INVESTIGATION ON PERFORMANCE RELIABILITY …€¦ · v Course-work Completion Certificate This is to certify that Mr. Pancholi Nilesh Hasamukhlal, enrolment no. 129990919010 is a

INVESTIGATION ON PERFORMANCE RELIABILITY

IMPROVEMENT BY OPTIMIZING MAINTENANCE

PRACTICES THROUGH FAILURE ANALYSIS IN

CONTINUOUS PROCESS INDUSTRY

A Thesis submitted to Gujarat Technological University

for the Award of

Doctor of Philosophy

in

Mechanical Engineering

by

Pancholi Nilesh Hasamukhlal

129990919010

under supervision of

Dr. Mangal G. Bhatt

GUJARAT TECHNOLOGICAL UNIVERSITY

AHMEDABAD

September 2019

ii

© Pancholi Nilesh Hasamukhlal

iii

DECLARATION

I declare that the thesis entitled “Investigation on Performance Reliability Improvement

by Optimizing Maintenance Practices through Failure Analysis in Continuous

Process Industry” submitted by me for the degree of Doctor of Philosophy

is the record of research work carried out by me during the period from October 2012 to

November 2017 under the supervision of Dr. Mangal G. Bhatt, Principal, Shantilal

Shah Engineering College, Bhavnagar and this has not formed the basis for the award

of any degree, diploma, associate ship, fellowship, titles in this or any other University or

other institution of higher learning.

I further declare that the material obtained from other sources has been duly acknowledged

in the thesis. I shall be solely responsible for any plagiarism or other irregularities, if

noticed in the thesis.

Signature of the Research Scholar: ………………………… Date:

Name of Research Scholar: Pancholi Nilesh Hasamukhlal

Place: Ahmedabad

iv

CERTIFICATE

I certify that the work incorporated in the “Investigation on Performance Reliability

Improvement by Optimizing Maintenance Practices through Failure Analysis in

Continuous Process Industry” submitted by Shri Pancholi Nilesh

Hasamukhlal was carried out by the candidate under my supervision/guidance. To

the best of my knowledge: (i) the candidate has not submitted the same research work

to any other institution for any degree/diploma, Associate ship, Fellowship or other

similar titles (ii) the thesis submitted is a record of original research work done by the

Research Scholar during the period of study under my supervision, and (iii) the thesis

represents independent research work on the part of the Research Scholar.

Signature of Supervisor: ……………………………… Date:

Name of Supervisor: Dr. Mangal G. Bhatt

Place: Ahmedabad

v

Course-work Completion Certificate

This is to certify that Mr. Pancholi Nilesh Hasamukhlal, enrolment no. 129990919010 is a PhD

scholar enrolled for PhD program in the branch Mechanical Engineering of Gujarat

Technological University, Ahmedabad.

(Please tick the relevant option(s))

He/She has been exempted from the course-work (successfully completed during M.Phil

Course)

He/She has been exempted from Research Methodology Course only (successfully

completed during M.Phil Course)

He/She has successfully completed the PhD course work for the partial requirement for the

award of PhD Degree. His/ Her performance in the course work is as follows-

Grade Obtained in Research Methodology Grade Obtained in Self Study Course (Core Subject)

(PH001) (PH002)

AB AB

Supervisor’s Sign

(Dr. Mangal G. Bhatt)

vi

Originality Report Certificate

It is certified that PhD Thesis titled “Investigation on Performance Reliability

Improvement by Optimizing Maintenance Practices through Failure Analysis in

Continuous Process Industry” by Mr. Pancholi Nilesh Hasamukhlal has been examined

by us. We undertake the following:

a. Thesis has significant new work / knowledge as compared already published or are under

consideration to be published elsewhere. No sentence, equation, diagram, table,

paragraph or section has been copied verbatim from previous work unless it is placed

under quotation marks and duly referenced.

b. The work presented is original and own work of the author (i.e. there is no plagiarism).

No ideas, processes, results or words of others have been presented as Author own work.

c. There is no fabrication of data or results which have been compiled / analysed.

d. There is no falsification by manipulating research materials, equipment or processes, or

changing or omitting data or results such that the research is not accurately represented in

the research record.

e. The thesis has been checked using Turnitin Plagiarism (copy of originality report

attached) and found within limits as per GTU Plagiarism Policy and instructions issued

from time to time (i.e. permitted similarity index <=10%).

Signature of the Research Scholar: …………………………… Date:


Place: Ahmedabad

Signature of Supervisor: ……………………………… Date:

Name of Supervisor: Dr. Mangal G. Bhatt

Place: Ahmedabad

vii

Turnitin Originality Report

viii

PhD THESIS Non-Exclusive License to

GUJARAT TECHNOLOGICAL UNIVERSITY

In consideration of being a PhD Research Scholar at GTU and in the interests of the

facilitation of research at GTU and elsewhere, I, Pancholi

Nilesh Hasamukhlal, having Enrollment No. 129990919010 hereby grant a non-exclusive,

royalty free and perpetual license to GTU on the following terms:

a) GTU is permitted to archive, reproduce and distribute my thesis, in whole or in part,

and/or my abstract, in whole or in part ( referred to collectively as the “Work”)

anywhere in the world, for non-commercial purposes, in all forms of media;

b) GTU is permitted to authorize, sub-lease, sub-contract or procure any of the acts

mentioned in paragraph (a);

c) GTU is authorized to submit the Work at any National / International Library, under

the authority of their “Thesis Non-Exclusive License”;

d) The Universal Copyright Notice (©) shall appear on all copies made under the

authority of this license;

e) I undertake to submit my thesis, through my University, to any Library and Archives.

Any abstract submitted with the thesis will be considered to form part of the thesis.

f) I represent that my thesis is my original work, does not infringe any rights of others,

including privacy rights, and that I have the right to make the grant conferred by this

non-exclusive license.

g) If third party copyrighted material was included in my thesis for which, under the

terms of the Copyright Act, written permission from the copyright owners is required,

I have obtained such permission from the copyright owners to do the acts mentioned

in paragraph (a) above for the full term of copyright protection.

h) I retain copyright ownership and moral rights in my thesis, and may deal with the

copyright in my thesis, in any way consistent with rights granted by me to my

University in this non-exclusive license.

i) I further promise to inform any person to whom I may hereafter assign or license my

copyright in my thesis of the rights granted by me to my University in this non-

exclusive license.

ix

j) I am aware of and agree to accept the conditions and regulations of PhD including all

policy matters related to authorship and plagiarism.

Signature of the Research Scholar: ……………………….


Date: Place: Ahmedabad

Signature of the Supervisor: ……………………….

Name of the Supervisor: Dr. Mangal G. Bhatt

Date: Place: Ahmedabad

Seal:

x

Thesis Approval Form

The viva-voce of the PhD Thesis submitted by Shri Pancholi

Nilesh Hasamukhlal, Enrollment No. 129990919010 entitled “Investigation on

Performance Reliability Improvement by Optimizing Maintenance Practices through

Failure Analysis in Continuous Process Industry” was conducted on

…………………….………… (day and date) at Gujarat Technological University.

(Please tick any one of the following option)

The performance of the candidate was satisfactory. We recommend that he/she be

awarded the PhD degree.

Any further modifications in research work recommended by the panel after 3

months from the date of first viva-voce upon request of the Supervisor or request of

Independent Research Scholar after which viva-voce can be re-conducted by the

same panel again.

(briefly specify the modifications suggested by the panel)

The performance of the candidate was unsatisfactory. We recommend that he/she

should not be awarded the PhD degree.

(The panel must give justifications for rejecting the research work)

----------------------------------------------------- -----------------------------------------------------

Name and Signature of Supervisor with Seal 1) (External Examiner 1) Name and Signature

------------------------------------------------------- -------------------------------------------------------

2) (External Examiner 2) Name and Signature 3) (External Examiner 3) Name and Signature

xi

Abstract

There are certain possibilities of the performance reliability improvement through

maintenance optimization of various systems or components of major process industries

that run on continuous basis. Looking to the past business volume of 6.5 lac metric tons

for aluminium wire and expected compound annual growth rate of 13.5 % during

2014-19 due to Government of India’s power for all drive, an aluminium wire

processing plant forms a noticeable sector of continuous process industry. The

maintenance time is about 20 to 25 % of the total time which leads to reliability losses.

Moreover, poor maintenance seems the prime reason of low productivity and profit. Such

facts and challenges of keeping the system in ready-state motivate a definite maintenance

plan to be modeled based on a live failure analysis to be executed during shutdown or

scheduled period. The purpose of this research study is to investigate the extent at

which the reliability of an aluminium wire rolling mill can be improved by

ameliorating current control and maintenance practices.

The deliverables are achieved by collecting the historical failure data i.e. downtime

and failure frequencies; from April 2013 to March 2014 at Sampat aluminium private

limited. Reliability modeling is done in a view to understand the failure pattern behaviour

of the rolling machine. The critical components like; bearings, gears and machining shafts

are discriminated based on these data and their functional failures, failure causes, effects

and repercussions of failures with existing control and maintenance practices has been

modeled based on live shop-floor study. Scores are assigned on 1 to 10 levels by analyzing

attributes effects from lowest to highest concern respectively for every modes of failure

through realistic brain-storming among maintenance team by incorporating some advanced

attributes like; maintainability, economic safety, economic cost and spares with basic

criteria in this study. The risk priority number (RPN) and maintainability criticality indices

(MCIs) are narrated by these score values through traditional as well as MCDM based

failure analysis models like; TOPSIS, COPRAS-G and PSI.

The primary findings of this research work are to propose improvements in the

maintenance plan of critical components like bearings, gears and shafts of an

aluminium wire rolling mill which are commonly representing the most critical

components in a large range of industrial processes. The common modes of failure (C5,

xii

C3, C4, C10, and C14) having large is covered with condition-based monitoring or

predictive type of approaches, modes of failure (C13, C7, C8, and C1) having medium

is covered with preventive measures where it is assumed that avoidance of failure is

better than restore and modes of failure (C2, C11, C12, C6, and C9) having small is

covered by remedial or corrective actions when breakdown prompts. Originality mainly

consists in the contemporary application of non-identical MCDM methods.

The PhD thesis will be helpful in explicating the drawbacks of maintaining matters of the

foremost processing plants and prescribed yield outputs where MCDM approaches are

advantageous.

xiii

Dedicated to

My Wife (Unnati)

and

My Daughter (Dhanvee)

xiv

Acknowledgement

I am deeply indebted to my supervisor Dr. Mangal. G. Bhatt, Principal, Shantilal Shah

Engineering College, Bhavnagar whose invaluable guidance and constant encouragement

at all stages of the research work provided me with valuable insight without which this

research work would not have been possible. He gave me complete freedom to finish the

work at my own without compromising the standards.

I am thankful to Doctoral Progress Committee members Dr. Kalpesh D. Maniya, C. K.

Pithawala College of Engineering and Technology, Surat and Dr. Harshit K. Dave, S. V.

National Institute of Technology, Surat. Their valuable advices and moral support during

earnest reviews provide me proper direction throughout the research work.

I am especially grateful to Mr. Samyak Deora, Director, Deora Group for permitting me to

do research work at his esteemed industry. I am also thankful to Mr. Shrikant Patel,

Executive Director, Sampat Aluminium Pvt. Ltd., Ahmedabad and his team of maintenance

personnel, managers, and shop floor executives for giving their kind and valuable technical

support in fulfillment of requirements directly or indirectly during this study.

I acknowledge Honorable Vice Chancellor Dr. Navin Sheth, Dr. N. M. Bhatt and Dr. Rajul

Gajjar, Deans (Ph. D.), Shri J. C. Lilani, Registrar and staff members of Ph. D. section for

their kind support. I am also thankful to Dr. S. B. Sharma, former outstanding scientist and

Dy. Director, Antenna system area, SAC, ISRO, Ahmedabad for motivating to dream about

Ph. D. course by igniting research attitude in me and Late Prof. C. C. Rajyaguru, Director

(Academics), Dr. K. M. Srivastav, Ex. Professor, Indus Institute of Technology and

Engineering, Ahmedabad for providing me path to fulfill such dream.

Last but not least, I would like to express my gratitude to all nears and dears who helped

me in the course of my journey to complete this work with their good wishes.

Pancholi Nilesh Hasamukhlal Date:

xv

Table of Contents

List of Abbreviations .......................................................................................................... xix

List of Symbols ................................................................................................................... xxi

List of Figures ................................................................................................................... xxiv

List of Tables ..................................................................................................................... xxv

CHAPTER 1 ......................................................................................................................... 1

Introduction and Literature Review .................................................................................. 1

1.1 Broad Area of Research ................................................................................................... 1

1.2 Overview and Significance of Study ............................................................................... 1

1.3 Brief Overview about Reliability ..................................................................................... 2

1.3.1 Reliability Concept ....................................................................................................... 2

1.3.2 Significance of Mortality (Bath-tub) Curve .................................................................. 3

1.3.3 System Reliability Models ............................................................................................ 4

1.3.3.1 Series networks .......................................................................................................... 4

1.3.3.2 Parallel networks ........................................................................................................ 5

1.3.3.3 Redundant Reliability Model ..................................................................................... 5

1.3.3.4 The x – out of – m structure ....................................................................................... 6

1.3.3.5 System with mixed mode failure ............................................................................... 6

1.3.4 Reliability Analysis through Failure Distribution Function ......................................... 7

1.4 Brief Overview about Maintenance ................................................................................. 7

1.4.1 Maintenance Philosophy ............................................................................................... 8

1.4.2 Systematic step-by-step method for planning maintenance program ......................... 10

1.5 Brief Overview about Failure Mode Effect and Criticality Analysis (FMECA) ........... 11

1.5.1 Concept ....................................................................................................................... 11

1.5.2 Types of FMEA/FMECA ........................................................................................... 12

1.6 Brief Overview about Multi-criteria Decision-making (MCDM) ................................. 12

1.6.1 Overview and Importance of MCDM Approaches ..................................................... 12

1.6.2 Multi-criteria Decision-making Process ..................................................................... 14

1.6.3 Multi-criteria Approaches ........................................................................................... 14

1.7 Historical Background of Reliability Analysis Issues ................................................... 15

1.8 Comprehensive Literatures Review related to Maintenance Planning .......................... 16

1.9 Literatures Review on Multi-criteria Decision-making (MCDM) based FMECA ........ 18

xvi

1.10 Motivation of Research ................................................................................................ 20

1.10.1 Outcome of Literature Review and Research Gap ................................................... 20

1.10.2 Definition of the Problem ......................................................................................... 20

1.11 Objective and the Scope of Work ................................................................................ 22

1.12 Research Approaches ................................................................................................... 24

1.13 Original contribution by the thesis ............................................................................... 26

1.14 Organization of Thesis ................................................................................................. 26

1.15 Summary ...................................................................................................................... 28

CHAPTER 2 ....................................................................................................................... 29

Data Collection, Reliability Modeling and Identification of Critical Components ..... 29

2.1 Overview of Identified Process Industry (Rolling Mill) ................................................ 29

2.1.1 Introduction and Background ..................................................................................... 29

2.1.2 Rolling Process ........................................................................................................... 30

2.1.3 Rolling Mill Components ........................................................................................... 30

2.1.4 Rolling Mill Properzi Process ..................................................................................... 32

2.1.5 Rolling Machine Sub-Components ............................................................................. 32

2.2 Major Reliability and Maintenance Issues during Preliminary Studies and Learnings

from them ............................................................................................................................. 34

2.3 Failure Data Collection and Analysis ............................................................................ 35

2.4 Reliability Modelling ..................................................................................................... 40

2.5 Discrimination of Critical Components of Rolling Mill ................................................ 45

2.6 Summary ........................................................................................................................ 48

CHAPTER 3 ....................................................................................................................... 49

Failure Pattern Study, Criteria Selection, Score Assignment and Traditional

Approach ............................................................................................................................ 49

3.1 Failure Pattern Study of Critical Components through Failure Mode and Effect

Analysis (FMEA) ................................................................................................................. 49

3.1.1 Overview ..................................................................................................................... 49

3.1.2 FMEA for discriminated Critical Components ........................................................... 50

3.2 Selection of Criteria for Criticality Assessment ............................................................ 53

3.2.1 Traditional Criteria...................................................................................................... 53

3.2.2 Advanced Criteria ....................................................................................................... 53

3.3 Score Assignment Methodology .................................................................................... 54

3.3.1 Score Assignment for Traditional Approach .............................................................. 54

xvii

3.3.2 Score Assignment for MCDM based Failure Analysis Models.................................. 56

3.4 Traditional Failure Analysis Approach .......................................................................... 58

3.4.1 Overview ..................................................................................................................... 58

3.4.2 Criticality Assessment based on Risk Priority Number (RPN) .................................. 59

3.4.3 Maintenance Planning Through Traditional FMECA ................................................ 61

3.4.4 Drawbacks of Traditional FMECA ............................................................................. 62

3.5 Summary ........................................................................................................................ 63

CHAPTER 4 ....................................................................................................................... 64

Multi-criteria Decision-making based Failure Analysis Models ................................... 64

4.1 Overview ........................................................................................................................ 64

4.2 TOPSIS based Failure Mode Effect and Criticality Analysis ........................................ 65

4.2.1 TOPSIS Methodology ................................................................................................. 65

4.2.2 Maintenance Planning through TOPSIS FMECA ...................................................... 71

4.3 COPRAS-G based Failure Mode Effect and Criticality Analysis ................................. 73

4.3.1 COPRAS-G Methodology .......................................................................................... 73

4.3.3 Significance of COPRAS-G ....................................................................................... 79

4.3.4 Maintenance Planning through COPRAS-G FMECA ................................................ 79

4.4 PSI based Failure Mode Effect and Criticality Analysis ............................................... 81

4.4.1 PSI Methodology ........................................................................................................ 81

4.4.2 Significance of PSI ..................................................................................................... 86

4.4.3 Maintenance Planning through PSI FMECA .............................................................. 86

4.5 Summary ........................................................................................................................ 88

CHAPTER 5 ....................................................................................................................... 89

Results and Discussion ....................................................................................................... 89

5.1 General Overview .......................................................................................................... 89

5.2 Results of Discrimination Process through Shop-floor Statistics .................................. 90

5.3 Results of Traditional FMECA Model .......................................................................... 92

5.4 Results of MCDM based FMECA ................................................................................. 93

5.5 Suggested Remedies ...................................................................................................... 95

5.6 Summary ........................................................................................................................ 96

CHAPTER 6 ....................................................................................................................... 99

Conclusion and Future Scope ........................................................................................... 99

6.1 General Overview .......................................................................................................... 99

xviii

6.2 Major Concluding Remarks ......................................................................................... 100

6.3 Recommendations for Future Scope of Work ............................................................. 101

6.4 Limitations of Proposed Research Work ..................................................................... 102

6.5 Summary ...................................................................................................................... 102

References .......................................................................................................................... 103

List of Publications ............................................................................................................ 112

xix

List of Abbreviations

AHP: Analytical Hierarchy Process

ANP: Analytical Network Process

BCR: Benefit-to-Cost Ratio

CAGR: Compound Annual Growth Rate

CMMS: Computerized Maintenance Management Systems

COPRAS-G: Grey-Complex Proportional Risk Assessment

DTMM: Delay Time Maintenance Model

ELECTRE: Elimination and Choice Translating Reality

FAHP: Fuzzy Analytic Hierarchy Process

FMEA: Failure Mode Effect Analysis

FMECA: Failure Mode Effect and Criticality Analysis

GA: Genetic Algorithm

GP: Goal Programming

HGA: Hybrid Genetic Algorithm

JIT: Just-in Time

LEPs: Large Engineering Plants

LOLP: Loss of Load Probability

MACBETH: Measuring Attractiveness by a Categorical Based Evaluation

Technique

MAVT: Multi-Attribute Value Theory

MCDM: Multi Criteria Decision Making

MCI: Maintainability Criticality Index

MDT: Mean Down Time

MSF: Modeling System Failure

MTBF: Mean Time between Failure

MTTF: Mean Time to Failure

MTTR: Mean Time to Repair

OJT: On the Job Training

PM: Preventive Maintenance

PROMETHEE: Preference Ranking Organization Method for Enrichment

Evaluation

PSI: Preference Selection Index

xx

QBD: Quasi-Birth–Death

RDCIS: Research & Development Centre for Iron and Steel

RNN: Recurrent Neural Network

RPN: Risk Priority Number

RRE: Reusable Rocket Engine

SA: Simulated Annealing

SAIL: Steel Authority of India Ltd.

TOPSIS: Technique for Order Preference by Similarity to Ideal Solution

TPM: Total Productive Maintenance

TQM: Total Quality Management

WSM: Weighted Sum Model

xxi

List of Symbols

Scores for probability of occurrence (P) for i

th alternative and j

th criteria in

traditional FMECA

: Operational availability

: Inherent availability

Scores for degree of detectability (D) for ith

alternative and jth

criteria in

traditional FMECA

C1: Improper lubrication & defective sealing

C2: Higher speed than specified

C3: Design defects, bearing dimension not as per specification

C4: Foreign matters/particles

C5: Sudden impact on the rolls

C6: Loss of power

C7: Inadequate lubrication - dirt, viscosity issues

C8: Improper meshing, case depth & high residual stresses

C9: Overheating at gear mesh

C10: Excessive overload & cyclic stresses

C11: High contact stresses due to rolling & sliding action of mesh

C12: Vibratory dynamic load from bearing

C13: Uneven bearing load

C14: Reverse & repeated cyclic loading

Degree of unity in percentage (%) contribution for ith

failure cause

Scores for degree of severity (S) for ith

alternative and jth

criteria in

traditional FMECA

D: Degree of detectability

: Distance between positive ideal solutions in TOPSIS

: Distance between negative ideal solutions in TOPSIS

Deviation in preference value in PSI jth

criteria in PSI

: Downtime

EC: Economic cost

ES: Economic safety

: Entropy of jth

criteria in TOPSIS

xxii

: Hazard rate

M: Degree of maintainability

Maximum value of relative weight of Maintainability criticality index

among all alternatives

: Maintainability criticality index for COPRAS-G

: Maintainability criticality index for PSI

: Maintainability criticality index for TOPSIS

: Mean Downtime

: Mean time between failures

: Mean time between maintenance,

: Mean time to repair

: Frequency of failure

Score for normalized decision matrix – X in PSI

Overall preference value for jth

criteria in PSI

P: Probability of chances of failure

Weighted mean normalized sums of beneficial criteria

Preference variation value for jth

criteria in PSI

Weighted mean normalized sums of non-beneficial criteria

Score for normalized decision matrix – X in TOPSIS

S: Degree of severity

: Positive ideal solution in TOPSIS

Negative ideal solution in TOPSIS

SP: Spare parts

: Total time

: Uptime

Score for decision matrix – X in TOPSIS

Maximum score value of each alternative

Minimum score value of each alternative

; : Lower and upper score value of decision matrix – X in grey interval in

COPRAS-G

Lower and upper score value of normalized decision matrix – X1 in grey

interval in COPRAS-G

xxiii

: Lower and upper score value of weighted normalized decision matrix –

X2 in grey interval in COPRAS-G

: Weight of jth

criteria in TOPSIS

xxiv

List of Figures

Figure

No. Title

Page

No.

1.1 Bathtub curve (Ebling Charles, 2016) 03

1.2 Effect of planned maintenance on failure rate (Moubray

John, 1997) 04

1.3 Series network 05

1.4 Parallel network 05

1.5 Redundant reliability model 06

2.1 Rolling mill plant layout 30

2.2 Rolling mill components with their functional details 31

2.3 Actual image of rolling machine (Fifteen stands) 31

2.4 Rolling mill process flow (Properzi) 32

2.5 Hazard rate curve for rolling mill 44

2.6 Availability curve for rolling mill 45

2.7 Criticality curve for reliability parameters (rolling

machine components) 46

2.8 Criticality curve based on losses in production volume

and cost 46

3.1 Some photographs of the observations made at shop-

floor study 51

3.2 General flow process of traditional FMECA 58

4.1 Flow Diagram of MCDM based FMECA process 65

5.1 Pie chart presentation for % failure contributions of

components 90

5.2 RPN for each failure cause based on traditional FMECA 93

5.3 MCI for each failure cause based on MCDM based

FMECA 94

xxv

List of Tables

Table

No. Title

Page

No.

