1

SYNOPSIS

of the Ph. D. thesis entitle

Investigation on Performance Reliability Improvement

by Optimizing Maintenance Practices through Failure

Analysis in Continuous Process Industry

Submitted in

Partial fulfillment of the degree of

DOCTOR OF PHILOSOPHY

of the

GUJARAT TECHNOLOGICAL UNIVERSITY, AHMEDABAD

by

Pancholi Nilesh Hasamukhlal

(Enrolment No:-129990919010)

Supervisor

Dr. M. G. Bhatt

GUJARAT TECHNOLOGICAL UNIVERSITY

CHANDKHEDA, AHMEDABAD

November 2017

2

1. Title of the thesis and abstract

1.1 Title

Investigation on Performance Reliability Improvement by Optimizing Maintenance Practices

through Failure Analysis in Continuous Process Industry

1.2 Abstract

The proposed research work addresses the reliability and maintenance issues of major process

industries. The work deals with the comprehension of failure pattern; reliability modeling and

discrimination of the critical components through substantial shop-floor failure data. These

data are collected for the period of April-2013 to March-2014 for aluminium wire rolling mill

plant situated at Ahmedabad, India. The research work narrates a method for evaluating risk

priority number (RPN) and maintainability criticality index (MCI) through traditional as well

as multi-criteria decision making (MCDM) approaches respectively for each failure cause of

identified critical components. There are three non-identical MCDM approaches discussed

namely; technique for order preference by similarity to ideal solution (TOPSIS), grey-

complex proportional risk assessment (COPRAS-G) and preference section index (PSI).

The primary findings of this research work are to prioritize the maintenance activities by

comparing results obtained through different failure analysis models. It is proposing

improvements in the maintenance plan of critical components like; bearings, gears, and shafts

of aluminium wire rolling mill which are commonly representing the most critical

components in a large range of industrial processes.

Originality mainly consists in the contemporary application of non-identical MCDM methods

(TOPSIS, COPRAS-G and PSI). It will help to elucidate maintenance issues of major process

industries and recommended deliverable keys where multi-criteria decision-making (MCDM)

approaches are very useful.

2. Brief description on the state of the art of the research topic

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

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

planning. Barlow and Proshan (1965, 1975) researched some practices in maintenance

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

3

classification of maintenance models. Sikorska (2008) presented the scope to improve the

quality of failure histories stored in computerized maintenance management systems.

Looking to the literature review; it seems many researchers have done various modifications

for improvement of FMECA to overcome drawbacks for different processing plants. Hwang

and Yoon (1981) highlighted the importance of MCDM, where multiple and conflicting

criteria are under consideration in different area like personal, public, academic or business

contents; Gilchrist (1993) incorporated failure cost to his modified FMECA model;

Bevilacqua et al. (2000) presented modified FMECA through Monte Carlo simulation in

power plant; Braglia (2000) developed multi-attribute failure mode analysis with economic

considerations; Xu et al. (2002) presented FMEA of engine system based on fussy assessment

concept; Braglia et al. (2003) presented fuzzy TOPSIS FMECA to overcome the limitations

of conventional US MIL-STD-1626A method; Sahoo et al. (2004) showed that FMECA is a

basic part of the maintenance plan and a strong tool to evaluate and improve system

reliability with reduction of overall maintenance cost; Sachdeva et al. (2009) presented a

multi-criteria decision-making approach to prioritizing failure modes for paper industry using

TOPSIS; Gargama and Chaturvedi (2011) presented risk factors in fuzzy linguistic variables

to generate fuzzy rank priority number; Maniya and Bhatt (2011) presented the multi-criteria

decision-making method to solve problems of facility layout design selection based on

preference selection index (PSI) method; Adhikary and Bose (2014) presented multi-factor

FMECA through COPRAS-G method for coal-fired thermal power; Fragassa et al. (2014)

presented an advanced application of FMECA used in integration with other quality tools

(FTA, RDA) for recognizing critical functions on diesel intake manifold in a view to

optimizing industrial processes where several parts are realized in aluminium (including

wires). Mobin et al. (2015) proposed an integration of a fuzzy analytic hierarchy process

(FAHP) and the complex proportional assessment of alternatives to grey relations (CORPAS-

G) to prioritize suppliers in an Iranian manufacturing industry; Zhang (2015) deduced

closeness coefficient for failure modes by integrating both subjective and objective weights to

avoid over or under estimation though fuzzy TOPSIS; Chanamool and Naenna (2016)

highlighted the importance of Fuzzy FMEA to prioritize and assess failures associated with

working process of hospital’s emergency department. Mittal et al. (2016) described the

ranking of major problems of plywood industries through multiple-attribute decision-making

(MADM) approach based fuzzy TOPSIS. Rathi et al. (2016) presented fuzzy MADM for

prioritizing six sigma projects through fuzzy VIKOR in the Indian auto sector. Rastegari et al.

4

(2017) addressed condition-based maintenance and its implementation with vibration

monitoring techniques in order to plan maintenance activities of the spindle units of the

automobile gear box manufacturing company in Sweden. Farley and Miller of Innoval

technology Ltd. presented three parts on maintaining rolling mill performance. In first part

they identified some factors responsible for unhealthy rolling mill performance over time. In

second part they discussed the overall equipment efficiency (OEE) based approaches to avoid

such decline in order to improve rolling mill performance and in third part they explained

guidelines to avoid initial decrease in performance of new mill through good design, training

and technical support.

