolant water contains the cations Mg2+, Na+, Fe2+, and anions SO42-, NO3-, Cl-, hence a dual bed deionizer or electrical energy is used to shift chemical equilibrium of the hard water. Therefore it is necessary to monitor water conductivity and the total dissolved substances (T.D.S) so as to prevent damage of the furnace inductor through clogging of the inductor water cooling pipes by solid deposits caused by hard water( Hricisin& Mudron,2006). Hence, the standard T.D.S of make-up water after deionization should be 5 and for circulating cooling water should be 250 .The conductivity of the make-up water after deionization should be in the range of 0-7 µse/cm and circulating cooling water should be maintaining in the range 250-350 µse/cm.
Refractory Lining
Furnace refractory lining is made from inorganic nonmetallic material, which can withstand high temperature without undergoing physical or chemical changes while remaining in contact with molten slag, metal and gases (Gupta O, 2002). The furnace refractory lining is often exposed to thermal shock during skimming, cleaning, fluxing and charging. The severity of these forces increases sharply with increasing furnace size and varies widely with operation practice. Mechanical shock and abrasion are severe where large quantities of cold scrap are directly charged with solid materials through the main door or through doors on one sidewall or are charged into a well containing molten metal. All this practice leads to massive energy losses and consequently, stresses generates in the lining that can cause damage of the refractory lining and eventually this would result in great loss of production time, equipment, and sometimes the product itself. (Schacht, 1995).Therefore monitoring of refractory Linings by real-time measurement of the temperature distribution of the exterior lining surface. Therefore, this enables protection of the induction coil against overheating and, the more so, against contact with molten metal is of vital importance for ensuring safe and reliable operation of an induction furnace. Crucible cracks and erosion are detected and localized reliably and precisely and normal refractory wear is under close control (Schmitz et al, 2006).
Condition Based Maintenance
Condition based maintenance is a maintenance type that utilizes on-condition tasks in order to monitor the condition over time and usage. It is done in order to give input to decide maintenance actions dynamically. A central part of condition-based maintenance is thus monitoring, often called condition monitoring. Condition monitoring can be performed using a number of various approaches and utilizing different levels of technology as shown on Figure 2.2.Hence, several fault detection and diagnostic technologies are employed ranging from simple thresholding to rule-based algorithms. However, these technologies have not specifically focused on the ability to predict the future condition (prognostics) of a machine based on the current diagnostic state of the machinery and its available operating and failure history data. An advanced prognostic capability is desired because the ability to forecast this future condition enables a higher level of condition-based maintenance for optimally managing total life cycle costs (LCC), the common point being that the activity is normally performed in an operating state (Jardine et al,2006).


Figure 2.2 Different approaches to condition monitoring (Bengtsson, 2007)
A number of different techniques exist to measure the condition of an item. Depending on the type of potential failure condition one is set out to measure, one or more techniques can be utilized. Moubray (1997), together with Tsang (1995) classify condition monitoring techniques according to the symptoms they are designed to detect, dynamic effects, such as vibration and sound, particles released into the environment, chemicals released into the environment, physical effects, such as cracks, fractures, wear, and deformation, temperature rises in the equipment, and electrical effects, such as resistance, conductivity, dielectric strength, etc.
Decision Support in Condition Based Maintenance
Traditional methods of decision support still exist for condition-based maintenance and usually these influence how long maintenance will take. Moubray (1997) and Starr (1997), point out that it is important that condition based maintenance is applied where it is appropriate, not as an overall policy. This is because many techniques are expensive, and it would not be cost effective. For the decision support, some companies use own knowledge and experience when choosing a maintenance strategy while others opt for time to failure modelling. Most companies use decision support tools such as Decision Trees, FMECA, Statistical Process Control and From/To Charts for their decision support. Technologically incompetent organizations use cheap but quantitative methods such as major overhauls, monthly reports and manufacturer recommendations. A new era dawns on decision support from agents; this is still a developing field whereby autonomous agents can offer decision support. This approach abandons all these traditional approaches thereby increasing an organizations flexibility.
Diagnostic Expert System
Involves a complex reasoning activity, it involves the application Artificial Intelligence techniques. These techniques use association rules, reasoning and decision making processes as would the human brain in solving diagnostic problems (Angeli & Atherton, 2001).Therefore, if the system is not functioning correctly, the algorithm should be able to determine, as accurately as possible, which part of the system is failing, and which kind of fault it is facing. The computations are based on observations, which then provide information on the current behavior. Hence, the heuristic method comprises the procedures, in principle; imitate the experts operations, their reasoning line based on the data correlated to their knowledge. This is a form of artificial intelligence. In software sense of meaning, artificial intelligence is first developed in the form of expert systems as shown on figure 2.3.

