Professional Documents
Culture Documents
Defect Control Analysis For Improving Quality and Productivity - An Innovatve Six Sigma Case Study
Defect Control Analysis For Improving Quality and Productivity - An Innovatve Six Sigma Case Study
3, 2011 259
N. Gunasekaran
Angel Collège of Engineering and Technology,
Angel Nagar, Dharapuram Main,
Road, Tirupur – 641 665, Tamilnadu, India
E-mail: guna_kct_cbe_tn_in@yahoo.com
1 Introduction
statistical tools/techniques are used to reduce the defect or process variations. In control
phase, the way of sustaining the improvement is formulated and put in force
(Pyzdek, 2001; Montgomery, 1998). The DMAIC frame work utilises various tools and
techniques like control charts, quality function deployment (QFD), failure mode and
effect analysis (FMEA), design of experiments (DoE) and statistical process control
(SPC) for variation management to drive out defects in operations. Among the available
collection of tools and techniques, application of DoE is at the heart of DMAIC cycle
(Breyfogle, 1999a).
DoE technique helps to identify key process parameters and to subsequently
adjust them in order to achieve sustainable performance improvements (Pande et al.,
2002). This technique determines significant factors and factor interactions that
affect variability within a product at the improvement stage of Six Sigma application
(Antony, 2001). There may be situations when there is a trade-off in the selection
of process parameter levels to identify the optimum parameter level (Antony et al.,
1999). To tackle the conflicting situations, Japanese scientist Taguchi (1986) has
developed robust design approach (Robinson et al., 2003). Robust design is an
efficient and systematic methodology that applies statistical experimental design for
improving product and manufacturing process design (Tsui, 1992). Robust design
approach analyses the signal-to-noise (S/N) ratios as a means of finding a robust
solution (Maghsoodloo, 1990). The loss function is a contribution of robust design
approach to address the process variability. This quadratic loss function informs
about money value and product variability to all management. Classical DoE do not
directly emphasise the reduction of process variability and translate this need to
economical considerations for management (Breyfogle, 1999). Moreover, classical
DoE approach is deemed too costly and time consuming because of full factorial
designs (Ross, 1988). Orthogonal array (OA)-based Taguchi DoE has been recognised as
an economical tool to optimise process parameters in industry for quality and
productivity (Nataraj et al., 2007). Taguchi technique found an alternative
experimental design strategy to drive out the fit fall in classical approach. The basic
promise between classical and Taguchi approach to higher quality is that one proposes
that higher quality costs more and the other proposes that higher quality costs less
(Kackar, 1985).
There are currently few publications that have spoken the application of
Taguchi methodology in Six Sigma projects. This study is an initiative to practice
Six Sigma DMAIC concepts with Taguchi’s experimental design. Since the casting
process is prone to defects, this new approach is proposed in a die casting industry.
Today, most of the automobile components are manufactured by die casting process;
any reduction in defect level can benefit not only the manufacturer but also the end
user. This paper presents a case study performed to eliminate the casting defects in two
stroke engine oil pump body. The organisation was encountering blow holes defect in oil
pump body casting. In Six Sigma improve phase, Taguchi experimental design approach
is used to optimise the casting process parameters. L27 OA is used to make casting
process optimisation with minimum amount of data. The application of a Six Sigma
methodology (DMAIC) with Taguchi’s DoE has developed a structured defect control
framework to improve productivity and product quality of die casting processes with less
cost and time.
262 M. Shanmugaraja et al.
Case 1 Case 2
Focus: productivity Focus: quality
Before After Before After
Total items produces 1,000 1,100 1,000 1,000
Defective items 100 220 100 10
Good items 900 880 900 990
Improvement 90% 80% 90% 99%
Improving productivity through quality can produce all the desired results: better quality,
less rework, greater productivity, lower unit cost, price elasticity, improved customer
satisfaction, larger profits and more jobs. After all, customers get high quality at a low
price, vendors get predictable long-term sources of business, and investors get profits, a
‘win-win’ situation for everyone.
that enables companies to use simple and powerful statistical methods for achieving
and sustaining operational excellence (Harry and Schroeder, 1999). The statistical
representation of Six Sigma describes quantitatively how a process is performing
(Banuelas et al, 2005).