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

quality research is lacking in the aluminium wire rolling mill which forms an important

segment of the process industry. Literature review shows that past researchers have not yet

considered the case of three multi-criteria decision-making approaches simultaneously

applied to any process industry. There is a huge scope for improvement in reliability by

optimizing maintenance practices through failure analysis based on different MCDM

approaches in aluminium wire rolling mill plant.

3. Definition of the Problem

After globalization, various process industries all over the world face the problem of keeping

the subsystem and components in efficient working condition for carrying out its designated

functions effectively for a sufficient possible longer time. Minimum downtime of

components is the pressing needs of major process industries such as the rolling mill, dairy

plant, chemical-petrochemical plant, sugar mill, textile mill, paper industry, fertilizer plant

etc. Reliability oriented maintenance is relatively a new tool for mechanical engineering in

India to addresses reliability issues in process industries. It analyzes the system and sub-

system of plant and tries to find out the failure modes, effects and consequences of the

failure. Also, the study can be at preventing or reducing such failures. The growing need for

higher reliability arises from the requirement to develop the maintenance plan which

continuously performs in the most efficient manner possible.

Out of various process industries as stated earlier, the aluminium wire rolling mill is selected

for study because aluminium transmission wire market size was about 6.5 lacs Metric Tons in

volume terms in financial year (FY) – 2015 in India (IEEMA’s 68th

Annual Report – 2014-

15). It is likely to grow at a compound annual growth rate (CAGR) of 13.5% between FY14-

19 due to inter-regional transmission network expansion, infrastructure, industrial demand

5

and Government of India’s “power for all” initiative (Indian Electrical Equipment Industry

Mission Plan 2012 – 22). In Gujarat also transmission lines network is likely to grow at a

CAGR of 7.8% between FY14-18 and Government of Gujarat will be investing US$4.6

billion in transmission and distribution by 2020 (BIG 2020 – Final Report Volume 1 –

Summary and Vision, August 2009).

During the detailed study of performance of identified aluminium wire rolling mill, it is

observed that approximately 20 to 25 % of possible production time goes towards

maintenance of equipment i.e. loss of reliability. After studying the facts related to aluminium

wire rolling mill plant, it is found that maintenance is the main cause of low productivity and

profit. Hence, it is required to upgrade current control practices associated with maintenance

system with a view to increasing the effective utilization of resources with little cost or

without any additional cost.

Literature reviews state that FMECA is widely used and accepted tool to enhance

maintenance practices in process industries. It is based on the systematic brainstorming

session to recognize the failures which may occur in system or process (Vandenbrande,

1998). It is devoted to determining the design reliability by considering the potential causes

of failures and their effects on the system under study (Dhillon, 1985 and O’Conner, 2002).

In present research study, the problem is defined for the scope of investigating the extent at

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

control and maintenance practices of aluminium rolling mill through three distinct MCDM

based FMECA approaches with traditional approach:

(i) Technique for order preference by similarity to ideal solution (TOPSIS) where; scores

are considered in weighted crisp value

(ii) Grey-complex proportional assessment (COPRAS-G) where; weighted scores are in

grey number range rather than in crisp value and;

(iii)Preference selection index (PSI) where; subjective weight consideration not required.

4. Objective and Scope of work

The main aim of this study is to enhance existing maintenance practices with modified

FMECA through MCDM approaches of identified aluminium wire rolling mill and to derive

the scope of improvement in maintenance strategy to process industries of similar or different

kinds in accordance with failure analysis.

The research objectives explored in this study are as under:

6

(i) To study reliability and maintenance issues faced by process industry deriving the

scope of maintenance optimization.

(ii) Collection and analysis of historical failure data including evaluation of major

reliability parameters for components of identified aluminium wire rolling mill.

(iii)Identification of the critical components based on their downtime, frequency of failure

and loss of production in terms of volume and cost.

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

present maintenance practices of identified critical components and assignment of

scores to each failure causes based on real shop-floor condition for further failure

analysis models.

(v) Optimizing maintenance activities of critical components through traditional as well

as multi-criteria decision-making (MCDM) approaches to prove its competency.

Scope of Work

The scope of proposed research work is summarized as below:

(i) Understanding the scenario of the performance of different process industries 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. (The data set can be used for comparison and

investigation of future failures with the highest probability of occurrence). Failure

data analysis and estimation of reliability parameters to generate necessary inputs.

(iii)Discrimination of critical components like; bearings, gears, and shafts of aluminium

wire rolling mill based on downtime, frequency of failures, loss of production in

terms of volume and cost. Understanding the failure modes, causes, effects and

consequences of failure pattern and problems faced in present maintenance practices.

(iv) Multi-criteria decision-making based different failure models for prioritizing

maintenance activities.

(v) Comparison of results of different MCDM models. Suggested improvement and

recommendations for future scope of the study.

(vi) It is assumed that presented failure model may not represent failures due to the first

point as adequate of design for such components are out of the scope of this study for

high failure rate.