1.1 Comparisons of few researchers’ considerations with

presented study

21

2.1 Format for master list of machineries/component 36

2.2 Format for breakdown maintenance records 37

2.3 Format for preventive maintenance check list 37

2.4 Part-wise failure data of rolling machine (Duration: April

- 2013 to March - 2014)

39

2.5 Month-wise summary of failure data of rolling mill –

reliability modeling (Duration: April -2013 to March -

2014)

42

3.1 FMEA of derived vital parts 52

3.2 Scores for probability of occurrence (P) 55

3.3 Scores for degree of detectability (D) 55

3.4 Scores for degree of severity (S) 55

35 Scores for degree of maintainability (M) 57

3.6 Scores for spare parts (SP) 57

3.7 Scores for economic safety (ES) 57

3.8 Scores for economic cost (EC) 58

3.9 Decision matrix – X for traditional FMECA 60

3.10 Risk priority number (RPN) for traditional FMECA 60

3.11 RPN based FMEA with existing practices and proposed

improvements in maintenance plan

61

4.1 Decision matrix – X for TOPSIS 67

4.2 Normalization of decision matrix – X for TOPSIS 68

4.3 Distances between positive and negative ideal solution 70

4.4 Maintainability criticality index and criticality

rank for TOPSIS

71

xxvi

4.5 based FMEA with existing practices and

proposed improvements in maintenance plan

72

4.6 Decision matrix – X for COPRAS-G 75

4.7 Normalized decision matrix – for COPRAS-G 76

4.8 Weighted normalized decision matrix – for

COPRAS-G

77

4.9 Maintainability criticality index and (%)

contribution for COPRAS-G

79



80

4.11 Decision matrix – X for PSI 82

4.12 Normalized decision matrix – for PSI 84

4.13 Multiplication matrix of and 85

4.14 Maintainability criticality index and rank for PSI 85



87

5.1 Outcome of shop-floor data regarding performance of

rolling machine parts

92

5.2 Outcome (RPN) of traditional FMECA and suggestions 93

5.3 Outcome ( ) of MCDM FMECAs and suggestions 94

5.4 Maintenance optimization action plan 97

1

CHAPTER 1

Introduction and Literature Review

1.1 Broad Area of Research

The present research study is focusing on the area of reliability engineering, maintenance

management and failure pattern of the process industries influencing their performance.

Moreover, the research work also discusses the multi-criteria decision making methods to

enhance failure models. The area as reliability engineering deals with the methods of

reducing the failure frequencies by identifying and rectifying the causes of failures. The

accountability of maintenance management is to ensure efficient functioning of the

processing units or their systems and components with optimized capacity. The failure

pattern study helps to understand the failure causes, effects and their consequences. The

study is exploring the scope of failure mode and effect analysis and its modification as a

promising tool for targeting reliability problems. The area of multi-criteria decision

making encompasses the instinctive assessment of various criteria with interdependency.

Such tools are reasonably new in the failure analysis problems associated with processing

units to effectively make decisions. The presented study is smoothly integrating all such

areas in consideration of research objectives.

1.2 Overview and Significance of Study

The reliability subject is essential for maintenance personnel like; service managers and

plant technocrats alike, for keeping the components in a ready state. The reliability

engineering helps to identify the fault-diagnosis of component or system, compare


2

several consequences of failure in a view to plan and maintain them effectively. It is

imperative to mix regular and planned maintenance retaining the reliability and

availability of components. A proper reliability analysis will help the plant managers

enhancing system’s working and maintaining task as well. The conditions of operation

and repair strategies of the processing unit are crucial in maintaining the operating

systems with the highest uptime. This is achieved only through the performance

evaluation and analysis of critical components of the process industry. Moreover,

maintenance planning and optimization are complex and stochastic, which require multi-

criteria approaches for enhancing mean time between failure (MTBF), mean time to

repair (MTTR) as well as other reliability parameters. The issues or drawbacks of

maintaining non-performing things of the processing unit are covered under the proposed

study. The system performance can be measured in reliability and analyzed in real

working conditions to optimize maintenance activities of concern aluminium rolling mill.

1.3 Brief Overview about Reliability

1.3.1 Reliability Concept

Reliability is defined as the probability that the part, sub-system or unit will perform its

required job adequately for given period of time under specified design life

(Balagurusamy, 1984). This parameter is important as the repairs should always be

considered in a determination of reliability for the repairable system. Reliability is a time-

dependent property which can be predicted at any time during the process but can be

evaluated after an elapsed time. The unreliability is defined as the probability that the

part, sub-system or unit will not perform its desired function for given orders within

specified design life.

Important features of reliability are;

Reliability works on the concept of probability and its value ranging between 0

and 1.

Use of well-established and simple design leads high reliability of the

components.

Redundancy increases high risk of failure and leads to lower the reliability.


Poor shop-floor activities, complexities in system, poor maintenance practices,

lack of skill and human errors increases unreliability.

Assessing system reliability by finding faults at each subsystem and their

contribution in overall unreliability is essential for performance reliability

improvement.

Efficient reliability analysis should be supported with good mathematical

reliability models.

Reliability can be enhanced by using high-rated component with design

improvement, simplicity and better maintenance practices.

1.3.2 Significance of Mortality (Bath-tub) Curve

The mortality curve is representing the failure rate of the components in three different

zones. The components are experiencing single or any combinations of zone during a

lifetime as shown in Fig. 1.1. According to the shape of curve, it is also called the bath-

tub curve (Mishra and Pathak, 2012).

FIGURE 1.1 Bathtub curve (Ebling Charles, 2016)

In the first stage, the rate of failure suddenly decreased from large initial failure due to

machining or commissioning errors which is called burn-in zone. Once the machine

components pass this stage, it is entering into the long-life zone without hurdles.

However, the scheduled maintenance may send such components back into the burn-in

zone as shown in Fig. 1.2. For example; if similar bearings groups are imposed planned

maintenance, they may suffer burn-in period. The proper maintenance strategies on


4

showing signs of imminent failure are good to increase the design life of such

components.

The second stage of bath-tub curve is constant failure rate region and third and final stage

experienced the sudden rise of the failure rate due to parts or components wear-out. The

bath-tub curve is very useful in understanding failure-form of major industrial processes.

The simple as well as the complex system is following such failure pattern which helps to

understand failure pattern of the components or system to enhance reliability.

The planned maintenance increases the maintenance task due to an increased failure rate

as the parts or components are experiencing infant mortality after maintenance. If the

parts or components are allowed to run till failure by incorporating proper condition-

based techniques, the average life of the components are extended.

FIGURE 1.2 Effect of planned maintenance on failure rate (Moubray John, 1997)

1.3.3 System Reliability Models

The important system reliability models with their configurations are discussed as under

(Balagurusamy, 1984):

1.3.3.1 Series networks

In the series model, the components are in a series manner and the system is said to be

successful only when all its components are successful. Consider reliability model in

which n components with reliabilities , …., are connected in series. The

reliability of the complete system would be;


……. (1.1)

FIGURE 1.3 Series network

In this configuration, the failure of any one component affects the system in shut down.

An alternate exponential formula for calculating the reliability of a series network is

given as;

(1.2)

Where = the total parts working in series manner

1.3.3.2 Parallel networks

In this configuration, it is possible for the system to be partially operative even if some of

its components are in non-working condition. Considering the configuration of n

components having reliabilities , …., which are connected in parallel. The

reliability of the system can be derived from the following equation;

(1.3)

Where; =

FIGURE 1.4 Parallel networks

1.3.3.3 Redundant Reliability Model

If the system contains both series and parallel components to obtain higher reliability,

then the reliability of it is calculated similarly as discussed in parallel and series


6

arrangements. The analysis of a system is done by combining all components into

equivalents serially. Following configuration helps to analyze the system effectively.

FIGURE 1.5 Redundant reliability models

1.3.3.4 The x – out of – m structure

This type of configuration is another important practical system in which spare

components are paralleled to reduce downtime. The binomial distribution is good to

determine reliability of the systems with identical and statically independent components.

According to the binomial theorem, the probability of x out of m components is evaluated

with the help of following expression;

(1.4)

Where;

p = the probability of components in working condition

( ) (1.5)

So, system reliability is the addition of binomial probabilities as follows;

R = ∑ (1.6)

1.3.3.5 System with mixed mode failure

For the component having double failures, reliability is expressed as;

(1.7)

Where; and are failure probabilities due to short mode and open mode respectively.

1.4 Brief Overview about Maintenance

1.3.4 Reliability Analysis through Failure Distribution Function

As demonstrated by Balagurusamy (1984), it is necessary for reliability failure data

which are collected from either experiment or field observations to have their appropriate

distribution due to random in nature. Sometimes it is difficult to gather sufficient data for

accurate analysis. Understanding the pattern such data can be possible through some

failure distributions as listed below through which proper reliability analysis can be done:

Binomial

Exponential

Gamma-Poison

Normal and logarithmic-normal

Weibull type

Normal and exponential distributions are good, however, Weibull distribution is an

important distribution used in the analysis of reliability of the system. By properly

selecting the parameters the curve obtained can be represented with field observations.


Since last decade, the maintenance has changed than any other management philosophy

because of increasing diversification of engineering resources including automation.

These assets should be maintained throughout their lifespan. Moreover, more complex

designs need to explore the scope of optimal ideas of up keeping resources. Maintenance

process is reacted and influenced for creating awareness about equipment failure

concerning quality, safety, environment, and increasing pressure to achieve high-

performance reliability with little or no cost.

Maintenance is the process of optimizing the available resources such as; manpower,

materials, machinery, tools and testing equipment within the asset in a view to achieve

objectives and goals of an organization (Khanna, 2010). Pintelon and Gelders (1992)

defined the maintenance activity as the process to restore as well as to keep the

equipment in a designated working environment.


8

The goal of any organization is profit making by performing tasks as cost-effective as

possible. The primary objectives of maintenance management as discussed by Van Raijn

(1987) for industries are:

(i) to maximize the volume and capacity with consideration of reliability and

availability of the component,

(ii) to optimize safety and environmental ease

1.4.1 Maintenance Philosophy

As discussed by Khanna (2010), the maintenance philosophy is to optimize productivity

and plant availability with minimum consistent maintenance team and safety. The right

mixing of the following strategies is effective to obtain such philosophy.

(i) Breakdown Maintenance: In this process, a machine is allowed to run until it fails.

Several machines are maintained in this way without financial justification. It

leads loss of reliability due to the excessive delay in the production.

(ii) Scheduled Maintenance: It is the regular time-interval procedure so that sudden

failures can be prevented. The inspection, repair and overhaul, lubrication etc. are

covered under this strategy.

(iii)Preventive Maintenance: In this type of maintenance plan, each critical machine is

shut down after a specified period of time and replacement or repair of worn-out

parts is carried out based on the inspection. It works on the principle of prevention

is better than cure. The life of the components can be improved with additional

safety and low cost. However, it cannot guard against deterioration between

overhauls.

(iv) Total Productive Maintenance (TPM): It is fundamentally the management tool in

which all employees of the organization are involved for maintenance

improvement task. It works on a Japanese concept where; total quality

management (TQM) and just in time (JIT) concepts are applied over equipment

maintenance area.

(v) Predictive Maintenance: In predictive maintenance, the condition of machine

components are regularly measured and recorded. The fundamental difference

between preventive maintenance and predictive maintenance is that first type is

done immediately upon the progress of pre-determined period. Whereas in


predictive maintenance inspection or condition monitoring is done at pre-decided

regular intervals to decide maintenance. This type of maintenance strategy is used

to overcome the limitations of previously discussed maintenance types.

(vi) Reliability – Centered Maintenance (RCM): Moubrey (1997) discussed the role of

reliability modelling focusing the maintenance activities in a manner to retain

reliability of the components or system. The process of RCM employs FMEA to

understand the plan of maintenance of components effectively.

There are several maintenance strategies as discussed above which can be implemented

in any manufacturing or processing plant. The manufacturing or process equipment is

complex in nature and needs skilled personnel which lead increasing the maintenance

costs (Albert and Tsang, 1995). The decision-making approach helps to decide better

tools in various processing units including rolling mill. Fulop (2005) recommended the

usefulness of the decision-making process in identifying maintenance approaches which

satisfy the objectives of process industry in terms of profit. Some efforts were made to

bring such models to different aspects of maintenance management in the past. However,

it is a new and broad tool which exhibits the scope for modelling maintenance

management activities in many process industries.

In the large process industry, the aim of the production department is to manufacture a

planned output over definite time based on sales demand. The system or sub-system of it

may be in one of the following conditions;

In production process

No need for production

Under scheduled maintenance

Under corrective maintenance due to unexpected failure

Non-operational due to a shortage of maintenance resources

The maintenance optimization model should be presented by considering the parameters

affecting the availability and condition of the plant to gather. So, the function of the

maintenance team is to repair, replace, correct or modify the system or component of the

processing plant to make it operational with optimum performance reliability.

The important engineered maintenance strategies are discussed as under in a view to

improving component reliability. (Khanna, 2010).


10

(i) Planned-maintenance: In this strategy, the major maintenance task is split into the

minor scheduling-based task by integrating activities pre-planned.

(ii) Emergency Maintenance: It works on two approaches; in the first approach,

exigencies are handled with routine scheduling thereafter pending issues are covered

with contractual or ad-hoc workers or by providing overtime to regular workers. It is

proved strategy for processing units to allow about 15% of emergencies. The second

approach includes; estimation of an amount of emergency maintenance and

assignment of a work order to skilled and dedicated workers.

(iii)Reliability Improvement: The reliability improvement approach offers a sound

alternative for improving the maintenance task in which critical and major

equipment’s historical records are observed to monitor MTBF behaviour.

(iv) Equipment Management Program: It deals with generating TPM plan to almost all

parts of an instrument in the view to provide the satisfactory improvement in

maintenance performances.

(v) Cost Reduction: This approach works on the principle of reducing costs with

alternative materials, methods of service and overhaul, tooling and equipment,

scheduling and time standards.

(vi) Training and Employee Motivation: The art of optimum performance is always skill-

based which required an on-the-job-training (OJT) program to ensure that person

should have enough skills with high motivation.

1.4.2 Systematic step-by-step method for planning maintenance program

The systematic step-by-step process of the planning a maintenance program as suggested

by Khanna (2010) is as under:

(i) Determine the critical plant units by dividing the plant into units considering the

nature of the plant process. Then, failure analysis and estimation of the loss of the

production and availabilities should be done.

(ii) Classify the plant into constituent items. The classification is complex for the units

having high criticalities.

(iii) Decide the appropriate process to each item or component and assign them a rank

based on the cost and safety viewpoint.

(iv) Establish a plan for the identified critical components based on the type and

complexity of the plants.

1.5 Brief Overview about Failure Mode Effect and Criticality Analysis (FMECA)

(v) Establish the on-line and off-line maintenance schedule, corrective or preventive

maintenance, condition-based maintenance, shutdown etc. Unexpected failures

should be planned under spare management.

1.5 Brief Overview about Failure Mode Effect and Criticality Analysis

(FMECA)

1.5.1 Concept

The FMEA is the structured process of identifying, analyzing and documentation of

possible issues of component or system (Mcdermott et al., 2009). The failure

consequences on performance are reviewed and appropriate remedial measures are

imposed to eliminate or control them. It includes criticality analysis called; FMECA

which is an important reliability approach for avoiding costs experienced from product

failure.

FMECA is useful in following important activities;

Overall design process

Concept development phase to reduce the cost of design changes

Testing

Design modifications etc.

Other tools like; brainstorming; verification, fault tree analysis, data and record keeping,

material selection and procurement etc. are considered together with FMECA in order to

gain advantages.

FMECA is a most common tool for the planning the maintenance activities of the process

industries through reliability analysis among various tools. FMECA mainly consist of

two different approaches; (i) failure-mode and analysis of its effect and (ii) criticality

analysis. The FMEA is the technological tool for explaining, recognizing or removing

issues of the design process (Omdahl, 1988). It is one of the earliest preventive measures

which are reducing system or design failures (Kececioglu, 2002).

In FMECA, RPN is calculated with a multiplication of the scores of criteria like;

probability of chances of failure (P), detectability (D), and severity of the effect (S). It

helps to identify defects or failures happen in processing or element through structured


12

conceptualized discussion among experts (Vandenbrande, 1998). Dhillon (1985) and

O’Conner (2002) defined FMECA as a committed tool to evaluate the system reliability

with a review of possible failures modes with their consequences. It also provides the

organized way to determine criticalities considering risks as the final point in failure

study problem (Holmberg, 1991).

1.5.2 Types of FMEA/FMECA

There are three common types of FMEA as discussed below (Mcdermott et al., 2009) :

(i) System or concept FMEA: It is focusing system related deficiencies by

integrating the concept of system or sub-system in early stages. It provides

effective coordination between a system and surrounding considering single

failures.

(ii) Design FMEA: It is focusing component related design deficiencies by

involving adjacent components ensuring reliable performance for useful life.

This type of FMEA is normally implemented at a component, sub-system or

system level.

(iii) Process FMEA: It is focusing on process related deficiencies by incorporating

manufacturing and assembly operations, inward, storage and dispatch of

components or spares, conveying and maintenance processes. It ensures the

product should be manufactured to design specifications with minimum

downtime.

1.6 Brief Overview about Multi-criteria Decision-making (MCDM)

1.6.1 Overview and Importance of MCDM Approaches

Due to rapid technological and economic growth from the past few decades, the process

industries are facing complex problems in taking decisions. MCDM is very useful in

taking a decision where several alternatives are possible. It is very common practice to

select the most preferable alternatives through several mathematical models which are

developed recently. The MCDM tools are attracting academicians, practitioners etc. to

make convenient decision-making during conflicting situations of various criteria jointly.

International society on multiple-criteria decision-making defined the MCDM as the

1.6 Brief Overview about Multi-criteria Decision-making (MCDM)

process of consolidating disputed and numerous criteria together by the ways of

decisions.

The MCDM approaches started during early 1971. The growth of MCDM approaches is

due to pressing need of considering several criteria to the reliability problems of major

industrial processes including rolling process. It helps maintenance practitioners’ to

enhance the decision-making with recent advancements in optimization methods. Many

research articles and studies reflected the effect of MCDM paradigm on business,

engineering, and science where experts of various branches are assembled (Wiecek et al.,

2008).

Eshlaghy and Homayonfar (2011) considered three hundred eighty-six research papers

about the applications of MCDM approaches in various areas to know the facts about

state of art of work in MCDM domain. The results show the various categories as

follows: energy affairs twenty papers (about 5%), transportation and logistics matters

seventy eight papers (about 20%), strategic planning forty three papers (about 11%),

business and financial management fifty papers (about 13%), manufacturing and

assembly consideration thirty five papers (about 9%), environment affairs thirty four

papers (about 9%), water management twenty two papers (about 6%), agricultural and

forest area twelve papers (about 3%), project management thirty eight papers (about

10%), social welfare eleven papers (about 3%) and military domain eight papers (about

2%) and other general topics like; general science, sports etc. thirty five papers (about

9%).

There is very little application of MCDM approaches in processing plants to improve the

performance in different aspects including maintenance optimization. The major

industrial processing units are facing crucial problems of loss of reliability due to

components failures and breakdowns. All such plant’s performance depends on various

parameters or criteria. It seems very difficult to establish interdependence among all

criteria through traditional approaches. The requirement leads MCDM tool as a better

alternative for handling the problems associated with components failures and planning

effective maintenance plan among decision-makers.


14

1.6.2 Multi-criteria Decision-making Process

Janssen (2001) and Macharis et al. (2004) explained that MCDM is an explicit and

structured way to gain reasonable outcome by enhancing objectives satisfactorily. Roy

(1996) proposed a MCDM based solution in conflicting conditions as follows:

(i) Define the set of actions; say set S and the group of criteria; say group C

(ii) Decide best actions subset of in accordance to group C

(iii) Sorting of set S into subsets with their features

(iv) Assign rank to the activities of set S on basis of good to bad

(v) Discuss bad effects in a structured manner in a view to evaluate them effectively

The decision-making process flows step-by-step, however in case of addition of new

data; the process is repeated from any desired step. Yoe (2002) explained the MCDM

process as given below:

(i) Define the MCDM problem and its objectives.

(ii) List and explain alternatives to satisfy objectives.

(iii) Interpret criteria evaluators for the fulfillment of various options.

(iv) Collect data to determine the criteria.

(v) Build the decision-matrix for alternatives versus criteria.

(vi) Obtain unbiased weights for each criterion.

(vii) Assign ranking to alternatives and present the results to concern team.

(viii) Decide to recommendations of concern team group to obtain results.

1.6.3 Multi-criteria Approaches

Roy (1996) categorized the MCDM approaches into the following groups according to

their similarities, one multi-attribute concept; second out-ranking approaches; and third

interactive tools as under:

(i) Unique criterion method: It involves combining many perspectives into the

definite function to be optimized. For example; TOPSIS approach.

(ii) Out-ranking method: It involves the growth of an out-ranking relationship with

choice of decision-makers, to facilitate decision-maker solving problems. For

example; ELECTRE approach.

1.7 Historical Background of Reliability Analysis Issues

(iii) Interactive local judgment method: This method is based on trial-error relation

who provides alternative calculations with successive compromising solutions,

in a view to have additional knowledge to the choice of decision-taker (Vincke,

1992).

Some important MCDM approaches are listed as under:

AHP; Analytical hierarchy process

ELECTRE; called elimination and choice translating reality

GP or goal programming

COPRAS-G; Grey-complex proportional risk assessment

MACBATH means; measuring attractiveness by a categorical-based evaluation

technique

MAVT i.e. multi-attribute value theory

PROMETHEE or preference ranking organization method for enrichment

evaluation

PSI; Preference selection index approach

TOPSIS called; a technique for order preference by similarity to ideal solution

WSM i.e. weighted sum model

Many MCDM methods (e.g. TOPSIS, COPRAS-G) use the weight criteria in deciding

criticalities. These weights are keys to evaluate the overall preferences of alternatives.

MCDM methods utilize the different weights due to different groupings. It is very

essential for a decision-maker to correctly apply weights to alternatives for satisfactory

results.

1.7 Historical Background of Reliability Analysis Issues

From last five decades, reliability engineering is considered an important branch of

science. A study of reliability engineering helps in understanding the difficulties

associated with repairable systems. Development of reliability engineering is due to

quality control and its need. American Society for Mechanical Engineers (ASME) in

association with Bell laboratories was using statistical quality control during early the

1920s. Thereafter mathematical consideration of reliability was used in military

technology during the Second World War in 1939. Weibull studied fatigue life in


16

materials and proposed failure distribution function to fit data and observations especially

for reliability analysis.

The reliability analysis issues were discussed almost half-century back. These issues are

considered as useful in the field of reliability modelling, risk analysis and maintenance

planning. Barlow and Proschan (1965) proposed some mathematical practices in

maintenance activities. Jensen (1995), Dekker (1996) and Pham and Wang (1996)

discussed the maintenance models and their classification. Sikorska (2008) presented the

scope to improve the quality of failure histories stored in computerized maintenance

management systems (CMMS).

1.8 Comprehensive Literatures Review related to Maintenance Planning

Ashayeri et al. (1996) proposed mixed-integer linear programming based modification in

preventive maintenance schedule and production order of processing plant. Knapp (1998)

proposed a model to distribute maintenance personnel in proper quantity and quality at

proper workplace and time by precisely indicating the role of employees to improve

performance reliability in processing units. Santos et al. (1999) suggested the need of

efficient management system for the process industries which covers major factors to

reduce operating costs. They prepared the genetic algorithm for production sections of

kraft pulp and paper industry to introduce forced shutdowns. Parida et al. (2000)

explained the joint efforts of SAIL and Rourkela steel plant, and research and

development centre for iron and steel (RDCIS) to implement total productive

maintenance (TPM). They recommended vibration monitoring and analysis to do such

task effectively. Marseguerra et al. (2002) proposed a multi-objective genetic algorithm

(GA) to optimize profit and availability for determining the optimal degradation level.

Chen et al. (2004) discussed the optimization of fed-batch bio-reactor with recurrent

neural network (RNN) and genetic algorithm (GA) jointly to obtain optimal feed rate.

Sortrakul et al. (2005) suggested an approach for integrating production timings and

precautionary maintenance management through heuristic based GA tool and show its

effectiveness numerically through various problems. Abdullah et al. (2006) described the

genetic algorithms (GA) based solutions to the problems of multiple objectives with

diverse fitness functions. Eti et al. (2006) discussed the need of condition monitoring in

PM through FMECA, root-cause analysis, fault-tree analysis etc. Ghosh and Roy (2006)

http://www.tandfonline.com/action/doSearch?action=runSearch&type=advanced&searchType=journal&result=true&prevSearch=%2Bauthorsfield%3A(ASHAYERI%2C+J)

1.8 Comprehensive Literatures Review related to Maintenance Planning

discussed the benefit-to-cost ratio (BCR) based methodology to optimize the preventive

maintenance of processing units whose failure distribution function is either exponential

or Weibull. Nastac and Thatte (2006) described the potential fault locations based on

software tool and genetic algorithm and presented the results achieved from this tool.