Figure 1 Non-central normal distribution with 3.4 ppm sigma quality level (see online version for
colours)
The numerical goal of Six Sigma is reducing defects to less than 3.4 parts per
million (PPM), reducing cycle time and dramatically reducing costs which impact
the bottom line (Behara et al., 1995; Goh and Xie, 2004). As a process improvement
method, Six Sigma is largely mechanistic or process focused (McAdam and Lafferty,
2004). Figure 1 illustrates the basic measurement concept of Six Sigma (Park,
2003). Sigma quality level includes a ± 1.5 σ shift in mean value to account typical
shifts and drifts of the process mean. Table 2 shows the number of PPM that would be
outside the specification limits if the data were normally distributed with mean shifted to
± 1.5 σ.
Figure 2 Six Sigma DMAIC information flow (see online version for colours)
Phase 4 (Control): This last phase is initiated by ensuring that the new process
conditions are documented and monitored via SPC methods. After the ‘settling
in’ period, the process capability is reassessed. Depending upon the outcome of
such a follow-on analysis, it may become necessary to revisit one or more of
the preceding phases.
2 One-way ANOVA considers the effect of one controlled parameter upon the
performance of a product or process. Here the total variation (SST) is broken into:
266 M. Shanmugaraja et al.
3 Two-way ANOVA is the highest order of ANOVA analysis where two controlled
parameters are in the experimental situation. Here the total variation (SST) is
decomposed into:
• variation due to factor A (SSA)
• variation due to factor B (SSB)
• variation due to interaction of factors A and B (SSA×B)
• variation due to error (SSe).
SST = SS A + SS B + SS A× B + SSe
Based on DoE theory, Taguchi’s OA provide a method for selecting an intelligent subset
of the parameter space. OA is a matrix having rows and columns filled with factor levels
orthogonally. This orthogonal arrangement assures each and every level of each and
every factor comes in combination with each and every level of each and every other
factor. The number of columns in an OA indicates the maximum number of factors that
can be studied. OA is an efficient Taguchi test strategy to evaluate several factors in a
minimum number of tests as shown in Table 3 and obtains much information from few
trails.
Table 3 OA configurations
Orthogonal array Factors and levels No. of tests No. of tests in classical DoE
Two level array
L4 3 factors at 2 levels 4 8
L8 7 factors at 2 levels 8 128
L12 11 factors at 2 levels 12 2,048
L16 15 factors at 2 levels 16 32,768
L32 31 factors at 2 levels 32 2.147E + 09
Three level array
L9 4 factors at 3 levels 9 81
L18 8 factors at 3 levels 18 6,561
L27 13 factors at 3 levels 27 1.594E + 06
Defect control analysis for improving quality and productivity 267
4 Case study
This case study deals with the reduction of casting defects in aluminium pressure die
casting industry. The problem was tackled using a Six Sigma DMAIC problem solving
methodology. The basic equation of Six Sigma, Y = f(x), defines the relationship between
a dependent variable ‘Y’ or the outcome of a process and a set of independent variables or
possible causes which affect the outcome. In this case study, ‘Y’ is the loss in
productivity due to an unacceptable number of defective castings.
is one such type being used in two wheeler’s engine as lubricating pump. Since it is the
heart of engine assembly, presence of any defect in the casing may lead to severe
consequences in operation. At the customer end, the castings are put in heat treatment
process for 24 hours before starting the machining operations. During machining, blow
holes are noticed particularly at the flange section of the component. Presence of such a
casting defect may result in oil seepage when the assembly is put in operation. At this
point of rejection, the casing costs considerable to the client company. Hence, it becomes
imperative to produce porous free castings not only to reduce rejections but also to
minimise the cost of rejection.
In this case, rejections due to blowholes are only concerned. Any other opportunities for
rejection are not accounted. Hence, the number of opportunities is one.