7

5. Original contribution by the thesis

In present research study, three distinct MCDM based models are proposed over traditional

FMECA to provide better alternative to maintenance practitioner for planning maintenance

activities accordingly. The research work will make following original contributions:

(i) The actual historical failure data like; run time, uptime, frequency of failures, average

repair or replace time are collected for the period of April-2013 to March-2014 for

identified aluminium wire rolling mill.

(ii) Reliability parameters like; mean time between failure (MTBF), mean time to repair

(MTTR), mean down time (MDT), hazard rate, availability are calculated based on

actual historical failure data to generate necessary inputs for further failure analysis.

(iii)The failure modes, causes, effects and consequences of failure pattern and problems

faced in present maintenance practices, as well as assignment of scores to failure

causes is based on actual shop-floor condition to find RPN and MCI through

traditional and MCDM approaches respectively.

(iv) The suggestions are discussed by comparing the results of three distinct MCDM

approaches. So, the work will present a strong case in developing the maintenance

plan to process industries of same or of a different kind.

6. Methodology of Research, Results / Comparisons

6.1 Overview, Failure Data Collection and Reliability Modeling

The aluminium wire rolling mill near Ahmedabad was selected for the research study. The

soft solid aluminium bar of 40 mm diameter is converted into 6 mm diameter wire through

series rolling process by fifteen stands with decreasing the diameter of wire by about 15-20 %

at each stand. In this study, the work is focused on fifteen stands, where reliability and

maintenance issues played a crucial role. Each stand has thirty-one components. The

historical failure data (Run time, Frequency of failures, Repair/Replace time) are collected

and recorded for all thirty-one components of all fifteen stands for the duration of April 2013

to March 2014 at Sampat Aluminium Pvt. Ltd., Ahmedabad, India. Then, the major reliability

parameters (MTBF, MTTR, MDT, Hazard rate, Availability etc.) are calculated from failure

data with the help of mathematical equations as follows (Balagurusamy 1984, Hoang Pham

2003, Khanna 2010, Mishra & Pathak 2012):

(1)

8

=

(2)

(3)

(4)

(

) (5)

(6)

(7)

Where; is Mean Time Between Failure, is frequency of failure, is hazard rate,

is Uptime, is Mean Down Time, is Down time, is Mean Time to Repair,

is Mean Time Between Maintenance, is Total time, is Operational Availability,

is Inherent Availability

6.2 Identification of Critical Components of Rolling Mill

The substantial reliability failure data are analyzed in a view to identifying the most critical

components based on their failure rate. The component criticality is decided based on the

downtime, frequency of failure, loss of production in terms of volume and cost for all

components of each stand. The critical components like; bearings, gears, and shafts are

identified which are common to major process industries. The failure modes, causes, effects

and consequences of failure pattern and problems faced in present maintenance practices are

derived based on actual shop floor condition.

6.3 Assignment of Score to Failure Causes

The score for each individual failure mode for every process input of critical components are

decided on; (i) historical failure data; which gives the comprehensive behavioral study of

failure pattern of critical components and (ii) questionnaires; to floor operators, managers and

maintenance personnel. The scores for each failure cause for every different criterion are

ranked on a scale of 1 – 10. The scale of 1 to 10 refers from least to most consideration of the

impact of criteria. The scores are assigned to three criteria; chances of failure (C), degree of

detectability (D) and severity of effect (S) for traditional FMECA. The scores are assigned to

9

six criteria; chances of failure (C), degree of detectability (D), degree of maintainability (M),

spare parts (SP), economic safety (ES) and economic cost (EC) for MCDM based FMECA

approaches. It is in crisp values for TOPSIS and PSI and in grey number range ] for

COPRAS-G. Table I and Table II displays the scores assigned as discussed above for

traditional and MCDM based FMECA respectively.

6.3 Failure Analysis Models

In FMECA, Risk Priority Number (RPN) is calculated by multiplying the scores of criteria;

chances of failure (C), degree of detectability (D) and severity of effect (S). The limitations

of FMECA are that it covers only limited criteria. Moreover, same importance is given to

each criterion without considering their relative importance. Also, a small variation of C or D

or S may change the value of RPN due to multiplication. To overcome the limitations of

FMECA, multi-criteria decision-making (MCDM) approaches are used with covering more

criteria. It compares alternatives relatively on weights which help decision-making process

effective.

TOPSIS is a multi-attribute decision-making system based on the measurement of the

Euclidean distance of each criterion from the ideal value which was first discussed in a crisp

version by Hwang and Yoon (1981). In this study, maintainability criticality index

( ) is calculated based on procedure as discussed by Sachdeva et al. (2009) for each

failure cause of critical components. COPRAS-G stands for Grey-Complex Proportional Risk

Assessment which works on grey number concept. The grey number is having upper and/or

lower limits whose value falls within an interval (Zavadskas et al. 2008, 2009 and Maity et al.

2012). This concept of grey was deduced from grey theory, which helps in dealing

uncertainty of real environment (Deng 1989, Chang et al. 1999 and Lin et al. 2008). In this

study, maintainability criticality index ( ) is calculated based on the procedure as

discussed by Zavadskas et al. (2008, 2009) and Maity et al. (2012). PSI stands for Preference

Selection Index. In PSI, maintainability criticality index ( ) is calculated through

method proposed by Maniya and Bhatt (2011).