Andrawus et al. (2007) highlighted two approaches; modelling system failure (MSF) and

delay time maintenance model (DTMM) to the windmill for maintenance planning. They

highlighted the relevance and applicability of these techniques for optimizing the

maintenance of wind turbines. Mohanta et al. (2007) used a genetic algorithm (GA) and

simulated annealing (SA) with stochastic reliability as an objective function to evaluate

the loss of load probability (LOLP) for maintenance of components of captive power

plant like; boilers, turbines and generators. Nguyen and Bagajewicz (2008) suggested

preventive maintenance optimization with the help of GA and Monte Carlo simulation

techniques to evaluate the cost and economic loss. Sikorska (2008) explored the scope of

quality improvement of historical records of failures, which are stored in computerized

maintenance management systems (CMMS) through cost evaluation algorithm.

Mahadevan et al. (2010) discussed the maintenance planning problems for a process

industry to decide substitution or repairing the subsystems with the use of a hybrid

genetic algorithm (HGA). Nguyen and Bagajewicz (2010) proposed the optimal

preventive maintenance to processing plants through the genetic algorithm to equipment

and labour allocation task. Sanjeev Kumar (2010) developed the optimized models in

terms of availability for units of fertilizer processing unit considering steady-state

behaviour. Verma et al. (2010) recommended three-pronged strategies to establish e-

maintenance for large engineering plants (LEPs) and derived the scope of their results as

inputs to upgrade the condition monitoring methods. Moghaddass et al. (2011) addressed

x - out-of - m configuration to dissimilar units for efficient assessment of the availability

of systems in order to get steady state solutions through a non-homogeneous process

called QBD i.e. quasi-birth death process. The correctness and efficiency of proposed

methods were demonstrated with the help of analogue Monte Carlo simulation in their

study. Mehdi et al. (2011) focused to get the optimal balance between costs and benefits

of maintenance with its execution time on the existing maintenance optimization models.

Braaskma et al. (2012) presented the application of FMEA for industrial practices and

summarized descriptions and assumptions of FMEA into six postulates which helps

maintenance practitioners to access their potentials for implementing FMEA. Godwin et

al. (2012) discussed the critical failures of various subsystems like; impact crusher, air


18

slide, conveyor and separator system, elevator and gear assembly etc. of the raw mill in a

view to developing the cost-effective maintenance plan through analytical hierarchy

process (AHP) and goal programming (GP). Zhixian and Guobin (2012) developed the

genetic algorithm based optimization model for aircraft maintenance to minimize the

cost. Lynch et al. (2013) described the effect of maintenance plan on the performance of

an industrial system by considering preventive maintenance optimization through a

genetic algorithm. Chen et al. (2013) modeled genetic algorithm based PM i.e. preventive

maintenance scheduling with the concept of life-age for RRE i.e. reusable rocket engine.

1.9 Literatures Review on Multi-criteria Decision-making (MCDM)

based FMECA

By reviewing and analyzing literature, it is found that various researchers enhanced

FMECA through MCDM approaches to overcome the drawbacks of many processing

plants. Hwang and Yoon (1981) proposed MCDM significance by considering disputed

and diverse criteria on educational or corporate affairs. Gilchrist (1993) suggested an

expected cost model with economic considerations through modifying FMECA;

Bevilacqua et al. (2000) incorporated operating state as new criteria for power processing

unit to enhance FMECA; Braglia (2000) integrated financial affairs in traditional

FMECA to analyze reliability and failure mode. Xu et al. (2002) evaluated an automobile

engine to alter FMEA through fuzzy logic. Braglia et al. (2003) suggested modifications

in conventional US MIL-STD-1629A approach using fuzzy TOPSIS to domestic

appliance producing industry of Italy. Sahoo et al. (2004) explained the FMECA is

essential for any maintenance plan and help to improve system reliability by reducing

overall maintenance cost. Sachdeva et al. (2009) approached paper processing unit with

MCDM based TOPSIS to re-order priority of causes of failure. Gargama and Chaturvedi

(2011) discussed risk factors in fuzzy linguistic variables to induce fuzzy rank priority

number. Maniya and Bhatt (2011) approached the problem of facility layout design

selection criteria through PSI based MCDM approach. Zammori and Gabbrielli (2011)

enhanced FMECA jointly with multi-criteria analytical network process (ANP) by

dividing fundamental criteria into sub-criteria to determine RPN. Liao et al. (2012)

illustrated preference orders of possible modes of failure through clouding FMECA for

1.9 Literatures Review on Multi-criteria Decision-making (MCDM) based FMECA

electrical transformers quantitatively. Feili et al. (2013) proposed FMEA approach to

comprehend the primary issues of units of geothermal power processing plant failures.

Adhikary and Bose (2014) discussed MCDM based COPRAS-G approach to thermal

power unit fueled through coal considering personnel’s features and working condition as

extra criteria. Fragassa et al. (2014) discussed FTA and RDA blended FMECA to

numerous diesel-suction systems with consideration of optimizing manufacturing

processing elements covering metal wire processing. Liu et al. (2015) highlighted the

joint use of fuzzy AHP and entropy method in fuzzy VIKOR for assigning weights to risk

factor for avoiding uncertainty and vagueness due to individual perception and

experience. Mobin et al. (2015) integrated the fuzzy AHP and COPRAS-G to assign

prime concern to suppliers of the household device producing company of Iran. Zhang

(2015) proposed fuzzy TOPSIS with unbiased weights for determining closeness-

coefficient for various causes of failure so that score calculation errors can be controlled.

Chanamool and Naenna (2016) incorporated fuzzy FMEA to the trauma center of a

hospital for planning and assessment of working failures of it. Fragassa and Ippoliti

(2016) discussed the use of FMECA in the complex situation of continuous production

and end with a list of modifications to be used for improving the plant design in the way

to reduce the maintenance task. Mittal et al. (2016) assigned ranks to severe issues of

plywood processing plant with the help of MCDM based fuzzy TOPSIS in order to refine

them effectively. Rathi et al. (2016) described MCDM based fuzzy VIKOR to look after

problems and their scenario of Indian automobile domain and strengthen the performance

of six-sigma. Rastegari et al. (2017) proposed monitoring the condition of vibrations for

spindles of gear-box producing enterprise situated in Sweden to establish a proper

maintenance plan. Farley and Miller of Innoval technology Ltd. presented three parts on

maintaining rolling mill performance. In the first part, they identified some factors

responsible for unhealthy rolling mill performance over time. In the second part, they

discussed the overall equipment efficiency (OEE) based approaches to improve rolling

mill performance and in third part they explained guidelines to avoid an initial decrease

in performance of new mill through good design, training and technical support.


20

1.10 Motivation of Research

The study is proposed in the field of reliability-centered maintenance based issues in

processing plant as its research interests are an excellent match for my academic

background. Competitive global market demands high reliability with optimized

maintenance. The motivational factors like; research gap and expected business growth of

an aluminium wire are discussed in following sections and then problem is defined.

1.10.1 Outcome of Literature Review and Research Gap

Literature reviews state that failure mode effect and criticality analysis (FMECA) is an

accepted tool for enhancing maintenance practices in the processing unit. Moreover, it

helps to identify defects or failures happen in processing or element through structured

conceptualized discussion among experts (Vandenbrande, 1998). Dhillon (1985) and

O’Conner (2002) defined FMECA as a committed tool to evaluate the system reliability

with a review of possible failures modes with their consequences.

Literature review seems certain possibilities of the performance reliability improvement

through maintenance optimization of systems or components of major process industries

that run on continuous basis. The challenges of keeping the system in ready-state

necessitate a definite maintenance plan to be modeled based on a live failure analysis to

be executed during shutdown or scheduled period. Table 1.1 shows comparisons of some

researchers’ contributions with presented work to exhibit motivation of research study.

Moreover, the exploration about literature shows that former researchers did not reflect a

case for three MCDM approaches simultaneously applied to any process industry. There

is a huge scope of improvement in reliability by optimizing maintenance practices

through MCDM based failure analysis models in processing plants and advocate a need

to address such interesting issues in form of the research study.

1.10.2 Definition of the Problem

Many researchers have presented modified FMECA approaches to various industries, but

quality research is lacking in aluminium wire rolling mill which forms a noticeable sector

of the process industry. Looking to the past business volume of aluminium wire about

6,50,000 metric tons during 2015 in India, an aluminium wire processing mill is consider

important for the study. (The sixty eighth report of IEEMA, 2014-15).

1.10 Motivation of Research

TABLE 1.1 Comparisons of few researchers’ considerations with presented study

Braglia et

al. (2003)

Sachdeva

et al. (2009)

Feili et.al

(2013)

Adhikari

and Bose

(2014)

Mittal et al.

(2016)

Presented study

Focused

Processing

Plant

Italian

domestic-

appliance

manufacturer

Paper

processing

unit

Geothermal

Plant

Coal-fired

steam power

plant

Plywood

processing

unit

Aluminium wire processing unit

Approaches Fuzzy

TOPSIS

TOPSIS Standard

FMECA

COPRAS-G Fussy

TOPSIS

(i) Traditional FMECA with basic criteria

(ii) TOPSIS with an assignment of scores in

crisp value with weighted attributes

(iii) COPRAS-G to express the criteria

values in intervals to avoid practical

difficulties and variations of maintenance

personnel

(iv) PSI by statistics instead of weight

assignments.

Consideration

of Criteria

Only three

basic criteria

Basic criteria

with

maintainability,

economic

safety and cost

consideration

Only three

basic criteria

Some

process

criteria with

basic criteria

Criteria like;

cost, safety,

maintenance,

environment

and

automation

Basic criteria with maintainability, economic

safety and cost consideration

Significance Limited to

three basic

criteria with

TOPSIS

Limited to

TOPSIS

Limited to

traditional

FMCEA

with basic

criteria

Limited to

COPRAS-G

Limited to

TOPSIS

RPN based criticalities with weighed

attributes; in exact value (TOPSIS) as well as

in upper and lower limits (COPRAS-G) and

PSI without calculating any weights


22

Moreover, due to a proposal of expanding electrical wire network, industry requirements,

infrastructure planning and Indian Government’s project about electrical power to every

corner of the country, it is expecting the growth of aluminium wire market about 13.5%

CAGR i.e. compound annual growth rate during 2014-2019. In Gujarat, a network of

transmission lines is expected to rise at 7.8% CAGR during the financial year 2014-

2018 and Government of Gujarat will be investing around US$ 4500 million in

transmission and distribution till year 2020 (BIG 2020 report, volume 1-2009).

During the detailed study of the performance of the identified processing plant, it is

observed that maintenance time is about 20 to 25 % of the total time which leads to

reliability losses. The yearly loss of production and profit is about 3000 tons and INR 45

crores respectively. The average usage of aluminium wire rolling mill in India is

approximately 75 - 80 %. After studying the facts related to this plant, it is found that

poor maintenance is the prime reason of low productivity and profit. Hence, it is required

to upgrade existing maintenance so that the highest utilization of resources is obtained by

incorporating small cost or without any additional cost.

In the present research study, it is investigated the scope of optimizing the maintenance

practices through actual failure analysis by applying traditional and three distinct MCDM

based FMECA approaches as under:

(i) TOPSIS i.e. technique for order preference by similarity to ideal solution where;

weighted scores are considered in the crisp value

(ii) COPRAS-G i.e. grey-complex proportional assessment where; weighted scores

are in grey range rather than in crisp value and;

(iii) PSI i.e. preference selection index where; subjective weight consideration not

required

1.11 Objective and the Scope of Work

The prime objective of the study is to investigate the scope of reliability improvement

and discuss non-identical failure analysis models for evaluating criticalities of various

failure causes of critical components of an aluminium wire rolling mill. It advances

refinement of planning the maintenance by modifying FMECA through MCDM

approaches of an identified processing unit. The results are helpful in prioritizing the

1.11 Objective and the Scope of Work

maintenance actions to the process industry of same or of different kinds in accordance

with failure analysis.

The objectives of the research explored are listed under:

(i) To study reliability and maintenance issues faced by the aluminium wire rolling

mill and deriving the scope of optimizing maintenance activities.

(ii) To collect and analyze the historical data associated with failures including

calculation of major reliability parameters for components of identified

aluminium wire rolling mill.

(iii) To identify the vital or critical parts of concern rolling mill with the help of

reliability parameters like; failure frequencies, downtime and loss of production

on volume and cost consideration.

(iv) To study failure pattern of the critical components, select various criteria or

attributes and assigning the scores to each failure cause for every criteria based

on real shop-floor condition in order to evaluate criticality level of these failure

causes.

(v) To optimize maintenance activities of the critical components through traditional

as well as MCDM based failure analysis models based on comparison of

criticalities achieved from them. Furthermore, suggestions regarding the remedial

measures for optimal performance of rolling mill based on findings.

The scope of the proposed research work is summarized as below:

(i) Understanding the current scenario about the working of an aluminium wire

rolling mill plant and to acquire relevant information about reliability and

maintenance issues. Study existing maintenance practices and its limitations,

deriving scope of improvements.

(ii) Collection of historical failure data of major components of concern rolling mill.

(The data may be used for comparison and investigation of future failures with

the highest probability of occurrence). Failure data analysis as well as reliability

parameter calculation to generate necessary inputs.

(iii) Identification of critical components of the rolling mill on reliability parameters

like; downtime, failure frequencies, production loss on volume and cost basic.


24

(iv) Interpretation of the failure modes, causes, effects and consequences of failure

pattern and problems faced in present maintenance practices.

(v) Traditional and MCDM based different failure models for prioritizing

maintenance activities.

(vi) Comparison of results of different failure models. Suggested improvement and

recommendations for a future scope.

1.12 Research Approaches

In the present research study, following approaches are applied to fulfil the objectives:

Objective (i): To study reliability and maintenance issues faced by the aluminium wire

rolling mill and deriving the scope of optimizing maintenance activities.

This objective is achieved by obtaining permission from Sampat aluminium private

limited (Deora group) situated at Rakanpur GIDC near Ahmedabad to do the research

work. The on-site study about basic components, product details, plant layout, reliability

and maintenance issues etc. of concern rolling mill is made. During preliminary study, it

is found that average maintenance cost or loss of reliability is about 20 to 25 % of total

production time with existing maintenance practices are either breakdown or planned

shutdown. The study seems fair verdict on a need of the scope of improvement in the

existing maintenance practices.

Objective (ii): To collect and analyze the historical data associated with failures

including calculation of major reliability parameters for components of identified

aluminium wire rolling mill.

This objective is achieved with suggesting some formats to gather the maintenance data.

Substantial failure data like; downtime and frequency of failures are collected for thirty-

one components of aluminium rolling machine for a period of one year (April, 2013 to

March, 2014) for further failure analysis.

Objective (iii): To identify the vital or critical parts of concern rolling mill with the help

of reliability parameters like; failure frequencies, downtime and loss of production on

volume and cost consideration.

1.12 Research Approaches

For fulfilling this objective, the comprehensive reliability failure data are scrutinized for

identifying the major vital components with the help of reliability parameters like; failure

frequencies, downtime, loss of production on volume and cost basis etc. for all components

of each stand of a rolling machine. The critical components identified for extended

failure analysis are; ball and tapered roller bearings of designation 6213, 32308, 30310,

32222, power transmission gears of bevel type with spigot and tapered shape and

primary and secondary machining shafts.

Objective (iv): To study failure pattern of critical components, select various criteria or

attributes and assigning the scores to each failure cause for every criteria based on real

shop-floor condition in order to evaluate criticality level of these failure causes.

This objective is achieved by studying failure modes, causes, effects, and consequences

of failure pattern with the present maintenance practices of identified critical components

of the aluminium wire processing mill such as; bearings, gears, shafts. The potential

FMEA development, criteria selection and score assignment is done by methods of a

question-based survey with maintenance team including managers, engineers, shop-floor

technicians and machine operators etc. The scores are assigned to various criteria on 1 to

10 point scales from minimal to greatest influence for each failure cause.

Objective (v): To optimize maintenance activities of the critical components through

traditional as well as MCDM based failure analysis models based on comparison of

criticalities achieved from them. Furthermore, suggestions regarding the remedial

measures for optimal performance of rolling mill based on findings.

This objective is achieved by incorporating traditional FMECA to calculate risk priority

number (RPN). Then, maintainability criticality indices are calculated by three

non-identical MCDM approaches; TOPSIS in crisp value, COPRAS-G in grey number

range and PSI without subjective weight consideration for each failure cause in a view to

sequence maintenance activities. The remedial measures are suggested for smooth or

optimal functioning of concern rolling mill.


26

1.13 Original contribution by the thesis

In the present research study, three distinct MCDM based failure analysis models are

deployed as a modified FMECA for providing the satisfactory alternatives to

maintenance practitioner for better maintenance strategies. The research will make the

following original contributions:

(i) The actual historical failure data like; runtime, uptime, frequency of failures,

average repair or replace time are collected for the duration of a year (April’ 2013

to March’ 2014) for indicated aluminium wire rolling mill.

(ii) Reliability terms such as; MTTR, MTBF, MDT, hazard rate, availability are

determined based on realistic historical failure data to generate necessary inputs

for further failure analysis.

(iii)The failure modes with their causes and effects, failure patterns with the

problems faced in present maintenance practices are studied. The potential

FMEA for major critical elements is deduced. Moreover, the assignment of

scores to each failure cause for every diversified criterion is done based on real

shop-floor condition in order to evaluate criticality ranks through different failure

analysis approaches.

(iv) The results of criticality are evaluated and compared with discrete failure analysis

approaches.

(v) A contribution is proposed through the failure analysis approaches in showing

case-study for the maintenance plan preparation to rolling mill processing plant

entirely in a research study.

(vi) Originality mainly consists of the contemporary application of three non-identical

MCDM based methods (TOPSIS, COPRAS-G and PSI). The results are helpful

in explicating the pitfalls of maintenance for foremost processing plants and

prescribed yield outputs.

1.14 Organization of Thesis

The thesis is split into six units in order to organize the research work effectively.

Chapter 1 discusses the broad area of research, an overview of the study and its

significance. It reviews the literature related to reliability, maintenance, FMECA and

1.14 Organization of Thesis

MCDM with historical background. It shows the comprehensive literature reviews about

maintenance optimization through different tools. Moreover, it discusses the reviews of

past research study on improvements or modifications of FMECA through MCDM

approaches. General aspects of reliability and maintenance are also addressed in this

chapter. It defines the research gap and problem statement about an investigation of the

scope of reliability improvement by maintenance optimization through MCDM based

FMECA approaches to aluminium wire rolling mill. It states the objectives and scope of

present research work and original contributions.

Chapter 2 presents an introduction and overview of the identified process industry with

plant layout, process flow, major components etc. It also shows the substantial historical

failure data for the duration of April 2013 to March 2014 at Sampat Heavy Engineering

Ltd., Ahmedabad, India. The reliability modelling and process of discrimination of the

critical parts are presented based on the failure data and shop-floor condition in this

chapter.

Chapter 3 discusses the FMEA derived through failure modes, causes, effects and the

failure pattern consequence and problems faced in the present maintenance practices of

indicated critical components. It also discusses the procedure of selection of criteria and

score assignment methodology to every failure cause traditional as well as MCDM based

failure analysis models. Moreover, it discusses the traditional failure models and

evaluation of RPN through it with maintenance planning.

Chapter 4 presents MCDM based failure analysis models with an addition of some more

advanced criteria. It also describes methods to evaluate MCI for each failure mode of

targeted critical components through three different MCDM failure analysis models

called; TOPSIS, COPRAS-G, and PSI for optimizing current maintenance strategies of

the critical components of the targeted unit.

Chapter 5 discusses the out-trend of literature, the results of discrimination process

through shop-floor data for critical components and criticality indices obtained through

traditional and various MCDM based FMECA approaches as discussed in chapter 4. The

comparison of results is displayed in form of tables, figures etc. for effective

understanding. Based on achieved RPN and MCI, remedial measures are suggested and

priority plan of existing maintenance activities is discussed in the chapter.


28

Chapter 6 concludes the present research study and the scope of future work

recommendations.

1.15 Summary

This chapter discusses the broad area of research, an overview of the study and its

significance. It reviews the literature related to reliability, maintenance, FMECA and

MCDM with historical background. It shows the comprehensive literature reviews about

maintenance optimization through different tools. Moreover, it discusses the reviews of

past research study on improvements or modifications of FMECA through MCDM

approaches. General aspects of reliability and maintenance are also addressed in this

chapter. It defines the research gap and problem statement about an investigation of the

scope of reliability improvement by maintenance optimization through MCDM based

FMECA approaches to aluminium wire rolling mill. It states the objectives and scope of

present research work and original contributions.

The next chapter presents an introduction and overview of the identified process industry

with plant layout, process flow, major components etc. It also shows the substantial

historical failure data for the duration of April 2013 to March 2014 at Sampat Heavy

Engineering Ltd., Ahmedabad, India. The reliability modelling and process of

discrimination of the critical parts are presented based on the failure data and shop-floor

condition in this chapter.

29

CHAPTER 2

Data Collection, Reliability Modeling and

Identification of Critical Components

2.1 Overview of Identified Process Industry (Rolling Mill)

2.1.1 Introduction and Background

Deora Group's Sampat aluminium rolling mill is well known for electrical conductor

(EC) grade aluminium transmission wire products of International designation 1350 at

national and international level. It deals the non-ferrous market with cost-effective and

product integrity features in products like; shots, ingots, notch bars, flip-coiled and EC

grade aluminium wires. The constant casting and hot rolling based Properzi process are

used to manufacture aluminium wires in various diameters. Such products have multiple

usages in the area of electrical as well as mechanical engineering including electricity

supply, transformers production etc. The targeted rolling mill is primarily supplying

good conductivity non-ferrous overhead cores to Australia so that requirements of steel

cores are minimized to a major extent.

The detailed layout of the aluminium wire rolling mill plant is given in Fig. 2.1. It mainly

consists of the furnace, caster wheel and rolling machine. The functional details of these

components are discussed in Section 2.1.3. The rolled wire is wound around the spool to

the desired size.

Data Collection, Reliability Modeling and Identification of Critical Components

30

FIGURE 2.1 Rolling mill plant layout

2.1.2 Rolling Process

The rolling is a manufacturing process in which untreated metal block is feed between

couples of rollers to form desired sizes. The process is termed as hot or cold rolling

based on the temperature of raw aluminium. The hot rolling and cold rolling are the

forming methods in which processing temperature is more and less than recrystallization

temperature respectively. The hot and cold rolling produces more output than any other

hot or cold working processes respectively.

2.1.3 Rolling Mill Components

The research study is concentrating the problems of loss of reliability as well as improper

maintenance of a specific aluminium wire rolling mill plant situated in Gujarat, India.

Fig. 2.2 demonstrates the actual view of the rolling mill components and there role to

convert raw aluminium into aluminium wire through molten state. The functional details

of these parts are discussed as follows:

(i) Furnaces: It is used to convert aluminium raw into melting state of

temperature about 800 0C. The identified mill is having oil-fired furnaces of

twelve ton and fifteen-ton capacity to meet demands.

(ii) Caster Wheel: The caster wheel of diameter 1.4 m helps to pass melted

aluminium from furnace to rolling machine for further forming operations.

This component works on the principle of constant casting with the use of

water as quenching media. The unprocessed molten metal is coming from a

furnace and transferred to the rolling machine through caster wheel by

converting molten state into a solid form at recrystallize temperature through

water quenching.

(iii) Rolling Machine: It is the primary functional part of processing mill where

reliability and maintenance issues are on high-priority. The research is


31

targeting the rolling machine for further failure study. The diameter of

unprocessed metal coming from caster wheel is decreased from 40 mm to 6

mm through 15 successive stands in a series network. Fig. 2.3 illustrates the

actual image of a rolling machine (fifteen stands) for conceptualize the real

view of a plant.

FIGURE 2.2 Rolling mill components with their functional details

FIGURE 2.3 Actual image of rolling machine (Fifteen stands)


32

2.1.4 Rolling Mill Properzi Process

The production of aluminium metal rod is possible through the properzi process as

shown in block diagram Fig 2.4. The properzi process is defined as the continuous

casting based rolling process of producing aluminium wire in long length directly

through the molten state. The step by step process of producing aluminium through the

Properzi Process is as under.

(i) Aluminium Ingot (billet) is fed into the furnace and melted at about 750-800 C.

(ii) The liquid aluminium is then fed to the aluminium caster i.e. casting wheel. Water is

used for cooling the hot aluminium to convert it into the soft solid bar of 40 mm

diameter.

(iii) The soft solid bar is then converted into 6 mm diameter rod through the series

rolling process by fifteen stands with a decrease of the diameter of wire by about 15-

20 % at each stand.

(iv) The coils of about 2-2.5 tons are wound.