Hence, defects per opportunities (DPO) is:
DPU
DPO =
1
By the same token, defects per million opportunities (DPMO) is:
DPMO = DPO × 1, 000, 000
The sigma quality level with ± 1.5 σ shift is determined (Breyfogle, 1999) by the
equation:
Table 4 summarises the sigma level calculation for all batches and it clearly indicates that
process performance is poor and it needs improvement. Existing process capability is
varying from 1.86 σ to 3.23 σ with an average of 2.51 σ and the defect level ranges
between 42,000 to 346,200 units with a mean of 17.22% out of a million outcomes.
Table 4 Sigma quality level of batches (see online version for colours)
Among the possible parameters, five (metal temperature, intensifier pressure, metal
degassing frequency, 2nd phase turns, and metal mixing ratio) are selected for further
analysis. The rest were distinguished as constant parameters and kept steady throughout
the analysis. This selection of parameters of interest and their working range are based on
field expertise and field executives’ opinion.
Step 2 Selection of OA
It is interested to study the two-factor interaction effects on the component in respect to
formation of blowholes. The selected interactions are:
• metal temperature [P] and intensifier pressure [Q]
• intensifier pressure [Q] and metal degassing frequency [R]
• metal temperature [P] and metal degassing frequency [R].
To select the appropriate OA matrix to fit the case study, it is required to count the
degrees of freedom (DOF). To reach near optimum parameter level combination, the
minimum number of tests that must be performed will be identifiable with DOF. Total
DOF is counted by adding DOF of the number of control parameters which is equal to
one less than the number of level. This case study required minimum of 22 tests to get
optimum parameter settings as shown in Table 6. To cater this situation, an OA with at
least 22 DOF needed to be selected. Hence, L27, a three-level OA with DOF = 27 which
is greater than DOF of the factors and interactions, is selected.
Table 6 Degrees of freedom
Factors/interactions DOF
P 2
Q 2
R 2
S 2
T 2
P×Q 2×2
Q×R 2×2
P×R 2×2
Total DOF 22
Table 7, the interactions PQ, QR and PR are assigned to columns 3 and 4, 8 and 11, and
6 and 7, respectively and resultant L27 matrix is shown in Table 8.
Objective
Columns
function
Test S/N
no. P Q PQ PQ R PR PR QR S T QR * * ratio
Run 1 Run 2
1 2 3 4 5 6 7 8 9 10 11 12 13
1 1 1 1 1 1 1 1 1 1 1 1 1 1 y1,1 y2,1 S/N1
2 1 1 1 1 2 2 2 2 2 2 2 2 2 y1,2 y2,2 S/N2
3 1 1 1 1 3 3 3 3 3 3 3 3 3 y1,3 y2,3 S/N3
4 1 2 2 2 1 1 1 2 2 2 3 3 3 y1,4 y2,4 S/N4
5 1 2 2 2 2 2 2 3 3 3 1 1 1 y1,5 y2,5 S/N5
6 1 2 2 2 3 3 3 1 1 1 2 2 2 y1,6 y2,6 S/N6
7 1 3 3 3 1 1 1 3 3 3 2 2 2 y1,7 y2,7 S/N7
8 1 3 3 3 2 2 2 1 1 1 3 3 3 y1,8 y2,8 S/N8
9 1 3 3 3 3 3 3 2 2 2 1 1 1 y1,9 y2,9 S/N9
10 2 1 2 3 1 2 3 1 2 3 1 2 3 y1,10 y2,10 S/N10