Significance of COPRAS-G

During a brainstorming session, maintenance personnel score a criticality factor into different

criticality levels so it is challenging to do criticality analysis of failure modes accurately.

Hence this practical difficulty can be solving by expressing the scores of a criticality factor in

10

an interval (grey number) instead of certain and the exact value of TOPSIS. The main idea of

COPRAS-G method is to express the criteria values in intervals.

Significance of PSI

In preference selection index (PSI) method, preference values of each attribute are calculated

using the concept of statistics rather than assignment of weight attributes as in TOPSIS and

COPRAS-G. This method is very helpful in deciding the relative importance between

attributes when the situation of the conflict occurred.

TABLE I

MAINTENANCE PLANNING THROUGH TRADITONAL FMECA APPROACH

Particulars

Current Controls

Standard FMECA

Key

Process Input

Potential

Failure Mode

Potential

Causes

Potential Failure

Effects

C

D

S

Ris

k P

rio

rity

Nu

mb

er

Ran

k

What is the

Process

Input?

In what ways

can the

Process Input fail?

What causes the

Key Input to go

wrong?

What is the

impact on the

Key Output Variables once

it fails (customer or

internal

requirements)?

What are

the existing

controls and procedures

that prevent either the

Cause or the

Failure Mode?

Ho

w w

ell

can

you

det

ect

the

Cau

se o

r th

e F

ailu

re

Mod

e?

Ho

w o

ften

does

cau

se o

r F

M o

ccu

r?

Ho

w S

ever

e is

th

e ef

fect

to

th

e cu

stom

er?

Rolling

Mill Bearing

Failure

Bearing high

temperature

Improper

lubrication &

defective sealing

Bearing gets jammed/Bearing

housing jammed

Lubricating

the parts

when occurred

5 8 7 280 7

Bearing

corrosion

Higher speed

than specified

Increase in vibration &

noise

Proper

coolant 3 6 4 72 12

Bearing

fatigue

Design defects,

Bearing

dimension, not

as per

specification

Life reduction Bearing

replacement 9 7 10 630 1

Roller balls wear- out

Foreign matters/particles

Sudden rise in thrust

Regular

cleaning of

parts

8 6 7 336 4

Bearing

misalignment & improper

mounting

Sudden impact on the rolls

Shaft damage &

Impact damage

on other parts

Routine check up

8 5 8 320 5

Electrical damage

Loss of power Operation interrupted

Electrical

wiring

check up

2 1 7 14 14

11

Rolling

Mill Gearing

Failure

Gear teeth wear-out

Inadequate

lubrication - Dirt, viscosity

issues

Rough

operation & considerable

noise

Routine

check-up of

lubrication

5 2 5 50 13

Gear teeth

surface fatigue

(Pitting)

Improper

meshing, case depth & high

residual stresses

Gear life reduction

Preventive maintenance

8 5 8 320 6

Gear teeth scoring

Overheating at gear mesh

Interference &

backlash

phenomenon

Lubricating

when

needed

7 6 4 168 9

Gear teeth fracture

Excessive

overload &

cyclic stresses

Sudden

stoppage of

process plant

Break down maintenance

8 6 8 384 3

Gear teeth

surface

cold/plastic

flow

High contact stresses due to

rolling &

sliding action of mesh

Slippage &

power loss

Gear

replace

when

needed

5 3 5 75 11

Rolling Mill Shaft

(Primary

& Secondary)

Failure

Shaft fretting

Vibratory

dynamic load

from bearing

Leads to sudden failure

Break down maintenance

5 6 5 150 10

Shaft

misalignment

Uneven bearing

load

Vibration &

fatigue

Preventive

maintenance 7 7 8 392 2

Shaft fracture

(Fatigue)

Reverse & repeated cyclic

loading

Sudden stoppage of

process

Preventive

maintenance 7 4 8 224 8

7. Results, Discussion and Suggestions

In this study, the historical failure data of aluminium wire rolling mill are collected and

analyzed which help to understand behavioral failure pattern of components of rolling mill.

Moreover, major reliability parameters are calculated as a part of reliability analysis. The

results of criticality analysis show that bearings (70 %), gears (4 %) and shafts – primary &

secondary (4 %) are most critical components, which needs detailed FMECA to enhance the

working condition of the overall process industry. Remaining components cover of about 22

% of contribution for reliability loss on failure basis.

The results of traditional FMECA; failure modes with RPN more than 500 are considered

most critical and required to perform predictive maintenance, RPN from 250 to 500 are

considered critical and recommended preventive maintenance and less than 250 are

considered normal failures which are suggested corrective maintenance.

During failure pattern study, it is observed that almost 70 % downtime is due to bearing

failure and replacement practice is 100 %, so it is suggested to select standardize bearing with

appropriate specifications and mount them properly during every replacement to avoid

12

bearing misalignment (C5) and minimizing reverse and repeated cyclic loading thus shaft

fatigue (C14) and gear tooth fracture (C10) can be avoided. It is suggested appropriate

condition monitoring to continuously record the condition of bearing damage and shaft

damage 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. Hence, 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 under

recommendations.