FIGURE 2.4 Rolling mill process flow (Properzi)

2.1.5 Rolling Machine Sub-Components

The research is targeting the rolling machine for failure study where; successive series of

15 stands suffer pivotal problems of maintenance. Each stand consisting of thirty-one

sub-components as listed below:

1. Primary shaft

2. Secondary shaft

3. Bearing secondary housing

4. Primary bevel gear spigot end

5. Primary bevel gear taper end

6. Secondary bevel gear ring

7. Pin for entry guide roller

Melting of Aluminium ingots in furnace

Input

Semi solid cast bar through water

sprinking

Caster

Diameter reduction

through 15 stands in

series

Rolling Machine

Coiling of rod Output Dispatch


33

8. Main chuck nut for primary shaft

9. Spline side chuck nut

10. Chuck nut for bearing

11. Bottom nut for secondary shaft

12. Shear pin for drive assembly

13. Top nut for secondary shaft

14. Lock nut bearing side

15. Cylinder pin for primary assembly

16. Special bolt for secondary

17. Spacer for outer

18. Spacer for inner

19. Secondary block housing

20. Bearing housing 110ɸ for primary

21. Bearing housing 120ɸ for primary

22. Bearing tapered roller with designation 32308


24. Bearing radial ball bearing with designation 6213


26. Oil ring with designation 701010 changed with ball bearings

27. Oil ring with designation 608010 changed with roller bearings

28. Oil ring with designation 629010 changed with roller bearings

29. Coiler bolt

30. Casting bolt

31. Coupler bolt

The primary and secondary bevel gearing arrangement provides the proper feed with

forward motion and punch pressing force in a view to reduce diameter to desired level

for the processing aluminium bar. These gears are mounted in association with primary

and secondary shafts respectively. The tapered ball and roller type bearings facilitate the

rotary motion and support the dynamic loading. The other sub-components are integral

and essential parts of these three components in order to maintain smooth functioning of

them.


34

2.2 Major Reliability and Maintenance Issues during Preliminary

Studies and Learning from them

The following issues were observed during the preliminary study which affects reliability

and needs attention to enhance maintenance plan. The other specific observations are

made during detailed shop-floor study and failure analysis models have been discussed.

(i) Excessive vibration: Rolling mill is experiencing excessive machine vibration from a

super calendar stack specifically at higher speeds. The excessive vibration leads to

create severe damage to the rolls and keep the mill away from running to its

designated speed till replacement. The mill is experiencing such problems from

several years and the root cause is unknown to date. Following general causes are

observed to such problems:

High vibration at a lower part of the stack

Jammed/barred/broken rolls (Age-related issue)

Flat spots on the rolls

Failed bearings due to thrust load

(ii) Critical equipment failures: Rolling mill is experiencing frequent, unexpected and

sudden failures of critical equipment, which leads to loss of reliability. This

phenomenon may occur due to;

Lack of proper reactive maintenance.

Production below capacity.

Higher waste rate.

(iii) Poor operations and its management: The challenging issues addressing this

problem are;

Lack of an effective reliability program.

High downtime due to preventable unplanned breakdown.

High replacement rate of spare parts.

2.3 Failure Data Collection and Analysis

35


In reliability problem, the information or data on which a conclusion is derived must be

the result of some observation. During the failure study of rolling machine, the data

collection is considered the random experiment, which is the result of a repeatedly

performed experiment or observation under similar conditions with the different

performance at each level (Balagurusamy, 1984).

The random experiment has the following characteristics:

- It has a well-defined set of rules.

- It is repetitive in nature.

- The result of each performance cannot be uniquely predicted.

Examples of the random experiment are;

(i) Computing the number of defective parts in a batch through random

selection.

(ii) Measuring the lifespan of definite bulbs through random selection from the

lot.

(iii) Observing the diameter of a circular bar produced by turning process.

(iv) Recording the run time of car before the failure of the brakes.

(v) Deciding the age of individual persons in a community etc.

The above points are contemplated during failure data collection process. The main

objectives of the data collection are to monitor the performance of process plant and to

identify the actual condition of it which helps in deciding an appropriate course of action

to improve the reliability and maintenance issues.

The reliability engineering is associated with the problems of evaluating risks and

consequences. Reliability theory basically depends on probability theory for its

application. Balagurusamy (1984) suggested some points for addressing any reliability

problem as under:

(i) Understanding the physical problems associated with the real situations under

consideration.


36

(ii) Presenting the physical problems into mathematical form and application of

suitable methods to solve them.

(iii) Converting the mathematical outcome into a statement of real situations and its

implementation.

Looking to above points addressed for any reliability problems, it is necessary that the

system performance should be quantitative or in measurable terms of statistics. Hence, it

is necessary to collect certain data like; non-operational or downtime, frequency of

failures etc. associated directly with the system failures. The evaluation of reliability

parameters as discussed in this Section is helpful to interpreting the failure pattern

behaviour of the components of specific rolling mill. The recording of data and

evaluation process of concern parameters; i.e. reliability modeling is discussed in Section

2.4.

Based on observations during a preliminary study of plant condition and existing

maintenance system, the maintenance department of the plant is recommended and

consulted to record failure data in specific required format as displayed in Table 2.1 to

2.3 for effective data recording and management.

Table 2.1 is helpful to list the machineries with their technical specification and history

cards. Table 2.2 is facilitating the records of breakdown faults of any components/sub-

components with their timing of occurrence, existing control and maintenance practices

followed and downtime. Table 2.3 is providing preventive maintenance schedule

checklist on daily basis for each month to be followed. These templates are

recommended as a part of precise and organized data collection process to

comprehended neat failure pattern study.

TABLE 2.1 Format for master list of machineries/component

Sr.

No.

Name of

Machinery

/Component

I.D.

No Model Type

Manufacturer

Name

Year /

Month of

Installation

Capacity Other

Detail


37

TABLE 2.2 Format for breakdown maintenance records

Month :

_________________

S

No

.

Date

Machine

/Component /

Part Details

Fault/

Breakdown

Detail

Reported

time

Root

cause

analysis

Disposal/

Immediate

action taken

corrective

action

Break down

close date &

time

breakdown

hours Remarks

TABLE 2.3 Format for preventive maintenance check list

MONTH Machine Name Frequency

Parameter /

Check Point

/ To do

Point

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31


38

The data collection process involves the prior discernment of the parameters like;

runtime, downtime, uptime and frequency of failures. These primary parameters are

directly associated with shop-floor conditions. The measure of these parameters

quantifies the performance of the plant. Hence, it is required to record and determine

these parameters. The concept of these reliability parameters and their sample

calculation is illustrated as follows. Moreover, the obtained values are recorded in Table

2.4.

Runtime ( ):

The runtime is the total assumed running time of the plant for a specific period of time.

Here, in this study, the duration of a year of 365 days (April 2013 to March 2014) is

selected for data collection with considering one day as shutdown per month.

So, total working days are; ( )

Downtime ( ):

The downtime is the total non-working time of the plant due to repair or replaces during

failures. In this study, it is directly recorded and listed in Table 2.4.

Uptime ( ):

The uptime is the difference between running time and downtime. It is calculated from

downtime as shown.

Frequency of Failure ( ):

It is a number of breakdowns or failures noted during the specified service period of the

plant. It is recorded part-wise for the entire study period.

Production loss in terms of volume and cost:

The loss of production is the actual loss suffered due to failures under downtime. The

sample calculation for loss of production in terms of volume is derived by considering

the plant capacity as 1.5 ton per hour.

For example; the production loss in volume (ton) for primary shaft is;

And the production loss in terms of cost is derived by considering the cost of finished

aluminium wire as Rs. 140 per kilogram (Kg)

So, the production loss in cost (Rs.) for the primary shaft is;

( )


39

TABLE 2.4 Part-wise failure data for aluminium rolling machine (April - 2013 to March - 2014)

Part

No.

Part

Name

Run

time

(Hrs.)

Up

time

(Hrs.)

TOTAL

Down

time

(hrs.)

Failure

frequencies

(n)

Production

Loss

(Tons)

Profit Loss

(Rs. in

Crores)

1 Primary Shaft 8472 8388 84 21 126 1.76

2 Secondary Shaft 8472 8432 40 20 60 0.84

3 Bearing Secondary Housing 8472 8448 24 12 36 0.50

4 Primary Bevel gear Spigot end 8472 8394 78 23 117 1.64

5 Primary Bevel gear Taper end 8472 8400 72 18 108 1.51

6 Secondary Bevel gear Ring 8472 8418 54 27 81 1.13

7 Pin for Entry Guide Roller 8472 8447 25 154 38 0.53

8 Main Chuck Nut for primary

shaft 8472 8446 26 13 39 0.55

9 Spline Side Chuck Nut 8472 8452 20 10 30 0.42

10 Chuck Nut for BRG 8472 8448 24 12 36 0.50

11 Bottom Nut for secondary

shaft 8472 8444 28 14 42 0.59

12 Shear Pin for Drive Assembly 8472 8472 0 0 0 0.00

13 Top Nut for secondary shaft 8472 8469 3 36 4 0.06

14 Lock Nut Bearing side 8472 8461 11 11 17 0.23

15 Cylinder Pin for Primary

Assembly 8472 8412 60 15 90 1.26

16 Special Bolt for Secondary 8472 8472 0 0 0 0.00

17 Spacer for Outer 8472 8472 0 0 0 0.00

18 Spacer for Inner 8472 8472 0 0 0 0.00

19 Secondary Block Housing 8472 8472 0 0 0 0.00

20 Bearing Housing 110ɸ for

Primary 8472 8456 16 8 24 0.34

21 Bearing Housing 120ɸ for

Primary 8472 8448 24 12 36 0.50

22 Bearing No. 32308 8472 8136 336 168 504 7.06

23 Bearing No. 30310 8472 8140 332 166 498 6.97

24 Bearing No. 6213 8472 8136 336 168 504 7.06

25 Bearing No. 32222 8472 8000 472 118 708 9.91

26 Oil ring with designation

701010 8472 8472 0 0 0 0.00


608010 8472 8472 0 0 0 0.00


629010 8472 8472 0 0 0 0.00

29 Coiler Bolt 8472 8472 0 18 0 0.00

30 Casting Bolt 8472 8436 36 12 54 0.76

31 Coupler Bolt 8472 8456 16 8 24 0.34


40

Table 2.4 shows the thirty-one sub-components of a rolling machine with their historical

failure data. The plant has been monitored for a period from April 2013 to March 2014 to

record these data which are used to identify the critical components in the next section.

2.4 Reliability Modelling

Reliability modelling helps to understand and quantify the plant performance in terms of

design life. The reliability modelling is performed on substantial shop-floor failure data.

The brief overview about reliability parameters and their significance in this study is

presented as below:

The MTBF (Mean Time between Failures) is the expected time between failures of a

repairable system. The MTTF (Mean Time to Failure) is the expected time to failure of

the non-repairable system. The lower and upper confidence levels for these parameters

can be calculated by dividing total operational time with frequency of failures. The

fundamental difference between these parameters is that MTBF and MTTR is

specifically important to observe the performance of repairable and non-repairable

system or component respectively. Mathematically, it is expressed as;

(2.1)

For MTTR n is assumed as unity considering replacement after every instant of failures.

The failure or hazard rate is the probability of system failure within time t and t + 1 unit,

given that the system is continuously operational until time t. This system parameter can

be calculated for specific points in time. The components having constant hazard rate are

not replaced at specific failure instant, then total failure time is included in operational

time to evaluate MTTR. Mathematically, it is expressed as;

=

(2.2)

The MDT (Mean Downtime) is the average non-working time of the components of

plant due to repair or replacement. Mathematically, it is expressed as;

(2.3)


41

The MTTR (Mean Time to Repair) is average repair time of the components of the plant.

It is generally considered as 30 % of mean downtime. Mathematically, it is expressed as;

(2.4)

MTBM (Mean Time between Maintenance) is the average time to retain the components

into operable state through service overhaul or repair between failures of the

components. Mathematically, it is expressed as;

(

) (2.5)

The operational availability is the probability that a system or component is in an

operable state at a specified time. Logistic delay times and administrative downtime for

maintenance are included in the calculation of operational availability. Operational

availability can be calculated by taking a ratio of MTBF over the sum of MTBF and

MDT for specific points in time with lower and upper confidence levels. Inherent

availability is the instantaneous availability in which delay or downtime is not included.

It can be calculated by taking a ratio of MTBF over the sum of MTBF and MTTR. The

evaluation of these parameters helps deciding the criteria responsible for administrative

or logistic delay in failure analysis models. Mathematically, they are expressed as;

( ) (2.6)

( ) (2.7)

In all equations from 2.1 to 2.7; is Mean-time between failure, is a frequency of

failure, is hazard rate, is Uptime, is Mean downtime, is Downtime,

is Mean-time to repair, is Mean-time between maintenance, is Total

time, is Operational availability, is Inherent availability

Table 2.5 shows the month wise summary of major reliability parameters evaluated from

failure data with the help of above discussed mathematical equations (Mishra and Pathak

(2012). The sample calculation of reliability parameters for the month of April 2013 is

also illustrated as under:


42

Sample Calculation:

=

(

)

(

) = 7.38

( )

( )

= 0.78 (78 %)

( )

( )

= 0.92 (92 %)


43

TABLE 2.5 Month-wise summary of failure data of rolling mill – reliability modelling (Duration: April -2013 to March - 2014)

Sr.

No. Month-Year

Total Run

Time (Hrs.)

[1 day

shutdown]

Total

Uptime

(Hrs.)

Total

Down Time

(hrs.)

Freq. of Failure

(n)

MTBF

(HRS.)

HAZARD RATE

(HRS.)

MDT

(HRS.)

OPERATION AVAILABILITY

(Aop)

MTTR

(HRS.)

MTBM

(HRS.)

INHERENT AVAILABILITY

(Ain)

1 Apr-13 696 546 150 74 7.38 0.14 2.03 0.78 0.61 7.38 0.92

2 May13 720 549 171 86 6.38 0.16 1.99 0.76 0.60 6.38 0.91

3 Jun-13 696 543 153 83 6.54 0.15 1.85 0.78 0.55 6.54 0.92

4 Jul-13 720 558 162 88 6.34 0.16 1.84 0.77 0.55 6.34 0.92

5 Aug-13 720 535 185 94 5.69 0.18 1.97 0.74 0.59 5.69 0.91

6 Sep-13 696 513 183 95 5.40 0.19 1.93 0.74 0.58 5.40 0.90

7 Oct-13 720 540 180 86 6.28 0.16 2.09 0.75 0.63 6.28 0.91

8 Nov-13 696 551 145 80 6.88 0.15 1.82 0.79 0.54 6.88 0.93

9 Dec-13 720 551 169 83 6.63 0.15 2.04 0.76 0.61 6.63 0.92

10 Jan-14 720 527 193 96 5.49 0.18 2.01 0.73 0.60 5.49 0.90

11 Feb-14 648 458 190 94 4.87 0.21 2.02 0.71 0.61 4.87 0.89

12 Mar-14 720 488 233 113 4.31 0.23 2.06 0.68 0.62 4.31 0.87


44

From the reliability modeling, the hazard rate (bathtub) curve and availability curve are

presented as Fig. 2.5 and Fig. 2.6 respectively in a view to understand the failure pattern

behaviour of the components of rolling machine. The hazard rate function is also called

the failure rate function and often used in reliability. It gives an instantaneous failure rate

at time t and used to predict the behaviour of the failure of components. The bathtub

curve is the simplified form of hazard rate function based upon linear and constant

failure rates which represents failure rate component-wise over time.

The lifespan of a rolling machine is assumed long but its failure-free performance mainly

depends on lifetime span of rotating components like; bearings whose lifespan is

expected about million rotations. The obtained curves show the segmental lifetime span

and availability of rolling mill as continuous replacement of components like; bearings

are common practice. Thus, the obtained curve shows a cyclic pattern throughout

lifetime span of rolling mill.

The inherent or operational availability curve shows the probability that the rolling

machine is in the operable state during the specific month. The administrative downtime

is considered in operational availability.

FIGURE 2.5 Hazard rate curve for rolling mill

0.00

0.05

0.10

0.15

0.20

0.25

Haz

ard

Rat

e

Apr-13

May-13

Jun-13

Jul-13

Aug-13

Sep-13

Oct-13

Nov-13

Dec-13

Jan-14

Feb-14

Mar-14

Hazard Rate 0.14 0.16 0.15 0.16 0.18 0.19 0.16 0.15 0.15 0.18 0.21 0.23

Hazard Rate (Bath Tub) Curve

2.5 Discrimination of Critical Components of Rolling Mill

45

FIGURE 2.6 Availability curve for rolling mill

The significance of both the curves is to illustrate the short-term and long-term failure

pattern of the components. The actual time period for failure distribution is widely

ranging considering various failure causes of the components or sub-system. Sometimes

infant mortality may retain till many years and sometime wear-out failure may happen

within a few months. The presentation of such curves helps to conceptualize the overall

failure pattern of the system or plant at instantaneous time span so that appropriate

actions can be accelerated.


The substantial historical failure data such as downtime, failure frequencies etc. as

displayed in Table 2.4 are used to identifying vital components. The frequency of

failures represents the total instant of failures of each part of all fifteen stands as a whole.

The downtime represents the average non-working time after every failure. For example;

primary shafts from any stand failed twenty-one times as a whole during the period of

study. Also, each primary shaft failure needs four hours (recorded based on shop-floor

practices for each component) to restore and/or replacement with total eighty-four hours

downtime. Fig. 2.7 shows graphical presentation of the criticalities through important

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

Ava

ilab

iliti

es

Apr-13

May-13

Jun-13

Jul-13

Aug-13

Sep-13

Oct-13

Nov-13

Dec-13

Jan-14

Feb-14

Mar-14

Operational Availability 0.78 0.76 0.78 0.77 0.74 0.74 0.75 0.79 0.76 0.73 0.71 0.68

Inherent Availability 0.92 0.91 0.92 0.92 0.91 0.90 0.91 0.93 0.92 0.90 0.89 0.87

Availability Curve


46

reliability parameters as discussed. Fig. 2.8 shows graphical presentation of the criticality

analysis based on loss of production in terms of volume and cost.

FIGURE 2.7 Criticality curve for reliability parameters (rolling machine components)

FIGURE 2.8 Criticality curve based on losses in production volume and cost

(Part-wise)

0

50

100

150

200

250

300

350

400

450

500

Cri

tica

lity

Val

ue

Rolling Mill Components

Total DownTime (Hrs)

Frequency ofFailures

0

200

400

600

800

1000

1200

Pri

mar

y Sh

aft

Brg

Sec

on

dar

y H

ou

sin

g

Pri

mar

y B

evel

ge

ar…

Pin

fo

r En

try

Gu

ide

…

Splin

e S

ide

Ch

uck

Nu

t

Bo

tto

m N

ut

for…

Top

Nu

t fo

r…

Cyl

ind

er P

in f

or…

Spac

er

for

Ou

ter

Seco

nd

ary

Blo

ck…

Brg

Ho

usi

ng

12

0ɸ

fo

r …

Be

arin

g N

o. 3

03

10

Be

arin

g N

o. 3

22

22

Oil

Seal

. 60

80

10

…

Co

iler

Bo

lt

Cri

tica

lity

Val

ue

Rolling Mill Components

Losses inProduction(Tonnes)Losses inProduction(Cost)


47

The vital parts or components by explicating shop-floor data are as follows:

1. Bearings: The rolling machine consists of two types of bearings; ball bearings

(No. 6213) and taper roller bearings (No. 32308, 30310, 32222). The bearings

are used to impart motion and support dynamic loads. These are rotating

machining elements and found most critical with about 70 % failure

contributions. It is observed 100 % replacement of failed or time-worn bearings

with obscene practice of mountings which leads much failure rate.

2. Gears: The rolling machine is having bevel gearing like; the primary bevel gears

of spigot end, primary bevel gear of tapered end and secondary bevel gear – ring

type. These gears are used to transmit power and feed motion to unprocessed

aluminium rod. The gears are contributing 4 % which is comparatively low

against bearings (70%) but their role in association with bearings leads them a

second most vital part in a rolling machine.

3. Shafts: In rolling machine, the primary and secondary machining shafts are fitted

for power transmission with bearings and gears. Their failure contributions are

about 4% but seem critical due to their functional link with bearings and gears.

The appropriate functioning of above parts provides a proper feed to the aluminium wire

for size reduction at every stage of a stand. These parts are working in conjunction with

each other at high speed with dynamic loads due to which they are considered the most

vital parts of an aluminium rolling machine.

The other remaining parts or components are contributing about 22 % loss due to their

failures. It is noted that all other parts are not having a significant effect on overall

performance of a rolling machine and they are replaced under the standard replacement

of three identified vital parts; bearings, gears and shafts. These critical components

(bearings, gears, and shafts) are usual components to nearly all processing units.

Hence, it is decided to have modified failure mode effect and criticality (FMECA)

analysis of such components through different MCDM methods to prioritize

maintenance practices to enhance overall reliability of the plant.


48

2.6 Summary

This chapter presents an introduction and overview of the identified process industry

with plant layout, process flow, major components etc. It also shows the substantial

historical failure data recorded for the duration of April 2013 to March 2014 at Sampat

Heavy Engineering Ltd., Ahmedabad, India. The reliability modelling and process of

discrimination of the critical parts are presented based on the failure data and shop-floor

condition in this chapter.

The next chapter discusses the FMEA derived through failure modes, causes, effects and

the failure pattern consequence and problems faced in the present maintenance practices

of indicated critical components. It also discusses the procedure of selection of criteria

and score assignment methodology to every failure cause traditional as well as MCDM

based failure analysis models. Moreover, it discusses the traditional failure models and


49

CHAPTER 3

Failure Pattern Study, Criteria Selection, Score

Assignment and Traditional Approach

3.1 Failure Pattern Study of Critical Components through Failure

Mode and Effect Analysis (FMEA)

3.1.1 Overview

Failure Pattern study involves analyzing functional failures, failure causes, effects and

repercussions of failures with existing control and maintenance practices. Failure Mode

Effect Analysis (FMEA) is commonly used tool for planning such activities of

processing plants through reliability analysis. The FMEA is technological tool to explain,

explore and reduce possible issues associated with parts or components of the

manufacturing unit (Braaksma et al., 2013)

The FMEA consists of following step by step procedure:

(i) Deciding the key process inputs

(ii) Identifying the potential failure modes to each process inputs

(iii) Finding the reasons for each potential failure modes

(iv) Discussing the effects of every reasons/causes

Failure Pattern Study, Criteria Selection, Score Assignment and Traditional Approach

50

The primary elements of FMEA are;

(i) Failure mode: It is the way by which the system or component fails to

perform as designated.

(ii) Effect: It is the impact incorporated due to failure mode; and

(iii) Cause: It is the reason due to which the system or component bring about

failure mode.

It is very interesting to know about the relationship between these elements. There is a

chance of single cause with multiple effects or a combination of causes with a single

effect. To understand FMEA properly, it is required to have an idea about types of

different failure modes. The potential failure modes are; complete or partial failures,

intermittent failures, over-time failure, incorrect or premature operation, and failure due

to non-functioning of parts prematurely or up to design life. It is necessary to

contemplate that a part may have single or combinational failures.

3.1.2 FMEA for discriminated Critical Components

The shop-floor activities and live failure pattern including; visual inspection of failed

parts – its condition, wear, tear, processing temperature, misalignment, noise etc. of

critical components as discriminated in Chapter 2 is closely studied and observed in a

view to understanding;

(i) important process inputs;

(ii) potential failure modes; i.e., how process inputs failed

(iii) failure causes; i.e., why process inputs failed

(iv) failure effects; i.e., how the impact affected due to failure (external or

internal)?

It provides the concept about current control practices and problems faced during routine

activities. The behaviour of the failure pattern was illustrated by the observations of

definitely failed components. Fig. 3.1 shows some photographs of actual shop-floor

conditions and observations made for such components.

From this comprehensive study, the FMEA of vital components of aluminium wire

rolling mill is modeled. Table 3.1 highlights the FMEA of derived vital components.

3.1 Failure Pattern Study of Critical Components through Failure Mode and Effect Analysis (FMEA)

51

FIGURE 3.1 Some photographs of the oobservations made at shop-floor study


52

TABLE 3.1 FMEA of derived vital parts

Important

Process

Input

Potential

Failure

Mode

Possible Causes Potential Effects

Notation What is

process

input?

How

Process

inputs fail?

Why important

input failed?

How the impact affected

due to failure (customer

or internal)?