11 2 1 2 3 2 3 1 2 3 1 2 3 1 y1,11 y2,11 S/N11
12 2 1 2 3 3 1 2 3 1 2 3 1 2 y1,12 y2,12 S/N12
13 2 2 3 1 1 2 3 2 3 1 3 1 2 y1,13 y2,13 S/N13
14 2 2 3 1 2 3 1 3 1 2 1 2 3 y1,14 y2,14 S/N14
15 2 2 3 1 3 1 2 1 2 3 2 3 1 y1,15 y2,15 S/N15
16 2 3 1 2 1 2 3 3 1 2 2 3 1 y1,16 y2,16 S/N16
17 2 3 1 2 2 3 1 1 2 3 3 1 2 y1,17 y2,17 S/N17
18 2 3 1 2 3 1 2 2 3 1 1 2 3 y1,18 y2,18 S/N18
19 3 1 3 2 1 3 2 1 3 2 1 3 2 y1,19 y2,19 S/N19
20 3 1 3 2 2 1 3 2 1 3 2 1 3 y1,20 y2,20 S/N20
21 3 1 3 2 3 2 1 3 2 1 3 2 1 y1,21 y2,21 S/N21
22 3 2 1 3 1 3 2 2 1 3 3 2 1 y1,22 y2,22 S/N22
23 3 2 1 3 2 1 3 3 2 1 1 3 2 y1,23 y2,23 S/N23
24 3 2 1 3 3 2 1 1 3 2 2 1 3 y1,24 y2,24 S/N24
25 3 3 2 1 1 3 2 3 2 1 2 1 3 y1,25 y2,25 S/N25
26 3 3 2 1 2 1 3 1 3 2 3 2 1 y1,26 y2,26 S/N26
27 3 3 2 1 3 2 1 2 1 3 1 3 2 y1,27 y2,27 S/N27
⎛1 r
1 ⎞
S / N NB = − log ⎜
⎜r ∑y ⎟
2 ⎟
⎝ i =1 i ⎠
Test No. of good items out of 500 S/N ratio Test No. of good items out of 500 S/N ratio
no. Run 1 Run 2 (HB) no. Run 1 Run 2 (HB)
1 401 452 52.55 15 458 478 53.39
2 462 450 53.17 16 490 486 53.76
3 421 436 52.63 17 369 389 51.56
4 390 401 51.94 18 385 368 51.5
5 369 354 51.15 19 436 455 52.97
6 469 476 53.48 20 395 421 52.19
7 485 468 53.55 21 459 462 53.26
8 359 346 50.93 22 463 475 53.42
9 310 264 49.07 23 401 426 52.31
10 418 431 52.55 24 485 476 53.63
11 479 469 53.51 25 495 486 53.81
12 352 378 51.22 26 425 435 52.66
13 301 320 49.82 27 465 452 53.22
14 329 365 50.77
Factors/levels L1 L2 L3 Factors/levels L1 L2 L3
P 406.3 403.6 450.7 P 52.06 52.02 53.06
Q 432.1 413.1 415.4 Q 52.7 52.22 52.23
R 436.3 402.4 421.9 R 52.7 52.03 52.38
S 420.8 419.4 420.4 S 52.40 52.34 52.39
T 419.7 410.5 430.4 T 52.36 52.14 52.63
P×Q 442.4 425.5 405.4 P×Q 52.60 52.50 52.00
Q×R 412.2 436.1 412.3 Q×R 52.20 52.70 52.20
P×R 425.4 420.9 414.2 P×R 52.60 52.40 52.20
Figure 6 Response curves for process parameters (see online version for colours)
276 M. Shanmugaraja et al.
It is interested to identify the most influencing process parameter causes blow holes.
Pooled ANOVA analysis is done on the output response data at 95% confidence interval
as deployed in Table 12. The ANOVA summary showed that the factor Q, S and T has no
influence on the response where as the factors P (metal temperature) and R (degassing
frequency) have a substantial influence on the response. Also interactions of P with Q
and with R played a considerable role in blowholes formation.
Table 12 Pooled ANOVA analysis on response data (see online version for colours)
To validate the optimum factor level setting for future process, confident interval and the
mean response are estimated as below:
μGood = P3 + Q1 + R1 + S1 + T 3 − 4 M good
where
P3 mean response at level 3 of factor P
Q1 mean response at level 1 of factor Q
R1 mean response at level 1 of factor R
S1 mean response at level 1 of factor S
T3 mean response at level 3 of factor T
Mgood overall mean response value.