Table III shows the comparison of results for traditional FMECA as well as MCDM based

FMECA (TOPSIS/COPRAS-G/PSI) approaches. It is suggested to modify the current control

practices as listed in Table I and Table II that failure causes (C5, C3, C4, C10, C14) with at

least large value of should be kept under predictive maintenance, failure cause (C13, C7,

C8, C1) with moderate value of should be kept under preventive maintenance and

failure causes (C2, C11, C12, C6, C9) with low should be kept under corrective

maintenance.

7. Achievements with respect to objectives

In present research study, three different MCDM based FMECA models are discussed for

optimizing existing maintenance practices to accomplish the objectives.

Objective (i):

This objective is achieved by visiting many process industries like; dairy plant (Sagar-Amul),

petrochemical plant (Nova), paper industry (Utility print pack), textile mill (Welspun), rolling

mill (Deora group) etc. to study reliability and maintenance issues faced by them. It is found

that average cost of maintenance or loss of reliability is about 8 to 25 % with existing

maintenance practices which are either breakdown or planned shutdown to various process

industries. It is derived the scope of improvement in existing maintenance practices.

Objective (ii):

This objective is achieved by identifying aluminium wire rolling mill situated near

Ahmedabad for further study. It is observed certain issues which affect reliability and needs

attention to enhancing maintenance plan during preliminary study. It is suggested certain

formats to gather the failure data and maintenance records. The failure data are collected and

analyzed for a period of a year which leads to reliability analysis and modeling.

13

TABLE II MAINTENANCE PLANNING THROUGH TOPSIS, COPRAS-G & PSI FMECA APPROACHES

Particulars DECISION MATRIX

(TOPSIS/PSI ) DECISION MATRIX (COPRAS-G)

TOPSIS COPRAS-G PSI

No

tati

on

Key

Process

Input

Potential

Failure

Mode

Potential

Causes

Potential

Failure

Effects

Current

Controls C D M S

P

E

S

E

C C D M SP ES EC

What is

the

Process

Input?

In what ways

can the

Process Input

fail?

What causes

the Key

Input to go

wrong?

What is the

impact on

the Key

Output

Variables once it fails

(customer

or internal requirement

s)?

What are the

existing

controls and

procedures

that prevent either the

Cause or the

Failure Mode?

Chan

ce o

f fa

ilu

re

Det

ecti

on

pro

bab

ilit

y o

f

fail

ure

Mai

nta

inab

ilit

y c

rite

ria

Sp

are

par

ts c

rite

ria

Eco

no

mic

saf

ety

cri

teri

a

Eco

no

mic

co

st c

rite

ria

Chan

ce o

f fa

ilu

re

Det

ecti

on

pro

bab

ilit

y o

f fa

ilu

re

Mai

nta

inab

ilit

y c

rite

ria

Sp

are

par

ts c

rite

ria

Eco

no

mic

saf

ety

cri

teri

a

Eco

no

mic

co

st c

rite

ria

MC

I

Crit

icali

ty R

an

k

MC

I

Crit

icali

ty R

an

k

MC

I

Crit

icali

ty R

an

k

xij xij xij xij xij xij

Xi

j xij xij yij

Xi

j yij xij yij xij yij xij yij

Bearing

Failure

Bearing high

temperature

Improper lubrication &

defective

sealing

Bearing

gets jammed/Be

aring

housing jammed

Lubricating the parts

when

occurred

9 8 1 3 3 3 8 9 7 8 1 2 2 3 3 4 3 4 0.4265 9 0.1297 9 0.7094 3 C1

Bearing corrosion

Higher speed

than

specified

Increase in

vibration

& noise

Proper coolant

8 6 2 2 4 3 8 9 5 6 1 2 2 3 3 4 4 5 0.3640 10 0.1244 10 0.5486 11 C2

Bearing

fatigue

Design

defects, Bearing

dimension, not as per

specification

Life

reduction

Bearing

replacement

1

0 7 6 3

1

0 9 9

1

0 7 8 6 8 3 5 9

1

0 9

1

0 0.7986 2 0.2156 1 0.7842 2

C3

Roller balls

wear- out

Foreign

matters/parti

cles

Sudden rise

in thrust

Regular

cleaning of

parts

9 6 5 3 7 5 7 9 6 7 4 5 3 5 7 8 5 6 0.5794 3 0.1662 4 0.6622 6 C4

14

Bearing

misalignmen

t & improper mounting

Sudden impact on

the rolls

Shaft

damage & Impact

damage on

other parts

Routine

check up

1

0 5 6 5 9

1

0 8

1

0 5 6 6 7 5 7 9

1

0 9

1

0 0.8051 1 0.2079 2 0.9391 1 C5

Electrical

damage

Loss of

power

Operation

interrupted

Electrical wiring check

up

9 1 1 3 5 2 7 9 1 2 1 2 3 4 5 6 2 3 0.2499 14 0.1062 12 0.6444 7 C6

Rolling Mill

Gearing

Failure

Gear teeth

wear-out

Inadequate

lubrication -

Dirt, viscosity

issues

Rough operation &

considerabl

e noise

Routine

check-up of lubrication

7 3 5 3 7 4 7 8 2 3 5 6 3 4 7 8 4 5 0.4419 8 0.1401 8 0.5538 10 C7

Gear teeth

surface fatigue

(Pitting)