Figure 2.3 Expert system architecture (Bulatovic, 2003)
Prognostics
Prognostics is the process of predicting the future state of a system In this acceptation, prognostic is also called the "prediction of a system's lifetime" as it is a process whose objective is to predict the remaining useful life (RUL) before a failure occurs given the current machine condition and past operation profile. Prognostics systems comprise sensors, a data acquisition system, and microprocessor-based software to perform sensor fusion, analysis, and reporting/interpreting of results with little or no human intervention, in real time or near real time. Prognostics systems comprise sensors, a data acquisition system, and microprocessor-based software to perform sensor fusion, analysis, and reporting/interpreting of results with little or no human intervention, in real time or near real time. A prognostic system should development is based on assessment criteria, whose limits depend on the system itself and on performance objectives. The prognostic consist of split two sub-activities: a first one to predict the evolution of a situation at a given time, and a second one to assess this predicted situation about an evaluation referential Figure 2.4. When a central problem is pointed out: the accuracy of a prognostic system is related to its ability to approximate and predict the degradation of equipment; and hence the prediction phase is a critical one (Jardine et al, 2006).

Figure 2.4 Prognostic as a prediction and assessment process
Condition monitoring and prognostic techniques for induction furnaces


















REFERENCES
Angeli, C. and Atherton, D.P. 2001. A Model Based Method for an on-line Diagnostic Knowledge-Based System Expert Systems, 18(3):150-158.
Bala, K.C "Design Analysis of an Electric Induction Furnace for Melting Aluminum Scrap", AU J.T. 9(2): 83-88 (Oct. 2005), Pages 83-88
Bengtsson, M., 2007. Decision-Making during Condition Based Maintenance Implementation. Faro, Portugal, 20th International Conference on Condition Monitoring and Diagnostic Engineering Management.
Bulatovic M., Expert System in Function of Technical Diagnostics, 7 International Research/Expert Conference "TMT 2003, Loret de Mar, 2003, Loret de Mar, Barcelona - Spain, 15-17 September, 2003
Beebe, R. (1978), ``Recent experience with condition monitoring of steam turbines by performance analysis'', Transactions of Mechanical Engineering, Institution of Engineers, Australia.
Djurdjanovic, D., Lee, J. and Ni, J., "Watchdog Agent – An Infotronics Based Prognostics Approach for Product Performance Assessment and Prediction", International Journal of Advanced Engineering Informatics, Special Issue on Intelligent Maintenance Systems, Vol. 17, No.3-4, pp. 109–125, 2003.
Davies, J.T. Journal of the South African Institute Of Mining and Metallurgy, July 1976
Gupta O.P, 2002 Elements of Fuels, Furnaces and Refractories, Khanna Publishers
Hricisin, C and Mudron, J. (2006). The Refractory Lining of Blast Furnaces and Modernization of Their Cooling System
A.K.S. Jardine, D. Lin and D. Banjevic, "A review on machinery diagnostics and prognostics implementing condition-based maintenance", Mech. Sys. & Sig. Pro., Vol. 20, pp. 1483-1510, 2006.
Lee, J. & Wang, Ben, (1999) Computer-aided Maintenance: methodologies and practices, Kluwer Academic Publishing
L.Swanson, The impact of new production technologies on the maintenance function: an empirical study. International journal of production research, vol.37, No.4, pp849-869, 1999.
Nisbett, B., 2006. Shigley's Mechanical Engineering Design. 8th ed. USA: McGraw-Hill.
Netzel H.H Manual for safe induction furnace operation, 2009
Rudnev V., Loveless D., Cook R., (2003). Handbook of Inductive Heating, Manufacturing Engineering and Material Processing.
Russell, S. & Norvig, P., 1995. Artificial Intelligence: A Modern Approach. Eaglewood
Cliffs, USA: Prentice-Hall.

Su, H. and Chong, K.T., 2007, "Induction machine condition monitoring using neural network modeling," IEEE Trans.Industrial Electronics, Vol. 54, No. 1, pp. 241-249.
Schacht C. A, Refractory Linings: Thermodynamically Design and Applications, M. Dekker, New York, p. 201 (1995)
Schmitz Wilfred , Frank Donsbach and Henrik Hoff.,2006 Development and use of a new optical sensor system for induction furnace crucible monitoring. Year: 2006 " Volume: 2 ". Conference: 67th World Foundry Congress Paper Lios Technology GmbH, Germany.
Thurston, M. and Lebold, M., 2001, "Open Standards for Condition Based Maintenance and Prognostic Systems", Pennsylvania State University, Applied Research Laboratory
Vachtsevanos; Lewis, Roemer; Hess, and Wu (2006). Intelligent fault Diagnosis and Prognosis for Engineering Systems. Wiley. ISBN 0-471-72999-X.



















Industrial Attachment Report-CAFCA LIMITED

The project is submitted in partial fulfilment of the requirements of the BEng (Hon) in
Industrial and Manufacturing Engineering program at the National University of Science and
Technology, Zimbabwe.


Design of a Self-Diagnostic and Prognostic System


16