Rolling Mill

Bearing

Failure

Bearing

high

temperature

Improper

lubrication &

defective sealing

Bearing gets

jammed/Bearing housing

jammed

C1

Bearing

corrosion

Higher speed than

specified

Increase in vibration &

noise C2

Bearing

fatigue

Design defects,

Bearing dimension

not as per

specification

Life reduction C3

Roller balls

wear- out

Foreign

matters/particles Sudden rise in thrust C4

Bearing

misalignme

nt &

improper

mounting

Sudden impact on

the rolls

Shaft damage due to

impact on other parts C5

Electrical

damage Power loss Process interruption C6

Rolling Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication - dirt,

viscosity issues

Rough operation and

considerable noise C7

Gear teeth

surface

fatigue

(Pitting)

Improper meshing,

case depth and

high residual

stresses

Gear life reduction C8

Gear teeth

scoring

Overheating at

gear mesh

Interference and backlash

phenomenon C9

Gear teeth

fracture

Excessive overload

and cyclic stresses

Sudden stoppage of

process plant C10

Gear teeth

surface

cold/plastic

flow

Large contact

stresses due to

rolling and sliding

meshing

Slippage and power loss C11

Rolling Mill

Shaft

(Primary &

Secondary)

Failure

Shaft

fretting

Vibratory dynamic

load from bearing Leads to sudden failure C12

Shaft

misalignme

nt

Uneven bearing

load Vibration & fatigue C13

Shaft

fracture

(Fatigue)

Reverse and

repeated cyclic

loading

Sudden stoppage of

process C14

3.2 Selection of Criteria for Criticality Assessment

53

Their failures are considered as key process inputs to deduce the FMEA. C1 to C14

represents the different potential failure modes/causes.

3.2 Selection of Criteria for Criticality Assessment

The selection of criteria or attributes needs instinctive decisions and constructive

thoughts for explicit determination of criticality of various failure modes. In this research

study, three basic as well as four advanced criteria are proposed in order to have better

criticality assessment, which are as follows:

3.2.1 Traditional Criteria

(1) Probability of chances of failure (P): This criterion represents the possibilities of

failure happens during runtime. The frequency of failures is having major impact

on criticality level so, this criterion is important in the study.

(2) Degree of detectability (D): This criterion highlights at what level maintenance

person can identify failures through visual looking or observing conditions of

elements. The lower score assumes easily detectable.

(3) Degree of severity (S): This criterion explains the impact of severity on working

of parts or components. The large value is assigned to high downtime and repairs

or replaces time.

The above criteria are incorporated in finding criticalities through traditional approach.

3.2.2 Advanced Criteria

(1) Maintainability (M): The term severity is replaced with the maintainability (M)

criteria which define the difficulty level of doing service, repairing or exchange.

The lower score value states less difficulty in maintaining components.

(2) Spare Parts (SP): Looking to the difference obtained between operational and

inherent availability during reliability modeling, spare part (SP) criterion

selected. This criterion interprets the level of reserve stocks of spare parts

available during sudden failures. The larger score value represents non-

availability of spares.

(3) Economic Safety (ES): This criterion is claiming the safety measures of resources

of the unit. The lower score refers to a safer environment.


54

(4) Economic Cost (EC): This criterion considers the various costs associated with

processing, spares, machinery, manpower etc. It states that large score value

refers more costly things.

The multi-criteria decision-making based failure analysis models are incorporated with

above four advanced criteria together with the probability of chances of failure (P) and

degree of detectability (D) looking to their impact on criticality assessment.

3.3 Score Assignment Methodology

3.3.1 Score Assignment for Traditional Approach

Traditional failure analysis needs score ranking for every mode of failure with different

criteria. Here, every failure modes of targeted vital parts are assessed with basic criteria

like; probability of chances of failure (P), degree of detectability (D) and degree of

severity (S).

The independent scores for each failure cause are decided with following consideration;

(i) past records of failures; that provides a detailed outline about failure consequence of

parts or component extracted and,

(ii) through questioning; to shop-floor technicians, plant engineers, service managers

etc.

These scores are assigned on 1 to 10 levels by considering criteria effects from lowest to

highest concern respectively. These values are chosen based on realistic shop-floor

condition.

With the help of the following tables; the scores are assigned to each failure cause for

every process input during FMEA. Table 3.2 shows the score for probability of failure

from no occurrence to high occurrence. Table 3.3 presents the score of this criterion from

immediate to impossible detectability of problems. Table 3.4 highlights the score of

severity from high service duration affected to almost nil.


55

TABLE 3.2 Scores for probability of occurrence (P)

Occurrence Criteria Score

Never More than three year 1

Very infrequent Once every 2-3 year 2

Infrequent Once every 1-2 year 3

Very less Once every 11-12 month 4

Less Once every 9-10 month 5

Medium Once every 7-8 month 6

Medium High Once every 5-6 month 7

High Once every 3-4 month 8

Too High Once every 1-2 month 9

Utmost High Less than 1 month 10

TABLE 3.3 Scores for degree of detectability (D)

Detection

Chances Non-detection level (%) Score

Instant < 10 1

Best 10 to 20 2

Better 21 to 30 3

Good 31 to 40 4

Easy 41 to 50 5

Periodic 51 to 60 6

Overdue 61 to 70 7

Hard 71 to 80 8

Very Hard 81 to 90 9

Impossible 91 to 100 10

TABLE 3.4 Scores for degree of severity (S)

Severity

impact Affected duration Score

Never < 30 min. 1

Very infrequent 1 hour 2

Infrequent 2 hour 3

Very less 3 hour 4

Less 4 hour 5

Medium 5 hour 6

Medium High 6 hour 7

High 7 hour 8

Too High 8 hour 9

Utmost High >8 hour 10


56

3.3.2 Score Assignment for MCDM based Failure Analysis Models

Multi-criteria based failure analysis needs accurate score ranking to each failure mode

with different criteria. The work emphasizes the score assigned to the failure causes (C1-

C14) of derived vital components on six different criteria P, D, M, SP, ES and ED as

discussed in Section 3.2.

The process of selection of score value is the perception of maintenance personnel

depending upon their experience and expertise. It is challenging to represent shop-floor

activities and past performance records in number. To accurately model such task special

care should be taken on certain aspects such as; manufacturing environment, processing

terms, the skill level of manpower etc.

These scores are given a number on 1 to 10 levels by considering attributes effects from

lowest to highest concern respectively. These scores are chosen based on realistic shop-

floor condition and questionnaires to maintenance personnel as discussed in earlier

section.

The probability of chances of failure (P) and degree of detectability (D) is kept same as

per traditional failure analysis model. Table 3.2 and 3.3 are used to model score for these

criteria. Table 3.5 shows the score of maintainability (M) criterion from extremely high

to almost nil maintainability of parts. Table 3.6 presents the score of spare parts (SP)

criterion from easy availability to impossible to procure with an urgent need of spare

parts. Table 3.7 proposed the score of economic safety (ES) criterion from extremely

less safe to the extremely safer working atmosphere. Table 3.8 represented the score of

economic cost (EC) criterion from extremely less costly to extremely more costly

resources.

The scores are in exact value for TOPSIS and PSI approach and in number range (lower

and upper limit) for COPRAS-G approach to compensate practical limitations and

variations of the maintenance personnel in a decision of scores.


57

TABLE 3.5 Scores for maintainability (M)

Chances of

detection Maintainability level (%) Score

Utmost High < 10 1

Too High 10 to 20 2

High 21 to 30 3

Medium High 31 to 40 4

Medium 41 to 50 5

Less 51 to 60 6

Very Less 61 to 70 7

Infrequent 71 to 80 8

Very infrequent 81 to 90 9

Never 91 to 100 10

TABLE 3.6 Scores for spare parts (SP)

Criteria of procurement and need Score

Easily procured and desirable 1

Easily procured and essential 2

Easily procured and too essential 3

Difficult to procure but desirable 4

Difficult to procure but essential 5

Difficult to procure but too essential 6

Rarely available and desirable 7

Rarely available and essential 8

Rarely available and very essential 9

Impossible to avail and acute 10

TABLE 3.7 Scores for economic safety (ES)

Criteria for economic safety Score

Extremely less 1

Very less 2

Less 3

Fair 4

Average 5

Medium 6

Medium high 7

High 8

Too high 9

Utmost high 10


58

TABLE 3.8 Scores for economic cost (EC)

Criteria for economic cost Score

Extremely less 1

Very less 2

Less 3

Fair 4

Average 5

Medium 6

Medium high 7

High 8

Too high 9

Utmost high 10

3.4 Traditional Failure Analysis Approach

3.4.1 Overview

The traditional failure analysis model is incorporated with criticality analysis in FMEA,

which is termed as failure mode effect and criticality (FMECA). Fig. 3.2 highlights the

general flow diagram of traditional FMECA; in which parts or components are reviewed

initially thereafter scores are allotted with brainstorming approaches to evaluate risk

priority number (RPN) for effective prioritization process.

FIGURE 3.2 General flow process of traditional FMECA

FMECA is composed of three steps: (i) analysis of failure modes and their effects with

consequences (ii) assignment of scores to each failure cause and (iii) criticality analysis.

In criticality analysis; the risk is quantified with RPN, which is determined through

multiplication of scores of basic criteria such as; probability of chances of failure (P),

degree of severity (S), and degree of detectability (D).


59

The step-by-step FMECA method for planning maintenance activities as suggested by

Mcdermott et al. (2009) is as under:

(i) Review of the process, system or component.

(ii) Brainstorming the modes of possible failure and gist their effects for further

analysis

(iii)Give a score to each cause of the failure based on its occurrence, severity and

detection criteria

(iv) Determine the RPN by multiplying the score of each criterion of failure mode

(v) Arrange the failure modes to their rank value order to prioritize the maintenance

action and implementation.

3.4.2 Criticality Assessment based on Risk Priority Number (RPN)

During the initial stage of interpretation of criticalities, it is necessary to establish the

decision-matrix by giving score-number to modes of failure (C1 to C14) for fundamental

criteria (P, D and S) as per method described in Section 3.2.1. Table 3.9 is representing

score assignment in form of decision-matrix for determining RPN through multiplication

of score-number as expressed below;

(3.1)

Where; i = 1, 2… n (choices) and j = 1, 2… m (attributes)

For example; RPN for failure cause C1 is;

and Table 3.10 displays the RPN obtained through traditional FMECA.


60

TABLE 3.9 Decision matrix – X for traditional FMECA

P D S

Poss

ible

cau

ses

of

Fai

lure

Pro

bab

ilit

y o

f

Chan

ce o

f fa

ilure

Deg

ree

of

Det

ecta

bil

ity

Deg

ree

of

Sev

erit

y

C1 5 8 7

C2 3 6 4

C3 9 7 10

C4 8 6 7

C5 5 5 8

C6 2 1 7

C7 5 2 5

C8 8 5 8

C9 7 6 4

C10 8 6 8

C11 5 3 5

C12 5 6 5

C13 7 7 8

C14 7 4 8

TABLE 3.10 Risk priority number (RPN) for traditional FMECA

Notation Possible or potential causes of failure Rank

C1 Improper lubrication & defective sealing 280 7

C2 Higher speed than specified 72 12

C3 Design defects, Bearing dimension not as

per specification 630 1

C4 Foreign matters/particles 336 4

C5 Sudden impact on the rolls 320 5

C6 Power loss 14 14

C7 Inadequate lubrication - dirt, viscosity

issues 50 13

C8 Improper meshing, case depth and high

residual stresses 320 6

C9 Overheating at gear mesh 168 9

C10 Excessive overload and cyclic stresses 384 3

C11 High contact stresses due to rolling and

sliding action of mesh 75 11

C12 Vibratory dynamic load from bearing 150 10

C13 Uneven bearing load 392 2

C14 Reverse and repeated cyclic loading 224 8


61

3.4.3 Maintenance Planning Through Traditional FMECA

By analyzing current maintenance practices with an outcome of RPN, modified and

constructive maintenance practices are proposed as per Table 3.11. The classification of

various failure modes is done through RPN value. The failure modes are considered most

vital having 500 or more RPN value and recommended predictive or condition-based

approaches. The failure modes having RPN value between 250 and 500 are considered

critical and recommended preventive strategies during a shutdown. The failure modes

having RPN value below 250 are termed as normal failures and proposed corrective

actions when prompted breakdown.

TABLE 3.11 RPN based FMEA with existing practices and proposed improvements in

maintenance plan

Particulars

Current

Controls

Su

gg

este

d I

mpro

vem

ent

in

Mai

nte

nan

ce P

lan

No

tati

on

Key

Process

Input

Potential

Failure Mode

Potential

Causes

Potential

Failure Effects

What is

process

input?

How Process

inputs fail?

Why

important

input failed?

How the

impact

affected due to

failure

(customer or

internal)?

What are the

existing

practices

through

which

failure mode

controlled?

Rolling

Mill

Bearing

Failure

Bearing high

temperature

Improper

lubrication &

defective

sealing

Bearing gets

jammed/Beari

ng housing

jammed

Lubricating

the parts

when

occurred

Preventive

Maintenance C1

Bearing

corrosion

Higher speed

than specified

Increase in

vibration &

noise

Proper

coolant

Corrective

Maintenance C2

Bearing

fatigue

Defects in

design,

Bearing

dimension

not as per

specification

Life reduction Bearing

replacement

Predictive

Maintenance C3

Roller balls

wear- out

Foreign

matters/partic

les

Sudden rise in

thrust

Regular

cleaning of

parts

Preventive

Maintenance C4

Bearing

misalignment

& improper

mounting

Sudden

impact on the

rolls

Shaft damage

due to impact

on other parts

Routine

check up

Preventive

Maintenance C5


62

Electrical

damage Power loss

Process

interruption

Electrical

wiring check

up

Corrective

Maintenance C6

Rolling

Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication –

Dirt, viscosity

issues

Rough

operation &

considerable

noise

Routine

check-up of

lubrication

Corrective

Maintenance C7

Gear teeth

surface

fatigue

(Pitting)

Improper

meshing, case

depth & high

residual

stresses

Gear life

reduction

Preventive

maintenance

Preventive

Maintenance C8

Gear teeth

scoring

Overheating

at gear mesh

Interference &

backlash

phenomenon

Lubricating

when

needed

Corrective

Maintenance C9

Gear teeth

fracture

Excessive

overload &

cyclic

stresses

Sudden

stoppage of

process plant

Break down

maintenance

Preventive

Maintenance C10

Gear teeth

surface

cold/plastic

flow

Large contact

stresses due

to rolling and

sliding

meshing

Slippage &

power loss

Gear replace

when

needed

Corrective

Maintenance C11

Rolling

Mill Shaft

(Primary &

Secondary)

Failure

Shaft fretting

Vibratory

dynamic load

from bearing

Leads to

sudden failure

Break down

maintenance

Corrective

Maintenance C12

Shaft

misalignment

Uneven

bearing load

Vibration &

fatigue

Preventive

maintenance

Preventive

Maintenance C13

Shaft fracture

(Fatigue)

Reverse &

repeated

cyclic loading

Sudden

stoppage of

process

Preventive

maintenance

Preventive

Maintenance C14

3.4.4 Drawbacks of Traditional FMECA

Some drawbacks of traditional FMECA are notes as under;

(i) The traditional FMECA is only working on specific fundamental criteria.

(ii) The same weightage is given to all criteria

(iii) Pertinence or inter-connection between criteria is not considered during the

assignment of scores and;

(iv) Minor changes in scores lead to discrepancy in RPN due to a multiplication of

mathematics. As RPN is the multiplication of score of three basic criteria, the

little changes of assignment of score due to personal perspective leads large

variations.

3.5 Summary

63

The above drawbacks of this failure analysis model can be resolved using MCDM based

failure models.

3.5 Summary

This chapter discusses the FMEA derived through failure modes, causes, effects and the

failure pattern consequence and problems faced in the present maintenance practices of

indicated critical components. It also discusses the procedure of selection of criteria and

score assignment methodology to every failure cause for traditional as well as MCDM

based failure analysis models. Moreover, it discusses the traditional failure models and


The next chapter presents MCDM based failure analysis models with an addition of

some more advanced criteria. It also describes methods to evaluate MCI for each failure

mode of targeted critical components through three different MCDM failure analysis

models called; TOPSIS, COPRAS-G, and PSI for optimizing current maintenance

strategies of them.

64

CHAPTER 4

Multi-criteria Decision-making based Failure

Analysis Models

4.1 Overview

The MCDM approach is very easy to use with covering more criteria. It compares the

alternatives relatively on weights which helps the decision-making process effective.

This chapter describes a method for evaluating i.e. maintainability criticality index

of all causes of failures C1 to C14 as suggested through three different MCDM failure

analysis models, first based on TOPSIS, second based on COPRAS-G and third based on

PSI in a view to optimize existing maintenance plan of critical components of aluminium

wire rolling mill.

Fig. 4.1 shows the general flow diagram of MCDM based FMECA process incorporated

in this study which shows the assignment of scores to designated failure causes for three

distinct failure models in a view to enhance priority by determining

4.2 TOPSIS based Failure Mode Effect and Criticality Analysis

65

FIGURE 4.1 Flow Diagram of MCDM based FMECA process


4.2.1 TOPSIS Methodology

TOPSIS stands for a technique for order preference by similarity to an ideal solution in

which, Euclidean distance is calculated from a classical point. It works on a multi-criteria

concept where attributes are scored in exact number and weights are considered to decide

preference level. The crisp version of TOPSIS was presented by Hwang and Yoon in

1981.

Multi-criteria Decision-making based Failure Analysis Models

66

The evaluation of of listed failure causes of vital or critical components of the

aluminium wire processing unit as discussed previously is done by the TOPSIS method

as recommended by Sachdeva et al. (2009).

The step-by-step process is described as under:

Step – 1: Preparing a decision matrix by tabulating causes of failure and designated

criteria into rows and columns correspondingly.

The said decision matrix is deduced for a set of beneficial attributes (P, D, M, SP, ES,

EC) against a set of fourteen causes of failure (C1 to C14) as per the description of

Section 3.3.2.

Step – 2: Development of decision matrix – by assigning score value in an exact

number

(4.1)

Where; i = 1, 2… n (choices) and j = 1, 2… m (attributes)

Table 4.1 displays decision matrix – which is developed by proper assignment of score

value to matrix cell through the methods described in Section 3.3.2.

Step – 3: Normalizing a decision matrix – .

The process of normalization of decision matrix is explained by Deng et al. (2000).

Moreover, Salabun (2013) proposed a method for calculation of the normalized value of

score for straight beneficial attributes as follows:

∑

(4.2)

The normalized matrix is presented in Table 4.2.


67

TABLE 4.1 Decision matrix – X for TOPSIS

P D M SP ES EC

Pote

nti

al F

ailu

re C

ause

s

Pro

bab

ilit

y o

f ch

ance

of

fail

ure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic s

afet

y

Eco

nom

ic c

ost

C1 9 8 1 3 3 3

C2 8 6 2 2 4 3

C3 10 7 6 3 10 9

C4 9 6 5 3 7 5

C5 10 5 6 5 9 10

C6 9 1 1 3 5 2

C7 7 3 5 3 7 4

C8 8 5 5 3 5 5

C9 5 4 2 3 3 3

C10 9 2 6 4 7 7

C11 3 6 3 3 3 3

C12 5 5 4 3 3 3

C13 8 5 5 3 6 6

C14 9 2 6 4 6 7

for profit criteria:

=

(

) (4.3)

=

(4.4)


68

Step – 5: Determining all weighted attributes.

With the help of Shannon’s entropy concept, weights of all criteria are calculated by first

evaluating entropy of jth

criteria as per the following expression,

= -

∑

(4.5)

TABLE 4.2 Normalization of decision matrix – X for TOPSIS

P D M SP ES EC

Pote

nti

al F

ailu

re

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f fa

ilure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic s

afet

y

Eco

nom

ic c

ost

C1 0.0826 0.1231 0.0175 0.0667 0.0385 0.0429

C2 0.0734 0.0923 0.0351 0.0444 0.0513 0.0429

C3 0.0917 0.1077 0.1053 0.0667 0.1282 0.1286

C4 0.0826 0.0923 0.0877 0.0667 0.0897 0.0714

C5 0.0917 0.0769 0.1053 0.1111 0.1154 0.1429

C6 0.0826 0.0154 0.0175 0.0667 0.0641 0.0286

C7 0.0642 0.0462 0.0877 0.0667 0.0897 0.0571

C8 0.0734 0.0769 0.0877 0.0667 0.0641 0.0714

C9 0.0459 0.0615 0.0351 0.0667 0.0385 0.0429

C10 0.0826 0.0308 0.1053 0.0889 0.0897 0.1000

C11 0.0275 0.0923 0.0526 0.0667 0.0385 0.0429

C12 0.0459 0.0769 0.0702 0.0667 0.0385 0.0429

C13 0.0734 0.0769 0.0877 0.0667 0.0769 0.0857

C14 0.0826 0.0308 0.1053 0.0889 0.0769 0.1000


69

Later on, weights are determined for jth

criteria through following expression;

= –

∑ –

(4.6)

Step – 6: Determining the distance between a positive ideal solution and negative ideal

solution

respectively.

The expressions as listed below are used to determine the distance from ideal value;

√∑

(4.7)

√∑

(4.8)

Table 4.3 represents the value of

for each failure cause.

Step – 7: Determining criticality index for TOPSIS;

The same is determined as per the following expression;

=

(4.9)

Table 4.4 shows the obtained value of and its criticality rank.


70

TABLE 4.3 Distances between positive and negative ideal solution

P D M SP ES EC

Pote

nti

al

Fai

lure

Cau

ses

Pote

nti

al

Fai

lure

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f

fail

ure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic

safe

ty

C1 0.0005 0.0170 0.0000 0.0634 0.0418 0.0000 0.0111 0.0028 0.0444 0.0000 0.0545 0.0011

C2 0.0019 0.0118 0.0052 0.0323 0.0268 0.0017 0.0251 0.0000 0.0327 0.0009 0.0545 0.0011

C3 0.0000 0.0231 0.0013 0.0466 0.0000 0.0418 0.0111 0.0028 0.0000 0.0444 0.0011 0.0545

C4 0.0005 0.0170 0.0052 0.0323 0.0017 0.0268 0.0111 0.0028 0.0082 0.0145 0.0278 0.0100

C5 0.0000 0.0231 0.0117 0.0207 0.0000 0.0418 0.0000 0.0251 0.0009 0.0326 0.0000 0.0711

C6 0.0005 0.0170 0.0635 0.0000 0.0418 0.0000 0.0111 0.0028 0.0227 0.0036 0.0712 0.0000

C7 0.0042 0.0076 0.0324 0.0052 0.0017 0.0268 0.0111 0.0028 0.0082 0.0145 0.0401 0.0044

C8 0.0019 0.0118 0.0117 0.0207 0.0017 0.0268 0.0111 0.0028 0.0227 0.0036 0.0278 0.0100

C9 0.0118 0.0019 0.0207 0.0116 0.0268 0.0017 0.0111 0.0028 0.0444 0.0000 0.0545 0.0011

C10 0.0005 0.0170 0.0466 0.0013 0.0000 0.0418 0.0028 0.0112 0.0082 0.0145 0.0100 0.0278

C11 0.0231 0.0000 0.0052 0.0323 0.0151 0.0067 0.0111 0.0028 0.0444 0.0000 0.0545 0.0011

C12 0.0118 0.0019 0.0117 0.0207 0.0067 0.0151 0.0111 0.0028 0.0444 0.0000 0.0545 0.0011

C13 0.0019 0.0118 0.0117 0.0207 0.0017 0.0268 0.0111 0.0028 0.0145 0.0081 0.0178 0.0178

C14 0.0005 0.0170 0.0466 0.0013 0.0000 0.0418 0.0028 0.0112 0.0145 0.0081 0.0100 0.0278


71

TABLE 4.4 Maintainability criticality index and criticality rank for TOPSIS

Notation Potential Failure Causes Rank

C1 Improper lubrication & defective sealing 0.4265 9

C2 Higher speed than specified 0.3640 10

C3 Design defects, bearing dimension not as

per specification 0.7986 2

C4 Foreign matters/particles 0.5794 3

C5 Sudden impact on the rolls 0.8051 1

C6 Power loss 0.2499 14


issues 0.4419 8

C8 Improper meshing, case depth and high

residual stresses 0.4981 7

C9 Overheating at gear mesh 0.2515 13

C10 Excessive overload and cyclic stresses 0.5636 4

C11 Large contact stresses due to rolling and

sliding meshing 0.3460 12

C12 Vibratory dynamic load from bearing 0.3525 11

C13 Uneven bearing load 0.5505 5

C14 Reverse and repeated cyclic loading 0.5455 6

4.2.2 Maintenance Planning through TOPSIS FMECA

It is noted from the Table 4.4 that failure-cause sudden impact on roll (C5) looks the vital

or most-critical and power loss (C6) looks minimal critical. Table 4.5 shows the failure

mode effect analysis with current control practices and suggested maintenance plan

based on their criticalities. It is recommended updating existing maintenance practices as

listed in Table 4.5 that causes of failure (C5, C3, C4, C10, and C13) with high

is covered with condition monitoring based predictive strategy, causes of failure (C14,

C8, C7, C1, C2) with medium is covered with preventive approaches which

are targeted during plant shutdown and causes of failure (C12, C11, C9, C6) with small

is approached remedial measures when prompted.