Improper meshing,

case depth & high residual

stresses

Gear life reduction

Preventive maintenance

8 5 5 3 5 5 8 9 4 5 5 6 3 4 4 5 5 6 0.4981 7 0.1444 7 0.5797 9 C8

Gear teeth

scoring

Overheating

at gear mesh

Interference

& backlash

phenomeno

n

Lubricating

when needed 5 4 2 3 3 3 4 5 3 4 2 3 2 3 2 3 3 4 0.2515 13 0.0863 14 0.5127 12

C9

Gear teeth fracture

Excessive

overload & cyclic

stresses

Sudden

stoppage of process

plant

Break down maintenance

9 2 6 4 7 7 9 10

2 4 6 7 3 4 7 8 7 8 0.5636 4 0.1700 3 0.6967 4 C10

Gear teeth surface

cold/plastic

flow

High contact stresses due

to rolling &

sliding

Slippage &

power loss

Gear replace

when needed 3 6 3 3 3 3 3 4 6 7 3 4 3 4 2 3 3 4 0.3460 12 0.1022 13 0.4518 14 C11

Rolling

Mill

Shaft (Primar

y &

Secondary)

Failure

Shaft fretting

Vibratory

dynamic

load from

bearing

Leads to

sudden failure

Break down

maintenance 5 5 4 3 3 3 5 6 4 5 3 5 3 4 3 4 3 4 0.3525 11 0.1096 11 0.4617 13 C12

Shaft

misalignment

Uneven

bearing load

Vibration

& fatigue

Preventive

maintenance 8 5 5 3 6 6 8 9 4 5 5 6 4 5 4 5 5 6 0.5505 5 0.1477 6 0.6131 8 C13

Shaft

fracture (Fatigue)

Reverse & repeated

cyclic

loading

Sudden

stoppage of process

Preventive

maintenance 9 2 6 4 6 7 9

1

0 2 3 6 7 3 4 5 6 6 7 0.5455 6 0.1526 5 0.6760 5 C14

15

TABLE III COMPARISON OF RESULTS FOR TRADITIONAL AS WELL AS MCDM (TOPSIS/COPRAS-G/PSI) FMECA

APPROACHES

Objective (iii):

In order to achieve this objective, the comprehensive reliability failure data are analyzed in a

view to identifying the most critical components based on the downtime, frequency of failure,

loss of production in terms of volume and cost for all components of each stand. The

identified critical components for further failure analysis are; bearings (bearing number

32308, 30310, 6213, 32222), gears (primary & secondary bevel gears with spigot and taper

end) and shafts (primary and secondary).

Objective (iv):

This objective is achieved by analyzing recorded historical failure data and actual shop-floor

conditions by methods of questionnaires to floor operators, managers and maintenance

personnel. The potential FMECA, as well as six criteria are derived and scores are assigned

on the scale of 1 to 10 from least to most consideration of the impact of criteria to failure

causes.

Objective (v):

This objective is achieved by applying traditional FMECA initially and results are derived

based on RPN. Then, maintainability criticality indices through three distinct multi-

criteria decision-making approaches; TOPSIS in crisp value, COPRAS-G in grey number

range and PSI without subjective weight consideration, are calculated for each failure cause

in a view to prioritize maintenance activities.

8. Conclusion

Looking to the actual failure analysis, lack of proper maintenance planning is the main reason

for the loss of reliability and poor productivity in process industries. In such condition, it is

necessary to optimize existing traditional maintenance practices based on real shop-floor

Method Result Analysis

1 Traditional FMECA

Most Critical (RPN: > 500) Critical (RPN: 250 < RPN <

500) Normal (RPN: < 250)

C3 C13, C10, C4, C5, C8, C1 C14, C9, C12, C11, C2, C7,

C6

MCDM Methods High MCI Moderate MCI Low MCI

1 TOPSIS Model C5, C3, C4, C10, C14 C13, C8, C7, C1, C12 C2, C11, C9, C6

2 COPRAS-G C3, C5, C10, C4, C14 C13, C8, C7, C1, C2 C12, C6, C11, C9

3 PSI C5, C3, C1, C10, C14 C4, C6, C13, C8, C7 C2, C9, C12, C11

16

conditions. In this research work, the failure pattern study of aluminium wire rolling mill

plant is demonstrated based on actual shop floor conditions through three distinct MCDM

failure models. Following conclusions are drawn from presented research work:

(i) Critical components are identified based on downtime, frequency of failures, loss of

production in volume and cost by analyzing actual historical failure data.

(ii) Research work is focused on potential failure causes of critical components like;

bearings, gears, and shafts of aluminium wire rolling mill which are common to major

process industries.

(iii)Scores for different failure causes are assigned on real shop-floor conditions

(iv) Maintenance planning is proposed in RPN for traditional FMECA and to overcome

the drawback of traditional FMECA, maintainability criticality indices are calculated

through three distinct multi-criteria decision-making approaches are; 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 process industry of

same or of different kinds in accordance with failure analysis.

(vi) The proposed study is challenging and interdisciplinary work; it will help to

understand about the working lives of components and associated failures, which lead

to re-engineer new technologies efficiently and to gain the operational advantage.

(vii) The study will be helpful in designing optimized maintenance plan to improve

plant efficiency as a whole.

9. Recommendations for Future Scope of Work

The research work presented can further be extended as under:

(i) The similar work can be extended to other process industries in a view to deciding

suitable maintenance strategy.

(ii) The results can be validated with similar or different kinds of process industries to

prove competency of MCDM based failure analysis models.

(iii)Similar work can be extended to other process industries such as; petrochemical plant,

textile mill etc. with other MCDM based approaches like; analytical hierarchy process

(AHP), qualitative flexible multi-criteria (QUALIFLEX) etc.

(iv) It is recommended to consider other criteria like; manpower skill, operating

conditions, environmental effect etc. to prioritize maintenance activities of process

industries.

17

(v) It is recommended the scope of the study to present failure model by considering

failures due to the first point with adequate of design for such components at high

failure rate.

(vi) Heuristic approaches can be further used to optimize maintenance activities by

considering reliability parameters as criteria of optimization.

(vii) Failure analysis approaches can be developed for various process industries

considering contemporaneous failures among various systems.

10. List of all publications arising from the thesis (Please Refer Table at the end)

11. References

Research Papers:

1. 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.

2. Barlow and Proshan (1965), Mathematical Theory of Reliability, Wiley, New York.

3. Barlow and Proshan (1975), Statistical Theory of Reliability and Life Testing, Holt,

Rinehart and Winston, New York.

4. 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, 313–324.

5. Braglia M., (2000), “MAFMA: multi-attribute failure mode analysis,” International

Journal of Quality & Reliability Management, vol. 17, no. 9, pp. 1017–1033.

6. 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.

7. 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.

8. 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.

9. Chang C. L., Wei C. C., and Lee Y. H., “Failure mode and effects analysis using

fuzzy method and grey theory,” Kybernetes, vol. 28, no. 9, pp. 1072–1080, 1999.

10. Dekker (1996), Applications of maintenance optimization models: A review and

analysis, Reliability Engineering and System Safety 51 (3), 229–240

11. Deng J. L., (1989), “Introduction to grey system theory,” The Journal of Grey Theory,

vol. 1, no. 1, pp. 1–24.

18

12. 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) 297-310, ISSN 1800-6450.

13. 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.

14. Gilchrist W., (1993), “Modeling failure modes and effects analysis,” International

Journal of Quality & Reliability Management, vol. 10, pp. 16–23.

15. Hwang C. L., Yoon K., (1981), Multiple Attribute Decision Making: Methods and

Applications, vol. 186 of Lecture Notes in Economics and Mathematical Systems,

Springer, New York, NY, USA.

16. Jensen (1995), Stochastic models of reliability and maintenance: an overview, In

Ozekici, S. (Ed.), Reliability and maintenance of complex systems, NATO ASI series,

vol- Springer, Berlin, 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.

17. 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.

18. 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.

19. 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.

20. 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, DOI:

10.1080/23311916.2016.1155839.

21. 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. (p. 1). American Society for

Engineering Management (ASEM).

22. Pham, Wang (1996), Imperfect maintenance, European Journal of Operational

Research, 94, 425–438.

23. Rastegari A., Archenti A., Mobin M., (2017), “Condition-based maintenance of

machine tools: Vibration monitoring of spindle units”, Reliability and Maintainability

Symposium (RAMS), pp. 1-8, IEEE, At: Florida (FL), USA.

19

24. Rathi R., Khanduja D., Sharma S. K., (2016), “A fuzzy-MADM based approach for

prioritising 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.

25. 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.

26. 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.

27. Sikorska J, (2008), “Identifying Failure Modes Retrospectively Using RCM Data”.

28. 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.

29. Zavadskas E. K., Kaklauskas A., Turskis J., and Tamosaitiene J., (2009), “Multi-

attribute decision-making model by applying grey numbers,” Informatica, vol. 20, no.

2, pp. 305–320.

30. Zhang F., (2015), “Failure modes and effects analysis based on fuzzy TOPSIS,” in

Proceedings of the IEEE International Conference on Grey System and Intelligent

Services (GSIS), pp. 588–593, Leicester, UK.

Books:

1. Balagurusamy E, “Reliability Engineering”, ISBN-13: 978-0-07-048339-2, Tata

Mcgraw Hill, New Delhi, 1984

2. Dhillon B. S., “Quality Control, Reliability, and Engineering Design”, ISBN: 0-8247-

7278-4, Marcel Dekker, New York, 1985

3. Hoang Pham, “Handbook of Reliability Engineering”, ISBN: 978-1-85233-453-6,

Springer, 2003

4. Khanna O.P., “Industrial engineering and management”, ISBN-13: 978-818992835,

Dhanpat rai & sons, 2010

5. Mishra, R C & Pathak K, “Maintenance Engineering and Management”, ISBN: 978-

81-203-4573-7, Prentice Hall of India Pvt Ltd., New Delhi, 2012

6. O’Connor P., “Practical Reliability Engineering, ISBN 13: 978-0-470-84462-5, John

Wiley, England, 2002

Websites Reports:

1. “Indian Electrical & Electronics Manufacturers’ Association”, 68TH ANNUAL

REPORT 2014 – 2015 (September 2015), accessed via:

ieema.org/wpcontent/uploads/2015/09/IEEMA-Annual-Report_2014-15.pdf

2. “Indian Electrical Equipment Industry Mission Plan 2012-2022” Central electricity

authority, CEA website,

20

http://www.cea.nic.in/reports/monthly/installedcapacity/2016/installed_capacity-

03.pdf.