72

TABLE 4.5 based FMEA with existing practices and proposed improvements

in maintenance plan

Particulars

Current

Controls

Su

gg

este

d I

mpro

vem

ent

in

Mai

nte

nan

ce P

lan

No

tati

on

Key

Process

Input

Potential

Failure Mode

Potential

Causes

Potential

Failure Effects

What is

process

input?

How Process

inputs fail?

Why

important

input failed?

How the

impact

affected due to

failure

(customer or

internal)?

What are the

existing

practices

through

which

failure mode

controlled?

Rolling

Mill

Bearing

Failure

Bearing high

temperature

Improper

lubrication &

defective

sealing

Bearing gets

jammed/Beari

ng housing

jammed

Lubricating

the parts

when

occurred

Preventive

Maintenance C1

Bearing

corrosion

Higher speed

than specified

Increase in

vibration &

noise

Proper

coolant

Preventive

Maintenance C2

Bearing

fatigue

Design

defects,

Bearing

dimension

not as per

specification


replacement

Predictive

Maintenance C3

Roller balls

wear- out

Foreign

matters/partic

les

Sudden rise in

thrust

Regular

cleaning of

parts

Predictive

Maintenance C4

Bearing

misalignment

& improper

mounting

Sudden

impact on the

rolls

Shaft damage

& Impact

damage on

other parts

Routine

check up

Predictive

Maintenance C5

Electrical

damage

Loss of

power

Operation

interrupted

Electrical

wiring check

up

Corrective

Maintenance C6

Rolling

Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication –

Dirt, viscosity

issues

Rough

operation &

considerable

noise

Routine

check-up of

lubrication

Preventive

Maintenance C7

Gear teeth

surface

fatigue

(Pitting)

Lack of

proper

meshing of

case depth &

high residual

stresses

Gear life

reduction

Preventive

maintenance

Preventive

Maintenance C8

Gear teeth

scoring

Overheating

at gear mesh

Interference &

backlash

phenomenon

Lubricating

when

needed

Corrective

Maintenance C9

Gear teeth

fracture

Excessive

overload &

cyclic

stresses

Sudden

stoppage of

process plant

Break down

maintenance

Predictive

Maintenance C10

4.3 COPRAS-G based Failure Mode Effect and Criticality Analysis

73

Gear teeth

surface

cold/plastic

flow

Large contact

stresses due

to rolling and

sliding

meshing

Slippage &

power loss

Gear replace

when

needed

Corrective

Maintenance C11

Rolling

Mill Shaft

(Primary &

Secondary)

Failure

Shaft fretting

Vibratory

dynamic load

from bearing

Leads to

sudden failure

Break down

maintenance

Corrective

Maintenance C12

Shaft

misalignment

Uneven

bearing load

Vibration &

fatigue

Preventive

maintenance

Predictive

Maintenance C13

Shaft fracture

(Fatigue)

Reverse and

repeated

cyclic loading

Process

suddenly

stopped

Preventive

maintenance

Preventive

Maintenance C14


4.3.1 COPRAS-G Methodology

COPRAS-G stands for grey-complex proportional risk assessment which works on grey

number concept. As discussed by Zavadskas et al. (2008, 2009), the data facts are

grouped into three numbers called; black, white and grey. The black number has no

higher as well as lower ranges. A white number is an exact number which means, the

upper and lower values are the same. Whereas the grey number has dissimilar values

with both above and below ranges so, it falls within an interval called the grey range

(Maity et al., 2012). This concept of the grey range was deduced from grey theory, which

helps in dealing uncertainty of real environment (Deng 1989, Chang et al. 1999 and Lin

et al. 2008).

The evaluation of of listed failure causes of vital or critical components of an

aluminium wire processing unit is based on following procedure (Zavadskas et al., 2008,

2009 and Maity et al., 2012):

Step – 1: Preparing a decision matrix by tabulating causes of failure and designated

criteria in to rows and columns correspondingly.

The decision matrix is deduced for a set of beneficial attributes (P, D, M, SP, ES, EC)

against a set of fourteen causes of failure (C1 to C14) as per description of Section 3.3.2.

Step – 2: Development of decision matrix – by assigning scores value in upper and

lower grey range as under.


74

= ] = [

] (4.10)

Where; and represents low and high scores in grey-range respectively.

i = 1, 2… m representing causes of failure

j = 1, 2… n representing profit criteria

The decision matrix – as shown in Table 4.6 is developed by proper assignment of

score values in a range to matrix cell through the methods described in Section 3.3.2.

Step – 3: Normalizing decision matrix- .

Table 4.7 shows the normalized matrix- which derived from expression as under;

=

∑ ∑

(4.11)

=

∑ ∑

(4.12)

= [

] (4.13)

The primary objectives of the normalizing process to organize the data with acceptable

logic and dependencies with less redundancy.

Step – 4: Determining weighted attributes.

With the help of Shannon’s entropy concept, weights of all criteria are calculated by first

evaluating entropy for both range and of jth

criteria as per the following

expression,

= -

∑

(4.14)

= -

∑

(4.15)

Thereafter weights are calculated similarly as follows;


75

= –

∑ –

(4.16)

= –

∑ –

(4.17)

TABLE 4.6 Decision matrix – X for COPRAS-G

P D M SP ES EC

Pote

nti

al F

ailu

re

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f fa

ilure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic s

afet

y

Eco

nom

ic c

ost

C1 8 9 7 8 1 2 2 3 3 4 3 4

C2 8 9 5 6 1 2 2 3 3 4 4 5

C3 9 10 7 8 6 8 3 5 9 10 9 10

C4 7 9 6 7 4 5 3 5 7 8 5 6

C5 8 10 5 6 6 7 5 7 9 10 9 10

C6 7 9 1 2 1 2 3 4 5 6 2 3

C7 7 8 2 3 5 6 3 4 7 8 4 5

C8 8 9 4 5 5 6 3 4 4 5 5 6

C9 4 5 3 4 2 3 2 3 2 3 3 4

C10 9 10 2 4 6 7 3 4 7 8 7 8

C11 3 4 6 7 3 4 3 4 2 3 3 4

C12 5 6 4 5 3 5 3 4 3 4 3 4

C13 8 9 4 5 5 6 4 5 4 5 5 6

C14 9 10 2 3 6 7 3 4 5 6 6 7

Step – 5: Evaluation of weighted normalized matrix-

This matrix- is generated based on the following equations as displayed in Table 4.8.

= (4.18)

= (4.19)

= [

] (4.20)


76

TABLE 4.7 Normalized Decision Matrix – X1 for COPRAS-G

P D M SP ES EC P

ote

nti

al F

ailu

re

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f

fail

ure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic s

afet

y

Eco

nom

ic c

ost

C1 0.0737 0.0829 0.1069 0.1221 0.0161 0.0323 0.0396 0.0594 0.0390 0.0519 0.0400 0.0533

C2 0.0737 0.0829 0.0763 0.0916 0.0161 0.0323 0.0396 0.0594 0.0390 0.0519 0.0533 0.0667

C3 0.0829 0.0922 0.1069 0.1221 0.0968 0.1290 0.0594 0.0990 0.1169 0.1299 0.1200 0.1333

C4 0.0645 0.0829 0.0916 0.1069 0.0645 0.0806 0.0594 0.0990 0.0909 0.1039 0.0667 0.0800

C5 0.0737 0.0922 0.0763 0.0916 0.0968 0.1129 0.0990 0.1386 0.1169 0.1299 0.1200 0.1333

C6 0.0645 0.0829 0.0153 0.0305 0.0161 0.0323 0.0594 0.0792 0.0649 0.0779 0.0267 0.0400

C7 0.0645 0.0737 0.0305 0.0458 0.0806 0.0968 0.0594 0.0792 0.0909 0.1039 0.0533 0.0667

C8 0.0737 0.0829 0.0611 0.0763 0.0806 0.0968 0.0594 0.0792 0.0519 0.0649 0.0667 0.0800

C9 0.0369 0.0461 0.0458 0.0611 0.0323 0.0484 0.0396 0.0594 0.0260 0.0390 0.0400 0.0533

C10 0.0829 0.0922 0.0305 0.0611 0.0968 0.1129 0.0594 0.0792 0.0909 0.1039 0.0933 0.1067

C11 0.0276 0.0369 0.0916 0.1069 0.0484 0.0645 0.0594 0.0792 0.0260 0.0390 0.0400 0.0533

C12 0.0461 0.0553 0.0611 0.0763 0.0484 0.0806 0.0594 0.0792 0.0390 0.0519 0.0400 0.0533

C13 0.0737 0.0829 0.0611 0.0763 0.0806 0.0968 0.0792 0.0990 0.0519 0.0649 0.0667 0.0800

C14 0.0829 0.0922 0.0305 0.0458 0.0968 0.1129 0.0594 0.0792 0.0649 0.0779 0.0800 0.0933


77

TABLE 4.8 Weighted normalized decision matrix – X2 for COPRAS-G

P D M SP ES EC P

ote

nti

al

Fai

lure

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f

fail

ure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic

safe

ty

Eco

nom

ic c

ost

C1 0.0217 0.0235 0.0178 0.0195 0.0014 0.0032 0.0027 0.0058 0.0081 0.0100 0.0071 0.0090

C2 0.0217 0.0235 0.0127 0.0146 0.0014 0.0032 0.0027 0.0058 0.0081 0.0100 0.0095 0.0112

C3 0.0244 0.0261 0.0178 0.0195 0.0083 0.0127 0.0040 0.0096 0.0243 0.0250 0.0213 0.0225

C4 0.0190 0.0235 0.0153 0.0171 0.0056 0.0080 0.0040 0.0096 0.0189 0.0200 0.0119 0.0135

C5 0.0217 0.0261 0.0127 0.0146 0.0083 0.0111 0.0066 0.0135 0.0243 0.0250 0.0213 0.0225

C6 0.0190 0.0235 0.0025 0.0049 0.0014 0.0032 0.0040 0.0077 0.0135 0.0150 0.0047 0.0067

C7 0.0190 0.0209 0.0051 0.0073 0.0069 0.0096 0.0040 0.0077 0.0189 0.0200 0.0095 0.0112

C8 0.0217 0.0235 0.0102 0.0122 0.0069 0.0096 0.0040 0.0077 0.0108 0.0125 0.0119 0.0135

C9 0.0109 0.0131 0.0076 0.0098 0.0028 0.0048 0.0027 0.0058 0.0054 0.0075 0.0071 0.0090

C10 0.0244 0.0261 0.0051 0.0098 0.0083 0.0111 0.0040 0.0077 0.0189 0.0200 0.0166 0.0180

C11 0.0081 0.0104 0.0153 0.0171 0.0042 0.0064 0.0040 0.0077 0.0054 0.0075 0.0071 0.0090

C12 0.0136 0.0157 0.0102 0.0122 0.0042 0.0080 0.0040 0.0077 0.0081 0.0100 0.0071 0.0090

C13 0.0217 0.0235 0.0102 0.0122 0.0069 0.0096 0.0053 0.0096 0.0108 0.0125 0.0119 0.0135

C14 0.0244 0.0261 0.0051 0.0073 0.0083 0.0111 0.0040 0.0077 0.0135 0.0150 0.0142 0.0157


78

Step – 6: Calculation of weighted-mean normalized additions of profit criteria whose

higher value is preferred and non-beneficial criteria , whose smaller value is

preferred.

They are calculated as follow;

= -

∑

(4.21)

= -

∑

(4.22)

where; i = 1,2,….,m (total criteria)

It is assumed that profit attributes are k then (m - k) is non-profit attributes. It is common

practice to keep profit attributes at the beginning followed by non-profit attributes.

Step – 7: Measure the relative weight of for available choices

It is deduced with following expressions;

= ∑

∑

(4.23)

= ∑

∑

(4.24)

Where; represents the least additions of weighted-mean normalized values for

unfavourable attributes ,

The mode of failures are prioritized with arranging results of in ascending

manner i.e. the higher value of is given top priority over others.

Step – 8: Determining the (%) contribution for ith

causes of failure;

The percentage contribution is determined to present unity as per following expression

and ranks are decided with ;


79

=

(4.25)

Table 4.9 highlighted the achieved results of , (%) contribution and rank

order.

TABLE 4.9 Maintainability criticality index and (%) contribution for

COPRAS-G

Notation Potential Failure Causes (%) Rank

C1 Improper lubrication & defective sealing 0.1297 60 9

C2 Higher speed than specified 0.1244 58 10


per specification 0.2156 100 1

C4 Foreign matters/particles 0.1662 77 4

C5 Sudden impact on the rolls 0.2079 96 2

C6 Loss of power 0.1062 49 12

C7 Inadequate lubrication - girt, viscosity

issues 0.1401 65 8

C8 Improper meshing, case depth & high

residual stresses 0.1444 67 7

C9 Overheating at gear mesh 0.0863 40 14

C10 Excessive overload and cyclic stresses 0.1700 79 3


sliding meshing 0.1022 47 13

C12 Vibratory dynamic load from bearing 0.1096 51 11

C13 Uneven bearing load 0.1477 68 6

C14 Reverse and repeated cyclic loading 0.1526 71 5

4.3.3 Significance of COPRAS-G

It is very difficult for maintenance personnel to assign the score consistently at various

stages of assessment of criticalities. Thus, an exact interpretation of causes of failure

looks practically difficult. This challenge may be handled by incorporating score-values

into interval called a grey number rather than an exact and specific value of TOPSIS.

The primary objective of this approach is to present score-values in range.

4.3.4 Maintenance Planning through COPRAS-G FMECA

The results of Table 4.9 highlight that failure cause; design defects and lack of

specification in bearing dimensions (C3) looks vital or most-critical and overheating at

gear mesh (C9) looks less important. Table 4.10 shows the failure mode effect analysis

with current control practices and suggested maintenance plan based on criticalities. It is


80

recommended to update the current control practices as listed in Table 4.10. It highlights

that causes of failure (C3, C5, C10, C4, and C14) with large may be covered

with condition monitoring or predictive type of approaches, causes of failure (C13, C8,

C7, C1, and C2) with medium is covered with preventive measures during

shut down and failure causes (C12, C6, C11, and C9) with small value of is

covered by remedial or corrective actions when breakdown prompts.

TABLE 4.10 based FMEA with existing practices and proposed

improvements in maintenance plan

Particulars

Current

Controls

Su

gg

este

d I

mpro

vem

ent

in

Mai

nte

nan

ce P

lan

No

tati

on

Key

Process

Input

Potential

Failure Mode

Potential

Causes

Potential

Failure Effects

What is

process

input?

How Process

inputs fail?

Why

important

input failed?

How the

impact

affected due to

failure

(customer or

internal)?

What are the

existing

practices

through

which

failure mode

controlled?

Rolling

Mill

Bearing

Failure

Bearing high

temperature

Improper

lubrication &

defective

sealing

Bearing gets

jammed/Beari

ng housing

jammed

Lubricating

the parts

when

occurred

Preventive

Maintenance C1

Bearing

corrosion

Higher speed

than specified

Increase in

vibration &

noise

Proper

coolant

Preventive

Maintenance C2

Bearing

fatigue

Design

defects,

bearing

dimension

not as per

specification


replacement

Predictive

Maintenance C3

Roller balls

wear- out

Foreign

matters/partic

les

Sudden rise in

thrust

Regular

cleaning of

parts

Predictive

Maintenance C4

Bearing

misalignment

& improper

mounting

Sudden

impact on the

rolls

Shaft damage

& Impact

damage on

other parts

Routine

check up

Predictive

Maintenance C5

Electrical

damage Power loss

Process

interruption

Electrical

wiring check

up

Corrective

Maintenance C6

Rolling

Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication –

Dirt, viscosity

issues

Rough

operation &

considerable

noise

Routine

check-up of

lubrication

Preventive

Maintenance C7

4.4 PSI based Failure Mode Effect and Criticality Analysis

81

Gear teeth

surface

fatigue

(Pitting)

Improper

meshing, case

depth & high

residual

stresses

Gear life

reduction

Preventive

maintenance

Preventive

Maintenance C8

Gear teeth

scoring

Overheating

at gear mesh

Interference &

backlash

phenomenon

Lubricating

when

needed

Corrective

Maintenance C9

Gear teeth

fracture

Excessive

overload &

cyclic

stresses

Sudden

stoppage of

process plant

Break down

maintenance

Predictive

Maintenance C10

Gear teeth

surface

cold/plastic

flow

Large contact

stresses due

to rolling and

sliding

meshing

Slippage &

power loss

Gear replace

when

needed

Corrective

Maintenance C11

Rolling

Mill Shaft

(Primary &

Secondary)

Failure

Shaft fretting

Vibratory

dynamic load

from bearing

Leads to

sudden failure

Break down

maintenance

Corrective

Maintenance C12

Shaft

misalignment

Uneven

bearing load

Vibration &

fatigue

Preventive

maintenance

Preventive

Maintenance C13

Shaft fracture

(Fatigue)

Reverse and

repeated

cyclic loading

Process

suddenly

stopped

Preventive

maintenance

Predictive

Maintenance C14


4.4.1 PSI Methodology

Maniya and Bhatt (2011) discussed PSI approach. This method is very useful for

incompatible situations to take an effective decision. It is multi-criteria based crisp

approach which works on the principle of statistical tabulation without consideration of

weights to decide preferences.

The evaluation of of listed modes of failure of vital or critical parts of the

aluminium wire processing unit as discussed previously is done by the PSI method as

recommended by Maniya and Bhatt (2011).

The step-by-step process is described as under:

Step – 1: Preparing a decision matrix by arranging designated criteria and failure

causes into columns and rows correspondingly.


82

The said decision matrix is deduced for a set of beneficial attributes (P, D, M, SP, ES,

EC) against a set of fourteen failure causes (C1 to C14) as per description stated in

Section 3.2.2.

Step – 2: Development of decision matrix – by assigning score value in exact number

= = [

] (4.26)

Where; represents fixed scores.

i = 1, 2… m representing causes of failure

j = 1, 2… n representing profit criteria

The decision matrix – as displayed in Table 4.11 is developed by proper assignment of

score value to matrix cell through the methods described in Section 3.3.2.

TABLE 4.11 Decision matrix – X for PSI

P D M SP ES EC

Pote

nti

al

Fai

lure

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f

fail

ure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic

safe

ty

Eco

nom

ic c

ost

C1 9 8 1 3 3 3

C2 8 6 2 2 4 3

C3 10 7 6 3 10 9

C4 9 6 5 3 7 5

C5 10 5 6 5 9 10

C6 9 1 1 3 5 2

C7 7 3 5 3 7 4

C8 8 5 5 3 5 5

C9 5 4 2 3 3 3

C10 9 2 6 4 7 7

C11 3 6 3 3 3 3

C12 5 5 4 3 3 3

C13 8 5 5 3 6 6

C14 9 2 6 4 6 7


83

Step – 3: Normalizing decision matrix -

Considering likelihood level of attributes, normalizing of score-values is done with the

help of expressions as described under and the same is presented in Table 4.12.

Considering likelihood more for profit attributes;

=

(4.27)

Considering likelihood less for profit attributes;

=

(4.28)

where; and are the higher and lower score-value possible causes of failure

correspondingly;

Then;

Normalized decision matrix is as follows;

= [

] (4.29)

Step – 4: Determine deviation in preference

The deviation in preference is determined with an expression as under;

= ∑ ] (4.30)

where; =

∑

(4.31)

Step – 5: Determine deviation of parameter

The deviation in parameter is determined with an expression as under;

= (4.32)


84

TABLE 4.12 Normalized decision-matrix – for PSI

P D M SP ES EC

Pote

nti

al F

ailu

re

Cau

ses

Pro

bab

ilit

y o

f

chan

ce o

f fa

ilure

Deg

ree

of

Det

ecta

bil

ity

Mai

nta

inab

ilit

y

Spar

e par

ts

Eco

nom

ic s

afet

y

Eco

nom

ic c

ost

C1 0.900 1.000 0.167 0.600 0.300 0.300

C2 0.800 0.750 0.333 0.400 0.400 0.300

C3 1.000 0.875 1.000 0.600 1.000 0.900

C4 0.900 0.750 0.833 0.600 0.700 0.500

C5 1.000 0.625 1.000 1.000 0.900 1.000

C6 0.900 0.125 0.167 0.600 0.500 0.200

C7 0.700 0.375 0.833 0.600 0.700 0.400

C8 0.800 0.625 0.833 0.600 0.500 0.500

C9 0.500 0.500 0.333 0.600 0.300 0.300

C10 0.900 0.250 1.000 0.800 0.700 0.700

C11 0.300 0.750 0.500 0.600 0.300 0.300

C12 0.500 0.625 0.667 0.600 0.300 0.300

C13 0.800 0.625 0.833 0.600 0.600 0.600

C14 0.900 0.250 1.000 0.800 0.600 0.700

Step – 6: Determine overall preference parameter

The is evaluated as per expression;

=

∑

(4.33)

Table 4.13 highlights the matrix of a product of and values.

Step – 7: Determine of each alternative.

The for all attributes are counted by expression;

= ∑ (4.34)

The preference of criticalities is arranged in an ascending manner of results of .

The higher should be considered the most pressing matter. Table 4.14 highlights

and rank preferences which may be helpful in recommending remedial measures.


85

TABLE 4.13 Multiplication matrix of and

C D M SP ES EC

Pote

nti

al

Fai

lure

Cau

ses

Pro

bab

ilit

y

of

chan

ce

of

fail

ure

Deg

ree

of

Det

ecta

bil

it

y

Mai

nta

inab

i

lity

Spar

e par

ts

Eco

nom

ic

safe

ty

Eco

nom

ic

cost

C1 0.2500 0.1071 -0.0320 0.2845 0.0621 0.0381

C2 0.2220 0.0804 -0.0643 0.1897 0.0828 0.0381

C3 0.2775 0.0938 -0.1930 0.2845 0.2071 0.1144

C4 0.2497 0.0804 -0.1608 0.2845 0.1450 0.0636

C5 0.2775 0.0670 -0.1930 0.4741 0.1864 0.1272

C6 0.2497 0.0134 -0.0322 0.2845 0.1035 0.0254

C7 0.1942 0.0402 -0.1608 0.2845 0.1450 0.0509

C8 0.2220 0.0670 -0.1608 0.2845 0.1035 0.0636

C9 0.1387 0.0536 -0.0643 0.2845 0.0621 0.0381

C10 0.2497 0.0268 -0.1930 0.3793 0.1450 0.0890

C11 0.0832 0.0804 -0.0965 0.2845 0.0621 0.0381

C12 0.1387 0.0670 -0.1287 0.2845 0.0621 0.0381

C13 0.2220 0.0670 -0.1608 0.2845 0.1243 0.0763

C14 0.2497 0.0268 -0.1930 0.3793 0.1243 0.0890

TABLE 4.14 Maintainability criticality index and rank for PSI

Notation Potential Failure Causes Rank

C1 Improper lubrication & defective sealing 0.7095 3

C2 Higher speed than specified 0.5486 11


per specification 0.7842 2

C4 Foreign matters/particles 0.6623 6

C5 Sudden impact on the rolls 0.9391 1

C6 Loss of power 0.6444 7


issues 0.5539 10

C8 Improper meshing, case depth & high

residual stresses 0.5797 9

C9 Overheating at gear mesh 0.5127 12

C10 Excessive overload & cyclic stresses 0.6968 4


sliding meshing 0.4519 14

C12 Vibratory dynamic load from bearing 0.4618 13

C13 Uneven bearing load 0.6131 8

C14 Reverse and repeated cyclic loading 0.6761 5


86

4.4.2 Significance of PSI

In PSI approach, the predilections of attributes or criteria are determined through the

statistical way. Moreover, the normalized numbers are not converted into weights as per

TOPSIS and COPRAS-G, which make this process more convenient. This method is best

suitable for incompatible situations where conclusions cannot be derived at common

points.

4.4.3 Maintenance Planning through PSI FMECA

Table 4.14 displays the outcomes in form of which are achieved from MCDM

assisted PSI approach as described in Section 4.4.1. The interpretation of these results

gives views about predilections of causes of failures. It conveys that failure cause C5;

sudden impact on the rolls due to bearing misalignment and improper mounting is vital

whereas cause of failure C11; large contact stresses due to rolling and sliding meshing is

minimal in terms of criticality.

This table also shows the failure mode effect analysis with current control practices and

suggested maintenance plan based on their criticalities. It is recommended to update the

current control practices as listed in Table 4.15 that modes of failure C5, C3, C1, C10,

and C14 having large is covered with condition-based monitoring or predictive

type of approaches, modes of failure C4, C6, C13, C8, and C7 having medium is

covered with preventive strategies on principle of prevention is good before collapsible

failures and modes of failure C2, C9, C12, and C11 having low is covered with

remedial actions when prompted.