3. “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

4. 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/

Aluminium Wire Rolling Mill Process Flow

Melting of Aluminium ingots in furnace

Input

Semi solid cast bar through water

sprinking

Caster

Diameter reduction

through 15 stands in

series

Rolling Mill

Coiling of rod Output Dispatch

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

0.00

0.20

0.40

0.60

0.80

1.00

Apr-13 May-13 Jun-13 Jul-13 Aug-13 Sep-13 Oct-13 Nov-13 Dec-13 Jan-14 Feb-14 Mar-14

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

21

Shop-Floor Activities/Observations at Aluminium Wire Rolling Mill

0

100

200

300

400

500

600

700

800

Pri

mar

y Sh

aft

Pri

mar

y B

evel

…

Pin

fo

r En

try…

Ch

uck

Nu

t fo

r…

Top

Nu

t fo

r…

Spec

ial B

olt

fo

r…

Seco

nd

ary…

Bea

rin

g N

o.…

Bea

rin

g N

o.…

Oil

Seal

. 629

01

0…

Co

up

ler

Bo

lt

Loss

of

Pro

du

ctio

n V

olu

e/C

ost

Rolling Mill Components

Criticality Curve Based on Loss of Production Volume/Cost

C

1

C

2

C

3

C

4

C

5

C

6

C

7

C

8

C

9

C

10

C

11

C

12

C

13

C

14

TOPSIS FMECA 0.40.30.70.50.80.20.40.40.20.50.30.30.50.5

COPRAS FMECA 0.10.10.20.10.20.10.10.10.00.10.10.10.10.1

PSI FMECA 0.70.50.70.60.90.60.50.50.50.60.40.40.60.6

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…COPRAS…

22

List of Publications:

Sr.

No. Research Paper Title

Month-

Year of

Publicati

on

Journal/Conference

Details

ISSN/ISBN/

DOI Remark

1 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

January

2015

2nd

National Conference

on Emerging trends in

Engineering, Technology

& Management

(NCEETM), IU,

Ahmedabad

ISBN: 978-93-

80867-75-5

2 Multi criteria FMECA Based

Decision-Making for Aluminium

Wire Process Rolling Mill through

COPRAS-G

June

2016

Journal of Quality and

Reliability Engineering,

Volume 2016, Article ID

8421916, 8 pages

http://dx.doi.org

/10.1155/2016/

8421916

SCImago Rank:

0.22

Scopus

CiteScore

2016:0.53

H Index:4

J-Gate

3 Performance Reliability

Improvement by Optimizing

Maintenance Practices through

Failure Analysis in Process Industry

– A Comprehensive Literature

Review

December

2016

IOSR Journal of

Mechanical and Civil

Engineering (IOSR-

JMCE), Volume 13, Issue

6 Ver. I (Nov. - Dec.

2016), PP 66-73

e-ISSN: 2278-

1684, p-ISSN:

2320-334X

doi:

10.9790/1684-

1306016673

4 Traditional and Multi-factor

Decision Making based FMECA

through Preference Selection Index

Method for Continuous Process

Industry

January

2017

International Journal of

Darshan Institute on

Engineering Research &

Emerging Technologies

(IJDI-ERET), Vol. 5, No.

2, 2016, www.ijdieret.in

ISSN: (Print):

2320-7590

Impact

Factor:

4.483

5 Identifying Critical Components of

Identified Process Industry through

Shop-floor Failure Data

February

2017

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

International Journal of

Engineering Technology,

Management and Applied

Sciences, February 2017,

Volume 5, Issue 2

ISSN: 2349-

4476

6 TOPSIS and COPRAS-G based

Maintenance Optimization of

Aluminium Wire Rolling Mill

Components

March

2017

Journal of Basic and

Applied Research

International, Vol. 20(3):

pp.189-201, 2017

International Knowledge

Press

ISSN: 2395-

3438 (P),

ISSN: 2395-

3446 (O)

EBSCOhost

(USA)

7 Maintenance Planning through

FMECA based Multi-criteria

Decision-making PSI Approach for

Aluminium Wire Rolling Mill Plant

April

2017

IEEE 2nd

International

Conference for

Convergence of

Technologies (I2CT)

ISBN: 978-1-

5090-4307-1/17

Scopus/

Ei

Compendex

8 Traditional and TOPSIS based

Failure Mode Effect and Criticality

Analysis for Maintenance Planning

of Aluminium Wire Rolling Mill

Components

July 2017

GIT – Journal of

Engineering and

Technology (Tenth

Volume, 2017

ISSN: 2249 –

6157

9 Quality Enhancement in

Maintenance Planning through Non-

identical FMECA Approaches

June 2017

In Printing

International Journal for

Quality Research

ISSN: 1800-

6450

SCImago Rank:

0.234

Scopus

CiteScore

2016:0.78

H Index:7

10 FMECA based Maintenance

Planning through COPRAS-G and

PSI

Accepted

Aug’2017

In Printing

Journal of Quality in

Maintenance Engineering,

Emerald

ISSN: 1355-

2511

SCImago Rank:

0.340

Scopus:

CiteScore

2016: 1.16

H Index:41