87

TABLE 4.15 based FMEA with existing practices and proposed improvements in

maintenance plan

Particulars

Current

Controls

Su

gg

este

d I

mpro

vem

ent

in

Mai

nte

nan

ce P

lan

No

tati

on

Key

Process

Input

Potential

Failure Mode

Potential

Causes

Potential

Failure Effects

What is

process

input?

How Process

inputs fail?

Why

important

input failed?

How the

impact

affected due to

failure

(customer or

internal)?

What are the

existing

practices

through

which

failure mode

controlled?

Rolling

Mill

Bearing

Failure

Bearing high

temperature

Improper

lubrication &

defective

sealing

Bearing gets

jammed/Beari

ng housing

jammed

Lubricating

the parts

when

occurred

Predictive

Maintenance C1

Bearing

corrosion

Higher speed

than specified

Increase in

vibration &

noise

Proper

coolant

Corrective

Maintenance C2

Bearing

fatigue

Design

defects,

Bearing

dimension

not as per

specification


replacement

Predictive

Maintenance C3

Roller balls

wear- out

Foreign

matters/partic

les

Sudden rise in

thrust

Regular

cleaning of

parts

Preventive

Maintenance C4

Bearing

misalignment

& improper

mounting

Sudden

impact on the

rolls

Shaft damage

& Impact

damage on

other parts

Routine

check up

Predictive

Maintenance C5

Electrical

damage Power loss

Process

interruption

Electrical

wiring check

up

Preventive

Maintenance C6

Rolling

Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication –

dirt, viscosity

issues

Rough

operation &

considerable

noise

Routine

check-up of

lubrication

Preventive

Maintenance C7

Gear teeth

surface

fatigue

(Pitting)

Improper

meshing, case

depth & high

residual

stresses

Gear life

reduction

Preventive

maintenance

Preventive

Maintenance C8

Gear teeth

scoring

Overheating

at gear mesh

Interference &

backlash

phenomenon

Lubricating

when

needed

Corrective

Maintenance C9

Gear teeth

fracture

Excessive

overload &

cyclic

stresses

Sudden

stoppage of

process plant

Break down

maintenance

Predictive

Maintenance C10


88

Gear teeth

surface

cold/plastic

flow

Large contact

stresses due

to rolling and

sliding

meshing

Slippage &

power loss

Gear replace

when

needed

Corrective

Maintenance C11

Rolling

Mill Shaft

(Primary &

Secondary)

Failure

Shaft fretting

Vibratory

dynamic load

from bearing

Leads to

sudden failure

Break down

maintenance

Corrective

Maintenance C12

Shaft

misalignment

Uneven

bearing load

Vibration &

fatigue

Preventive

maintenance

Preventive

Maintenance C13

Shaft fracture

(Fatigue)

Reverse and

repeated

cyclic loading

Process

stopped

suddenly

Preventive

maintenance

Predictive

Maintenance C14

4.5 Summary

This chapter presents MCDM based failure analysis models with an addition of some

more advanced criteria. It also describes methods to evaluate MCI for each failure mode

of targeted critical components through three different MCDM failure analysis models

called; TOPSIS, COPRAS-G, and PSI for optimizing current maintenance strategies.

The next chapter discusses the out-trend of literature, the results of discrimination

process through shop-floor data for critical components and criticality indices obtained

through traditional and various MCDM based FMECA approaches as discussed in this

chapter. The comparison of results is displayed in form of tables, figures etc. for

effective understanding. Based on achieved RPN and MCI, remedial measures are

suggested and priority plan of existing maintenance activities is discussed in the chapter.

89

CHAPTER 5

Results and Discussion

5.1 General Overview

This chapter describes the results or outcome achieved in previous chapters; i.e. out-

trends of literature review, outcome of the discrimination process for components

through shop-floor data and results obtained through traditional and MCDM based

failure analysis models. The traditional, as well as multi-criteria decision-making based

FMECA approaches are applied to critical components in a view to investigating the

scope of improving reliability by optimizing existing maintenance practices.

The epilogue of such models and their impact on planning the maintenance activities are

discussed. Remedial measures are also discussed based on the observations and outcome

of study. Tables and figures are constructed to show the results and their comparisons in

a most effective manner possible.

The upshot of literature review mentioned the challenges to keep the performance

reliability of major industrial processes by incorporating suitable maintenance strategies.

It highlights the failure analysis as better tool to reach to the root cause of failures.

Furthermore, there were modifications observed in FMECA to enhance the maintenance

plan by some researchers for various processing units. The outcome of the literature

review clears a scope in contemporary application of non-identical MCDM approaches

to the maintenance prioritization problems. The aluminium wire rolling mill appears to


90

be one of the important segments of processing units due to the expansion of

electrification in India. The strong exigency of disproportionate multi-criteria FMCEA

approaches simultaneously at a time to an aluminium wire rolling mill is advocated and

discussed.

5.2 Results of Discrimination Process through Shop-floor Statistics

In chapter 2, the critical components of aluminium wire rolling mill are differentiated by

explicating the historical failure data and real shop-floor practices. The past records

regarding non-performance or failures of parts or components associated with an

aluminium processing machine are gathered as well as disseminated based on downtime,

frequency of failures and loss of production in terms of volume (tons) and cost. The

failure pattern behaviour of these components is understood with modelling of reliability

terms in this study. The identification of critical components is done based on actual

shop-floor condition and historical records. Fig. 5.1 shows the results of failure records

in form of % failure contributions of each part through pie chart for better understanding.

FIGURE 5.1: Pie chart presentation for % failure contributions of components

5.2 Results of Discrimination Process through Shop-floor Statistics

91

The interpretation of data extracted major critical components as;

(i) Bearings

(ii) Gears

(iii) Shafts

The total downtime recorded is 2117 hours during the study period of a year for rolling

machine. Out of which 1476 hours are only due to bearings failures. Moreover, it is

noted that the bearings failed 620 times of total 1072 failure frequencies. It seems that

the bearings are most critical with about 70 % failure contributions. It is observed that

the gear’s overall percentage failure contribution is 4 % having 204 hours downtime of

total 1476 hours. The gears failed 68 times of total 1072 failure frequencies. The failure

contributions of shafting are about 4% with 124 hours downtime of total 1476 hours with

bearings and gears. They failed 41 times of total 1072 failure frequencies.

The above parts are essential and responsible for a proper feed of the aluminium wire for

size reduction at every stage of a stand. These parts are interdependently working with

high speed and experiencing various forces. This phenomenon leads them the most vital

parts of an aluminium rolling machine. Moreover, these critical components (bearings,

gears, and shafts) are usual components to nearly all processing units.

The other remaining parts or components are contributing about 22 % loss due to their

failures with 313 hours downtime of total 1476 hours. They failed together for 343 times.

It is noted that all such parts are not having a significant effect on overall performance of

a rolling machine. Also, it is observed that many such parts are replaced under the

standard replacement of three identified vital parts; bearings, gears and shafts.

Table 5.1 represents outcome of the process of shop-floor data collection during April

2013 to March 2014 regarding the performance of three discriminated vital parts over

other components of a rolling machine.


92

Table 5.1: Outcome of shop-floor data regarding performance of rolling machine parts

Sr.

No. Particulars Downtime (Hours)

Frequency of

failures (n)

1 Bearings (Critical part-1) 1476 620

2 Gears (Critical part-2) 204 68

3 Shafts (Critical part-3) 124 41

4 All remaining parts other than

critical 1, 2, and 3 313 343

Rolling mill as a whole unit

(All parts-Total) 2117 1072

5.3 Results of Traditional FMECA Model

In traditional FMECA, RPN of every failure mode is calculated for specific critical

components by multiplying scores of three basic criteria. The score are modeled for each

criterion for every failure mode is on a scale of 1 to 10 based on actual perspective of the

maintenance personnel and shop-floor conditions.

The classification of various failure modes is done through obtained RPN value in most

critical, critical and normal. The failure modes are scrutinized as most critical for RPN

value 500 or higher and need predictive or condition-based maintenance. The failure

modes are scrutinized as critical for RPN value ranging between 250 and 500 and

proposed preventive type of strategies. Whereas failure modes with RPN value below

250 are termed as normal failures and proposed cured when failed concept of corrective

strategies.

Table 5.2 shows the categorization of different failure causes based on their criticalities

evaluated through RPN. The failure mode bearing fatigue (C3) is the most critical failure

mode under study. The failure modes C1, C4, C5, C8, C10, C13 are found critical and

C2, C6, C7, C9, C11, C12, C14 are found normal.

5.4 Results of MCDM based FMECA

93

TABLE 5.2 Outcome (RPN) of traditional FMECA and suggestions

RPN Value Failure Causes Criticality

Level

Suggested

Maintenance

Plan

More than 500 C3 Most critical Predictive

Between 250 to 500 C13, C10, C4, C5, C8, C1 Critical Preventive

Less than 250 C14, C9, C12, C11, C2, C7, C6 Normal Corrective

The Fig. 5.2 shows the graphical representation of RPN for each failure cause based on

the traditional FMECA approach.

FIGURE 5.2 RPN for each failure cause based on traditional FMECA

5.4 Results of MCDM based FMECA

The for various modes of failure of targeted vital parts are determined based on

three non-identical MCDM failure analysis models called; TOPSIS, COPRAS-G, and

PSI as discussed in Chapter 4. The six criteria; P, D, M, SP, ES and EC as described in

Section 3.3.2 are assigned scores based on actual shop-floor practices in order to find

MCIs. The outcome of the study categorized the failure causes in most critical, critical

and normal failures for a high, moderate and low value of MCIs respectively. The

appropriate maintenance action has been suggested which is discussed in Section 5.5.

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14

RPN 280 72 630 336 320 14 50 320 168 384 75 150 392 224

0

100

200

300

400

500

600

700

RP

N

Traditional FMECA

RPN


94

Table 5.3 shows the categorization of different failure causes based on their criticalities

evaluated through MCDM approaches. Moreover, Fig. 5.3 shows the graphical

representation of MCIs for each failure cause based on different MCDM FMECA

approaches.

TABLE 5.3 Outcome (MCI) of MCDM FMECAs and suggestions

FIGURE 5.3 MCI for each failure cause based on MCDM based FMECA

Particulars Failure Causes

Met

hod

s

TOPSIS C5, C3, C4, C10,

C14

C13, C8, C7, C1,

C12

C2, C11, C9, C6

COPRAS-G C3, C5, C10, C4,

C14

C13, C8, C7, C1,

C2

C12, C6, C11, C9

PSI C5, C3, C1, C10,

C14

C4, C6, C13, C8,

C7

C2, C9, C12, C11

MCI Value Impact High Moderate Low

Criticality Level Most Critical Critical Normal

Suggested Plan Predictive

(Condition-based)

Preventive Corrective

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14

TOPSIS FMECA 0.4265 0.3640 0.7986 0.5794 0.8051 0.2499 0.4419 0.4981 0.2515 0.5636 0.3460 0.3525 0.5505 0.5455

COPRAS FMECA 0.1297 0.1244 0.2156 0.1662 0.2079 0.1062 0.1401 0.1444 0.0863 0.1700 0.1022 0.1096 0.1477 0.1526

PSI FMECA 0.7095 0.5486 0.7842 0.6623 0.9391 0.6444 0.5539 0.5797 0.5127 0.6968 0.4519 0.4618 0.6131 0.6761

0.0000

0.1000

0.2000

0.3000

0.4000

0.5000

0.6000

0.7000

0.8000

0.9000

1.0000

MC

I

MCDM Based FMECA

TOPSIS FMECACOPRAS FMECAPSI FMECA

5.5 Suggested Remedies

95

5.5 Suggested Remedies

During the failure pattern study, it is observed that almost 70 % downtime is due to

bearing failure and replacement practice is 100 %. The bearing is rotational machining

member under dynamic loading. The primary reason for high failure rate is the use of

non-standardize bearings. Moreover, it is noticed poor shop-floor management, lack of

proper documentation and monitoring regarding failure rate, inadequate or traditional

maintenance practice like; mounting and misalignment issues, lubrication issues etc.

Also, the gears and shafting functions are dependent on bearing performance. Such

issues make the bearings vital part of major industrial processing units like; aluminium

rolling mill.

It is suggested to select standardize bearing with appropriate specifications and mount

them properly during every replacement. This will help to avoid bearing misalignment

(C5) and minimizing reverse and repeated cyclic loading thus shaft fatigue (C14) and

gear tooth fracture (C10) can be controlled.

An appropriate condition monitoring methods are suggested to record the condition of

bearing damage and shaft damage which will help to prevent sudden breakdown and

starting thrust on these components. Also, it is suggested checking the condition of

lubricants and replacing them whenever necessary rather than routine clean up. So that,

failure causes such as; sudden impact on the rolls (C5), design defects with bearing

dimension/specification (C3), foreign matters/particles (C4), excessive overload & cyclic

stresses (C10) and reverse & repeated cyclic loading (C14) can be covered and

controlled under recommendations.

Table 5.4 shows the maintenance action plan based on a comparison of results obtained

through traditional as well as MCDM based FMECA (TOPSIS/COPRAS-G/PSI)

approaches respectively. The suggestions in the revision of the current control practices

are derived based on the concurrent effect of these results and same is displayed in Table

5.4. The common modes of failure C5, C3, C4, C10, and C14 having large is

covered with condition-based monitoring or predictive type of approaches, modes of

failure (C13, C7, C8, and C1) having medium is covered with preventive measures

where it is assumed that avoidance of failure is better than restore and modes of failure


96

(C2, C11, C12, C6, and C9) having small is covered by remedial or corrective

actions when breakdown prompts.

5.6 Summary

This chapter discusses the out-trend of literature, the results of discrimination process

through shop-floor data for critical components and criticality indices obtained through

traditional and various MCDM based FMECA approaches as discussed in chapter 3. The

comparison of results is displayed in form of tables, figures etc. for effective

understanding. Based on achieved RPN and MCI, remedial measures are suggested and

priority plan of existing maintenance activities is discussed in the chapter.

The next chapter includes the conclusions of a present research study and the scope of

future work recommendations.

5.6 Summary

97

TABLE 5.4 Maintenance optimization action plan

Key

Pro

cess

Inp

ut

Po

ten

tia

l

Ca

use

s

Cu

rren

t

Co

ntr

ols

No

tati

on

RPN Maintainability

Criticality Index Rank

Su

gg

esti

on

Tra

dit

ion

al

TO

PS

IS

CO

PR

AS

-G

PS

I

Tra

dit

ion

al

TO

PS

IS

CO

PR

AS

-G

PS

I

Bearing

Failure

Improper

lubrication &

defective

sealing

Lubricating

the parts

when

occurred

C1 280 0.4265 0.1297 0.7094 7 9 9 3 Preventive

Maintenance

Higher speed

than specified

Proper

coolant C2 72 0.3640 0.1244 0.5486 12 10 10 11

Corrective

Maintenance

Design defects,

Bearing

dimension, not

as per

specification

Bearing

replacement

C3 630 0.7986 0.2156 0.7842 1 2 1 2 Predictive

Maintenance

Foreign

matters/particles

Regular

cleaning of

parts

C4 336 0.5794 0.1662 0.6622 4 3 4 6 Predictive

Maintenance

Sudden impact

on the rolls

Routine

check up C5 320 0.8051 0.2079 0.9391 5 1 2 1

Predictive

Maintenance

Power loss

Electrical

wiring

check up

C6 14 0.2499 0.1062 0.6444 14 14 12 7 Corrective

Maintenance

Rolling

Mill

Gearing

Failure

Inadequate

lubrication -

dirt, viscosity

issues

Routine

check-up of

lubrication

C7 50 0.4419 0.1401 0.5538 13 8 8 10 Preventive

Maintenance

Improper

meshing, case

depth & high

residual stresses

Preventive

maintenance C8 320 0.4981 0.1444 0.5797 6 7 7 9

Preventive

Maintenance

Overheating at

gear mesh

Lubricating

when

needed

C9 168 0.2515 0.0863 0.5127 9 13 14 12

Corrective

Maintenance

Excessive

overload

& cyclic

stresses

Break down

maintenance C10 384 0.5636 0.1700 0.6967 3 4 3 4

Predictive

Maintenance

Large contact

stresses due to

rolling and

sliding meshing

Gear

replace

when

needed

C11 75 0.3460 0.1022 0.4518 11 12 13 14 Corrective

Maintenance

Rolling

Mill Shaft

(Primary

&

Secondary)

Failure

Vibratory

dynamic load

from bearing

Break down

maintenance C12 150 0.3525 0.1096 0.4617 10 11 11 13

Corrective

Maintenance

Uneven bearing

load

Preventive

maintenance C13 392 0.5505 0.1477 0.6131 2 5 6 8

Preventive

Maintenance


98

Reverse and

repeated cyclic

loading

Preventive

maintenance C14 224 0.5455 0.1526 0.6760 8 6 5 5

Predictive

Maintenance

99

CHAPTER 6

Conclusion and Future Scope

6.1 General Overview

In this thesis, the research study has been carried out to investigate the scope of

reliability improvement by optimizing the maintenance practices through actual failure

analysis with the help of traditional as well as multi-criteria FMECA approaches. In this

study, three distinct MCDM approaches namely; TOPSIS, COPRAS-G, and PSI are

discussed. The essences or criteria employed in this study are; probability of chances of

failure, degree of detectability and degree of severity for traditional approach. Moreover

along with probability of chances of failure and degree of detectability; some advanced

criteria like; maintainability, spare parts, economic cost, economic safety are selected

based on the outcome of shop-floor analysis and reliability modeling for MCDM based

failure analysis models.

The failure analysis models are applied to the identified vital components i.e.; bearings,

gears and machining shafts. The identification of critical components is done and FMEA

is deduced with real shop-floor practices. In order to obtain quantitative results,

criticality analysis is governed by assigning the scores to each failure causes of these

components. To do such task effectively, the plant has been monitored for a period from

April 2013 to March 2014 to record the historical failure data in the specific format as

suggested. Moreover, reliability modelling is done with the help of these data to

understand the behaviour of the failure pattern of the rolling mill.


100

To compensate the disadvantages or limitations of traditional failure analysis models,

MCDM based FMECA approaches are discussed. It is concluded from the outcome of all

failure analysis models that failure causes C3, C5, C10 and C14 have prescribed crucial

and advised intensive technical measures. These results are helpful in prioritizing

maintenance activities of process industry of same or of different kinds in accordance

with failure analysis.

6.2 Major Concluding Remarks

Lack of proper maintenance planning without absolute failure analysis is the main reason

of loss of reliability and poor productivity in process industries. In such condition, it is

necessary to optimize current control and maintenance practices. In this research work,

the novelty is primarily abiding by the demonstration of a case study of aluminium wire

rolling mill plant with contemporary application of three non-identical MCDM based

failure analysis models. The cardinal parts discussed here are usually revealing the large-

scale industrial processes along with metal rolling. Thus, the results and

recommendations are useful in explicating the drawbacks of the extant maintenance plan

and revise the same to comparable processing units.

Following conclusions are drawn from presented research work:

(i) The notable past failure statistics (downtime, frequency of failures, loss of

production in terms of tons and cost) of an aluminium wire rolling mill are

recorded with the help of templates suggested in Table 2.1, 2.2 and 2.3 for April

2013 to March 2014. These data are analysed based on reliability to extract an

explicative components i.e.; bearings, gears and machining shafts.

(ii) The study is concentrating prospective modes of various failures of imperative

parts such as; ball or roller type bearings, gears for shorter distance power

transmission and machining shafts of wire rolling process unit. These machining

parts are usually characterized in almost all manufacturing or processing units.

(iii) In traditional FMECA, usually three basic criteria as mentioned in chapter 3 are

considered. In MCDM based FMECA, the severity is replaced with maintainability

to impart value to the maintenance as a measurable term. Furthermore, the

economic safety, economic cost and spares are enumerated. To quantify the effect

6.3 Recommendations for Future Scope of Work

101

of these criteria statically, each criterion is scored on a scale of 1 to 10 based on

perspective of real shop floor context and its footprint on the diverse failure modes.

(iv) Maintenance planning is proposed in RPN for traditional FMECA and to overcome

the drawback of it, maintainability criticality indices are evaluated by three distinct

multi-criteria decision-making approaches; TOPSIS in crisp value, COPRAS-G in

grey number range and PSI where subjective weight consideration not required for

calculating .

(v) The results are helpful in prioritizing maintenance activities of a process industry of

alike or of dissimilar kinds in accordance with the failure analysis. The outcome of

the research in form of the maintenance prioritization will assist in indicating the

procedure for maintaining other non-performing issues of related industrial process

optimally to improve the plant to certain extent.

(vi) The proposed study is a demanding and collaborative type which assists in

comprehension of parts or components’ design life and connected defects. This

will lead to boost new technologies efficiently and to gain the operational

advantage.

6.3 Recommendations for Future Scope of Work

The research work presented can further be extended as under:

(i) Similar work can be extended to other related or different kinds of process industries

like; chemical or fertilizer unit, clothing or sugar mill, petroleum refineries etc. in a

view to deciding suitable maintenance strategy with other multi-criteria based

approaches such as; qualitative flexible multi-criteria (QUALIFLEX), heuristic tools

etc.

(ii) In this research, six criteria are considered during modelling of scores to each failure

cause for further failure analysis. However, there is a scope to incorporate some

more attributes such as; the skill of person, process environment, atmospheric

degradation etc. based on a need of industry.

(iii) Failure analysis approaches can be developed for various process industries

considering contemporaneous failures among various systems.


102

6.4 Limitations of Proposed Research Work

The proposed research study has some limitations as under:

(i) Failure pattern of various kinds of ball and roller bearings are considered identical

and subsequently, failure analysis models have been proposed.

(ii) Reliability parameters like; future forward time and backward time are not

considered during reliability modelling. This may be helpful specifically when it is

difficult to detect failure time.

(iii) The proposed failure models are not showing failures occurred with design

acceptance for parts having an excessive rate of failure and it is out of the scope of

this study.

(iv) The metallurgical aspect of raw aluminium is important for the quality of final

products. It is not the part of a proposed research study.

6.5 Summary

This chapter includes the conclusions of a present research study and the scope of future

work recommendation

103

References

1. Abbas Toloie Eshlaghy and Mahdi Homayonfar, (2011), “MCDM methodologies

and applications: A literature review from 1999 to 2009”, Research Journal of

International studies, vol. 21, Accessed from:

https://www.researchgate.net/publication/259933820

2. Adhikary D. D., Bose G. K., Bose D., and Mitra S., (2014), “Multi-criteria

FMECA for coal-fired thermal power plants using COPRAS-G,” International

Journal of Quality & Reliability Management, vol. 31, no. 5, pp. 601–614, ISSN:

0265-671X.

3. Albert H.C. Tsang, (1995), "Condition‐based maintenance: tools and decision

making", Journal of Quality in Maintenance Engineering, Vol. 1 Issue:

3, pp.3-17, https://doi.org/10.1108/13552519510096350.

4. Andrawus, J. A., Watson, J. and Kishk, M., (2007), “Wind turbine maintenance

optimization: principles of quantitative maintenance optimization”, Wind

Engineering, vol. 31 (2), pp. 101-110, ISSN: 0309-524X.

5. Ashayeri J., Teelen A., and W. Selenj W., (1996), “A production and

maintenance planning model for process industry”, International Journal for

Production Research, Issue 12, vol. 34, pp. 3311-3326, ISSN: 0020-7543.

6. DuyQuang Nguyen and Miguel Bagajewicz, (2008), “Optimization of preventive

maintenance scheduling in processing plants”, 18th European Symposium on

Computer aided Process Engineering, Elsevier, Volume 25, pp. 319–324.

7. Bevilacqua M., Braglia M., and Gabbrielli R., (2000), “Monte Carlo simulation

approach for a modified FMECA in a power plant,” Quality and Reliability

Engineering International, vol. 16, no. 4, pp. 313–324, ISSN: 1099-1638.

8. Braaksma, A., Klingenberg W., and Veldman J. (2013), “Failure mode and effect

analysis in asset maintenance: a multiple case study in the process industry”,

International Journal of Production Research, Issue 4, vol. 51, pp. 1–17, ISSN:

0020-7543.

9. Braglia M, Frosolini M., Montanari R. (2003), “Fuzzy TOPSIS approach for

failure mode, effect and criticality analysis”, Quality and Reliability Engineering

International, vol. 19, pp. 425-443, ISSN: 1099-1638.

http://www.emeraldinsight.com/doi/abs/10.1108/13552519510096350

http://www.emeraldinsight.com/doi/abs/10.1108/13552519510096350

https://doi.org/10.1108/13552519510096350

http://www.tandfonline.com/action/doSearch?action=runSearch&type=advanced&searchType=journal&result=true&prevSearch=%2Bauthorsfield%3A(ASHAYERI%2C+J)

http://www.tandfonline.com/action/doSearch?action=runSearch&type=advanced&searchType=journal&result=true&prevSearch=%2Bauthorsfield%3A(TEELEN%2C+A)

http://www.tandfonline.com/action/doSearch?action=runSearch&type=advanced&searchType=journal&result=true&prevSearch=%2Bauthorsfield%3A(SELENJ%2C+W)

104

10. Braglia M., (2000), “MAFMA: multi-attribute failure mode analysis,”

International Journal of Quality & Reliability Management, vol. 17, no. 9, pp.

1017–1033, ISSN: 0265-671X.

11. Braglia M., Frosolini M., and Montanari R., (2003), “Fuzzy TOPSIS approach

for failure mode, effects and criticality analysis” Quality and Reliability

Engineering International, vol. 19, no. 5, pp. 425–443, ISSN: 1099-1638.

12. Braglia M., Frosolini M., and Montanari R., (2003), “Fuzzy criticality assessment

model for failure modes and effects analysis,” International Journal of Quality &

Reliability Management, vol. 20, no. 4, pp. 503–524, ISSN: 0265-671X.

13. Chanamool N., T. Naenna, (2016), “Fuzzy FMEA application to improve the

decision-making process in an emergency department,” Applied Soft Computing,

vol. 43, pp. 441–453, ISSN: 1568-4946.

14. Chang C. L., Wei C. C., and Lee Y. H., (1999), “Failure mode and effects

analysis using fuzzy method and grey theory,” Kybernetes, vol. 28, no. 9, pp.

1072–1080.

15. Chen, LeiZhi, Sing Kiong Nguang, Xiao Dong Chen, and Xue Mei Li. (2004),

“Modeling and optimization of fed-batch fermentation processes using dynamic

neural networks and genetic algorithms” Biochemical Engineering Journal vol.

22, Issue 1, pp. 51–61. doi:10.1016/j.bej.2004.07.012, ISSN: 1369-703X.

16. Chen, Tao, Jiawen Li, Ping Jin, and Guobiao Cai., (2013), “Reusable rocket

engine preventive maintenance scheduling using genetic algorithm”, Reliability

Engineering & System Safety, vol. 114, pp. 52–60, ISSN: 0951-8320.

17. Dekker, (1996), “Applications of maintenance optimization models: a review and

analysis”, Reliability Engineering & System Safety vol. 51, no. 3, pp. 229–240,

ISSN: 0951-8320.

18. Deng J. L., (1989), “Introduction to grey system theory”, The Journal of Grey

Theory, vol. 1, no. 1, pp. 1–24.

19. Eti, M.C., S.O.T. Ogaji, and S.D. Probert, (2006), “Development and

implementation of preventive-maintenance practices in Nigerian industries”,

Applied Energy, vol. 83, no. 10, pp. 1163–1179, ISSN: 0306-2619.

20. Feili H. R., Akar N., Lotfizadeh H., Bairampour M., Nasiri S., (2013), “Risk

analysis of geothermal power plants using failure modes and effects analysis

(FMEA) technique”, Journal of energy conservation and management, Vol. 72,

pp.69-76, ISSN: 0196-8904.

105

21. Fragassa C, Ippoliti M., (2016), “Failure mode effects and criticality analysis

(FMECA) as a quality tool to plan improvements in ultrasonic mould cleaning

systems”, International Journal for Quality Research, vol. 10, Issue 4, 2016, pp.

847-870, ISSN: 1800-6450.

22. Fragassa C, Pavlovic A., Massimo S., (2014), “Using a total quality strategy in a

new practical approach for improving the product reliability in automotive

industry”, International Journal for Quality Research, vol. 8(3), pp. 297-310,

ISSN 1800-6450.

23. Fulop J., (2005), “Introduction to decision making methods,” BDEI Workshop,

pp. 1–15.

24. Gargama H., and Chaturvedi S. K., (2011), “Criticality assessment models for

failure mode effects and criticality analysis using fuzzy logic”, IEEE

Transactions on Reliability, vol. 60, no. 1, pp. 102–110, ISSN: 0018-9529.

25. Ghosh Devarun, Roy Sandip, (2009), “Maintenance optimization using

probabilistic cost-benefit analysis”, Journal of Loss Prevention in Process

Industry, Elsevier, vol. 22, pp. 403-407, ISSN: 0950-4230.

26. Gilchrist, W. (1993), “Modeling failure modes and effects analysis'',

International Journal of Quality & Reliability Management, vol. 10, No. 5, pp.

16-23, ISSN: 0265-671X.

27. Godwin Barnabas, Maran M, Nixon G S and I Ambrose, (2012), “Maintenance

cost optimization in process industry”, International Journal of Mechanical

Engineering and Robotics Research, Vol.1, No.3, pp. 81-90, ISSN 2278 – 0149.

28. Hwang C. L., Yoon K., (1981), “Multiple attribute decision making: methods and

applications”, Volume 186 of Lecture Notes in Economics and Mathematical

Systems, Springer, New York, USA.

29. Janssen R., (2001), “On the use of multi-criteria analysis in environmental impact

assessment in The Netherlands”, Journal of multi-criteria decision analysis, vol.

10, Issue 2, pp. 101-109, ISSN: 1099-1360.

30. Jensen (1995), “Stochastic models of reliability and maintenance: an overview”,

Proceedings of the NATO Advanced Study Institute on Current Issues and

Challenges in the Reliability and Maintenance of Complex Systems, Kemer-

Antalya, Turkey, June 12–22, pp. 3–36.

106

31. Knapp Gerald M., and Milind Mahajan, (1998), “Optimization of maintenance

organization and manpower in process industries” Journal of Quality in

Maintenance Engineering, vol. 4, no. 3, pp. 168–183, ISSN: 1355-2511.

32. Konak Abdullah, David W. Coit, and Alice E. Smith, (2006), “Multi-objective

optimization using genetic algorithms: a tutorial”, Reliability Engineering &

System Safety, vol. 91, no. 9, pp. 992–1007, ISSN: 0951-8320.

33. Liao R, Bian J., Yang L., Grzybowski S., (2012), “Cloud model-based failure

mode and effects analysis for prioritization of failures of power transformer in

risk assessment”, International transactions on electrical energy systems, vol. 23,

pp. 1172–1190, ISSN: 2050-7038.

34. Lin Y. H., Lee P. C., and Ting H. I., (2008), “Dynamic multi-attribute decision-

making model with grey number evaluations”, Expert Systems with Applications,

vol. 35, no. 4, pp. 1638–1644, ISSN: 0957-4174.

35. Liu H. C., You J. X., You X. Y., and Shan M. M., (2015), “A novel approach for

failure mode and effects analysis using combination weighting and fuzzy VIKOR

method”, Applied Soft Computing, vol. 28, pp. 579–588, ISSN: 1568-4946.

36. Lynch, P., K. Adendorff, V. S. S. Yadavalli, and O. Adetunji, (2013), “Optimal

spares and preventive maintenance frequencies for constrained industrial

systems”, Computers & Industrial Engineering, vol. 65, no. 3, pp. 378–387,

ISSN: 0360-8352.

37. Macharis C., Springael J., Brucker K, D., Verbeke A., (2004), “PROMETHEE

and AHP: The design of operational synergies in multi criteria analysis:

strengthening PROMETHEE with ideas of AHP”, European Journal of

Operational Research, vol. 153, pp. 307-317, ISSN: 0377-2217.

38. Mahadevan M.L, Poorana Kumar S, and Vinodh R., (2010), “Preventive

Maintenance Optimization of Critical Equipment in Process Plant Using

Heuristic Algorithms”, Proceedings of the 2010 International Conference on

Industrial Engineering and Operations Management, Dhaka, Bangladesh,

January 9 – 10.

39. Maity S. R., Chatterjee P., and Chakraborty S., (2012), “Cutting tool material

selection using grey complex proportional assessment method”, Materials and

Design, vol. 36, pp. 372–378, ISSN: 0264-1275.

107

40. Maniya K. D., Bhatt M. G., (2011), “A selection of material using a novel type

decision making method: preference selection method”, Materials and Design,

Vol. 31, pp. 1785-1789, ISSN: 0264-1275.

41. Marseguerra, Marzio, Enrico Zio, and Luca Podofillini, (2002), “Condition-based

maintenance optimization by means of genetic algorithms and Monte Carlo

simulation”, Reliability Engineering & System Safety, vol. 77, no. 2, pp. 151–165,

ISSN: 0951-8320.

42. Mehdi Vasili, Tang Sai Hong, Napsiah Ismail, Mohammadreza vasili, (2011),

“Maintenance optimization models: a review and analysis”, Proceedings of the

2011 International Conference on Industrial Engineering and Operations

Management, pp. 1131-1138, Kuala Lumpur, Malaysia, January 22-24.

43. Mittal K., Tiwary P. C., Khanduja D, Kaushik P., (2016), “Application of Fuzzy

TOPSIS MADM approach in ranking & underlining the problems of plywood

industry in India”, Journal of Cogent Engineering, vol. 3, Issue 1, ISSN: 2331-

1916.

44. Mobin M., Roshani A., Saeedpoor M., & Mozaffari M. M. (2015), “Integrating

FAHP with COPRAS-G method for supplier selection (case study: An Iranian

manufacturing company)”, Proceedings of the International Annual Conference

of the American Society for Engineering Management, American Society for

Engineering Management (ASEM).

45. Moghaddass R., M. J. Zuo and Jian Qu, (2011), “Reliability and availability

analysis of a repairable -out-of- System with repairmen subject to shut-off rules”,

IEEE Transactions on Reliability, vol. 60, no. 3, pp. 658–666. ISSN: 0018-9526.

46. Mohanta Dusmanta Kumar, Pradip Kumar Sadhu, and R. Chakrabarti, (2007),

“Deterministic and stochastic approach for safety and reliability optimization of

captive power plant maintenance scheduling using GA/SA-based hybrid

techniques: a comparison of results”, Reliability Engineering & System Safety,

vol. 92, no. 2, pp. 187–199, ISSN: 0951-8320.

47. Nastac, L., and A.A. Thatte., (2006), “A heuristic approach for predicting fault

locations in distribution power systems”, Power Symposium, NAPS 2006, North

American, doi:10.1109/NAPS.2006.360136.

48. Parida S., N.R.Kotu, M. M. Prasad, (2000), “Development and implementation of

reliability centered maintenance using vibration analysis: experiences at Rourkela

108

steel plant”, 15th

World Conference on Non-destructive Testing, Rome, Italy, 15-

21 October.

49. Pham, Wang (1996), “Imperfect maintenance”, European Journal of Operational

Research, vol. 94, pp. 425–438, ISSN: 0377-2217.

50. Pintelon L. M. and Gelders L. F., (1992), “Maintenance management decision

making”, European Journal of Operational Research, vol. 58, Issue 3, pp. 301-

317, ISSN: 0377-2217.

51. Rastegari A., Archenti A., Mobin M., (2017), “Condition-based maintenance of

machine tools: Vibration monitoring of spindle units”, IEEE Reliability and

Maintainability Symposium (RAMS), pp. 1-8, Florida (FL), USA.

52. Rathi R., Khanduja D., Sharma S. K., (2016), “A fuzzy-MADM based approach

for prioritizing six sigma projects in the Indian auto sector”, International

Journal of Management Science and Engineering Management, Page 1-8,

http://dx.doi.org/10.1080/17509653.2016.1154486.

53. Sachdeva A., Kumar D., and Kumar P., (2009), “Multi-factor failure mode

criticality analysis using TOPSIS”, Journal of Industrial Engineering,

International, vol. 5, no. 8, pp. 1–9, ISSN: 1735-5702.

54. Sahoo T., Sarkar P. K., and Sarkar A. K., (2014), “Maintenance optimization for

critical equipment in process industry based on FMECA method”, International

Journal of Engineering and Innovative Technology, vol. 3, no. 10, pp. 107–112,

ISSN: 2277-3754.

55. Salabun W., (2013), “The mean error estimation of TOPSIS method using a

fuzzy reference models”, Journal of Theoretical and Applied Computer Science,

vol. 7, Issue 3, pp. 40–50, ISSN: 2299-2634.

56. Santos, Amancio, and Antonio Dourado, (1999), “Global optimization of energy

and production in process industries: a genetic algorithm application”, Control

Engineering Practices, vol. 7, issue 4, pp. 549-554, ISSN: 0967-0661.

57. Sikorska J., (2008), “Identifying failure modes retrospectively using RCM data”,

ICOM Asset Management Conference, Fremantle, May 26-29.

58. Son, Young Kap, (2011), “Reliability prediction of engineering systems with

competing failure modes due to component degradation”, Journal of Mechanical

Science and Technology, vol. 25, no. 7, pp. 1717–1725, ISSN: 1738-494X.

59. Sortrakul, N., H. L. Nachtmann, and C. R. Cassady, (2005), “Genetic algorithms

for integrated preventive maintenance planning and production scheduling for a

https://econpapers.repec.org/article/eeeejores/v_3a58_3ay_3a1992_3ai_3a3_3ap_3a301-317.htm

http://www.tandfonline.com/toc/tmse20/current

http://www.tandfonline.com/toc/tmse20/current

http://dx.doi.org/10.1080/17509653.2016.1154486

109

single machine”, Computers in Industry, vol. 56, no. 2, pp. 161–168, ISSN: 0166-

3615.

60. Vandenbrande W. W., (1998), “How to use FMEA to reduce the size of your

quality toolbox”, Quality progress, vol. 31, Issue 11, pp. 97-100, ISSN: 0033-

524X.

61. Verma Ajit Kumar, A. Srividya, and P. G. Ramesh, (2010), “A systemic

approach to integrated E-maintenance of large engineering plants”, International

Journal of Automation and Computing, vol. 7, no. 2, pp. 173–179, ISSN: 1476-

8186.

62. Van Rjin C., (1987), “A system engineering approach for reliability, availability

and maintainability”, Conference on foundation of computer aided operations,

Salt Lake city, UT, July 5-10.

63. Wiecek M, M., Ehrgott M., Fadel G., Figueira J. R., (2008), “Multiple criteria

decision making for engineering”, Omega, The International Journal of

Management Science, vol. 36, pp. 337-339, ISSN: 0305-0483.

64. Xu K., Tang L. C., Xie Ho S. L., and Zhu M. L., (2002), “Fuzzy assessment of

FMEA for engine systems,” Reliability Engineering & System Safety, vol. 75, no.

1, pp. 17–29, ISSN: 0951-8320.

65. Yang Zhixian, and Guobin Yang, (2012), “Optimization of aircraft maintenance

plan based on genetic algorithm”, Physics Procedia, pp. 580–586,

doi:10.1016/j.phpro.2012.05.107.

66. Zammori F., Gabbrielli R., (2011), “ANN/RPN: a multi criteria evaluation of the

risk priority number”, Journal of quality and reliability engineering.

International, vol. 28, pp. 85–104, ISSN: 1099-1638.

67. Zavadskas E. K., Kaklauskas A., Turskis J., and Tamosaitiene J., (2008),

“Selection of the effective dwelling house walls by applying attributes values

determined at intervals”, Journal of Civil Engineering and Management, vol. 14,

no. 2, pp. 85–93, ISSN: 1392-3730.

68. Zavadskas E. K., Kaklauskas A., Turskis J., and Tamosaitiene J., (2009), “Multi-

attribute decision-making model by applying grey numbers,” Informatica, IOS

press, Amsterdam, vol. 20, no. 2, pp. 305–320.

110

69. Zhang F., (2015), “Failure modes and effects analysis based on fuzzy TOPSIS”,

Proceedings of the IEEE International Conference on Grey System and

Intelligent Services (GSIS), pp. 588–593, Leicester, UK.

Books:

1. Balagurusamy E., (1984), “Reliability Engineering”, Tata McGraw Hill, New

Delhi, ISBN-13: 978-0-07-048339-2.

2. Barlow and Proschan (1965), “Mathematical Theory of Reliability”, Wiley &

Sons, New York, Republication in 1996 by Society for Industrial and Applied

Mathematics, ISBN: 978-0-898713-69-5.

3. Dhillon B. S. (1985), “Quality Control, Reliability, and Engineering Design”,

Marcel Dekker, New York, ISBN: 0-8247-7278-4.

4. Ebeling Charles (2016), “An Introduction to Reliability and Maintainability

Engineering”, 19e reprint, McGraw Hill Education (India) Private Limited,

ISBN-13: 978-0-07-042138-2.

5. Hoang Pham (2003), “Handbook of Reliability Engineering”, Springer, ISBN:

978-1-85233-453-6.

6. Holmberg K., Folkeson A. (1991), “Operation reliability and systematic

maintenance”, Elsevier science publishers’ ltd., London, ISBN: 1-85166-612-5.

7. Kececioglu D. (2002), “Reliability Engineering Handbook”, vol. 1, DEStech

publication, ISBN: 1-932078-00-2.

8. Khanna O.P., (2010), “Industrial Engineering and Management”, Dhanpat rai &

sons, ISBN-13: 978-818992835.

9. Kumar Sanjeev, (2009), “Performance analysis and optimization of some

operating systems of a fertilizer plant”, NIT, Kurukshetra, Ph. D. thesis, 2009.

10. Mcdermott R. R., Mikulak R. J. (2009), Beauregard M. R., “Basics of FMEA”,

2nd

Edition, CRC press (Taylor and Francis Group), New york, ISBN: 13: 978-1-

56327-377-3.

11. Mishra R C & Pathak K (2012), “Maintenance Engineering and Management”,

Prentice Hall of India Pvt. Ltd., New Delhi, 2012, ISBN: 978-81-203-4573-7.

12. Moubray J. (1997), “Reliability-centered Maintenance”, 2nd edition, Industrial

Press Inc., ISBN: 0-8311-3078-4.

13. Omdahl T. P. (1988), “Reliability, Availability and Maintainability Dictionary”,

Milwaukee, ASQC Quality Press, ISBN: 0-873-890450.

111

14. O’Connor P. (2002), “Practical Reliability Engineering, John Wiley, England,

ISBN 13: 978-0-470-84462-5.

15. Roy B. (1996), “Multicriteria Methodology for Decision Aiding”, Kluwer

Academic Publishers, ISBN: 978-1-4757-2500-1 (e-book).

16. “Training and Operating Manual” – Sampat Aluminium Rolling Mill,

Ahmedabad, Gujarat

17. Vincke P. (1992), “Multicriteria Decision-Aid”, John Wiley & Sons, Chichester,

ISBN: 978-0-4719-3184-3.

Websites Reports:

1. Farley T., Miller D., “Maintaining Rolling Mill Performance Part I, II, III”,

Innoval Technology Ltd., accessed via: http://www.innovaltec.com/aluminium-

rolling-models-blog/

2. “Indian Electrical & Electronics Manufacturers’ Association”, 68TH ANNUAL

REPORT 2014 – 2015 (September 2015), accessed via:

http://ieema.org/wpcontent/uploads/2015/09/IEEMA-Annual-Report_2014-

15.pdf

3. “Indian Electrical Equipment Industry Mission Plan 2012-2022”, accessed via:

http://www.cea.nic.in/reports/monthly/installedcapacity/2016/installed_capacity-

03.pdf.

4. Michael R. Lyu, Software Reliability Engineering – A Roadmap, accessed from:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.109.8051&rep=rep1&t

ype=pdf

5. “Review of Blueprint for Infrastructure in Gujarat (BIG 2020) – Final Report

Volume 1 – Summary and Vision” (August 2009), accessed via

http://www.gidb.org/Document/2014-12-31_112.pdf

6. Yoe (2002), “Trade-Off Analysis Planning and Procedures Guidebook”,

Prepared for Institute for Water Resources, U.S. Army Corps of Engineers,

accessed from: http://www.iwr.usace.army.mil/Portals/70/docs/iwrreports/02-R-

2.pdf.

http://www.innovaltec.com/aluminium-rolling-models-blog/

http://www.innovaltec.com/aluminium-rolling-models-blog/

http://ieema.org/wpcontent/uploads/2015/09/IEEMA-Annual-Report_2014-15.pdf

http://ieema.org/wpcontent/uploads/2015/09/IEEMA-Annual-Report_2014-15.pdf

http://www.cea.nic.in/reports/monthly/installedcapacity/2016/installed_capacity-03.pdf

http://www.cea.nic.in/reports/monthly/installedcapacity/2016/installed_capacity-03.pdf

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.109.8051&rep=rep1&type=pdf

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.109.8051&rep=rep1&type=pdf

http://www.gidb.org/Document/2014-12-31_112.pdf

112

List of Publications

National/International Journals:

1. Pancholi Nilesh, Bhatt M. G., (2016), “Multi criteria FMECA based decision-

making for aluminium wire process rolling mill through COPRAS-G”, Journal of

Quality and Reliability Engineering, Volume 2016, Article ID 8421916, 8 pages,

ISSN: 2314-8055, http://dx.doi.org/10.1155/2016/8421916. (Scopus Cite Score

2016:0.53, SCImago Rank: 0.22, H Index: 4)

2. Pancholi Nilesh, Bhatt M. G., (2016), “Performance reliability improvement by

optimizing maintenance practices through failure analysis in process industry – a

comprehensive literature review”, IOSR Journal of Mechanical and Civil

Engineering (IOSR-JMCE), vol. 13, Issue 6, pp. 66-73, e-ISSN: 2278-1684, p-

ISSN: 2320-334X.

3. Pancholi Nilesh, Bhatt M. G., (2016), “Traditional and multi-factor decision

making based FMECA through preference selection index method for continuous

process industry”, International Journal of Darshan Institute on Engineering

Research & Emerging Technologies (IJDI-ERET), vol. 5, No. 2, ISSN: (Print):

2320-7590.

4. Pancholi Nilesh, Bhatt M. G., (2017), “Identifying Critical Components of

Identified Process Industry through Shop-floor Failure Data”, International

Journal of Engineering Technology, Management and Applied Sciences, vol. 5,

Issue 2, ISSN: 2349-4476.

5. Pancholi Nilesh, Bhatt M. G., (2017), “TOPSIS and COPRAS-G based

maintenance optimization of aluminium wire rolling mill components”, Journal

of Basic and Applied Research International, International Knowledge Press, vol.

20, Issue 3, pp.189-201, ISSN: 2395-3438 (P), ISSN: 2395-3446 (O).

6. Pancholi Nilesh, Bhatt M. G., (2017), “Traditional and TOPSIS based failure

mode effect and criticality analysis for maintenance planning of aluminium wire

rolling mill components”, GIT – Journal of Engineering and Technology, vol. 10,

ISSN: 2249 – 6157.

7. Pancholi Nilesh, Bhatt M. G., (2017), “Quality enhancement in maintenance

planning through non-identical FMECA approaches”, International Journal for

http://dx.doi.org/10.1155/2016/8421916

113

Quality Research, vol. 11(3), pp. 603-626, ISSN: 1800-6450. (Scopus Cite

Score 2016:0.78, SCImago Rank: 0.234, H Index: 7)

8. Pancholi Nilesh, Bhatt M. G., (In print), “FMECA based maintenance planning

through COPRAS-G and PSI”, Journal of Quality in Maintenance Engineering,

vol. 24, Issue 2, pp. 224-243, Emerald, ISSN: 1355-2511. (Scopus Cite Score

2016:1.16, SCImago Rank: 0.340, H Index: 41)

National/International Conference:

1. Pancholi Nilesh, Bhatt M. G., (2015), “Comparative study of traditional failure

mode effect and criticality analysis (FMECA) and TOPSIS based FMECA for

bearings of aluminium rolling mill plant – a case study”, 2nd

National Conference

on Emerging trends in Engineering, Technology & Management (NCEETM), IU,

Ahmedabad, ISBN: 978-93-80867-75-5.

2. Pancholi Nilesh, Bhatt M. G., (2017), “Identifying critical components of

identified process industry through shop-floor failure data”, International

Conference on Latest Concepts in Science, Technology and Management

(ICLCSTM-2017) at National Institute of Technical Teachers Training &

Research (NITTTR), MHRD, Govt of India, Chandigarh, ISBN: 978-81-932712-

4-7.

3. Pancholi Nilesh, Bhatt M. G., (2017), “Maintenance planning through FMECA

based multi-criteria decision-making PSI approach for aluminium wire rolling

mill plant”, IEEE 2nd

International Conference for Convergence of Technologies

(I2CT), ISBN: 978-1-5090-4307-1/17.

INVESTIGATION ON PERFORMANCE RELIABILITY …€¦ · v Course-work Completion Certificate This is to certify that Mr. Pancholi Nilesh Hasamukhlal, enrolment no. 129990919010 is a

Documents