Defect control analysis for improving quality and productivity 277
From Table 10, the mean response values are taken and estimated means is calculated as:
μGood = 450.7 + 432.1 + 436.3 + 420.8 + 430.4 − 4(420.18)
μGood = 489.58
Confidence interval for the population is calculated using the following formula (Ross,
1988):
⎛ ⎡ 1 ⎤⎞
CI = ⎜ Fα ;1;Ve Ve ⎢ ⎥ ⎟
⎜ ⎟
⎝ ⎣ ηe ⎦ ⎠
where
Fα ;1;Ve F ratio required for α (risk)
CI = (2.84)(1,929.65)[1 / 54]
CI = 10.1
The estimated mean response is 489.58 and at 90% CI, the predicted optimum output
would be:
Figure 7 Rejection run chart after study (see online version for colours)
Table 13 presents the consequential remarks of Six Sigma implementation with Taguchi
robust design methodology for controlling defects in aluminium die casting industry.
Table 13 Results of the research study
Process evaluation
Contribution
Process parameters SS’ Factor Optimum
SS DOF Cumulative
(SS/DOF) (SS’/total) level
Highly critical factors
Metal temperature [P] 25,150.26 2 12,575.13 54.23 54.23
Critical factor
Degassing frequency 13,790.48 2 6,895.24 29.74 83.97 320 shots/
[R] degas
Non-critical factors
Intensifier pressure 3,851.15 2 1,925.575 8.30 92.27 220 kg/cm2
[Q]
Metal mixing ration 3,567.37 2 1,783.685 7.69 99.96 60:40:00
[T]
2nd phase turns [S] 18.48 2 9.24 0.04 100 3 no
Total 23,188.87 100.00
Performance evaluation
Rejections Productivity
Percent Percent
Min Max Avg savings Total items Defective Good
Productivity savings
(%) (%) (%) produced items items
Before 4.2 34.62 17.22 12.42 10,000 1722 8278 82.78 10.74
study
After 3.2 7.6 4.8 20,000 1296 18704 93.52
study
Defect control analysis for improving quality and productivity 279
The cumulative contribution of process parameters up to 84% is found critical for casting
quality. Metal temperature and degassing frequency were found influencing parameters
for casting defect. Interactions of metal temperature with intensifier pressure and
degassing frequency also found critical for blowholes formation. 12.42% reduction in
defect level has resulted in productivity improvement by 10.74% with marginal cost of
quality control.
7 Conclusions
This paper has explored how best casting industries can take steps to control or minimise
defects through Six Sigma programme. Six Sigma DMAIC approach has been extremely
utilised for statistical analysis of the problem. Taguchi’s robust design practice was used
as improvement strategy to discover optimum process parameter settings. L27 OA-based
experimental design was carried out to study the variations in casting quality at different
levels of the process parameters. From the response graph analysis, it could be observed
that the highest process yield could be attained at the combined settings of parameters
P(3), Q(1), R(1), S(1), and T(3), i.e., metal temperature at 715 degree centigrade,
intensifier pressure at 220 kg/cm2, degassing frequency of 320 shots/degas, 2nd phase
turns at 3 nos. and metal mixing ratio as 60:40. Pooled ANOVA analysis has performed
and found metal temperature and degassing frequency are vital process parameters which
contributed nearly 84% of variations in output casting quality. The experimental results
have been validated by confirmation tests and found the casting process capability
280 M. Shanmugaraja et al.
improved from 2.51 σ to 3.03 σ by reducing the defect rate by 12.42% (17.22–4.8) as
illustrated in Figure 8. The results of this project provided better motivation for the future
applications of Six Sigma methodology across all the components manufactured by the
company. Taguchi DoE has become a part of working culture of the company on the
realisation its worth by the management. The benefits of adopting the Six Sigma problem
solving methodology have been convinced of and it has been linked directly to the
strategic goals of the company.
8 Research implications
Much of this study is devoted to improve the product quality through the process
optimisation. But the costs associated with this improvement drive are not accounted
anywhere in this study. Investing much on quality improvement will have adverse impact
on productivity. This study will have substantial support where the cost of poor quality is
more. Before going to improve either quality or productivity, the cost of improvement
against savings should be analysed. Even this may not be the case for many, as an
improvement drive, it should have favourable results in all aspects of the business. Future
research may be evoked to incorporate cost components into the analysis for
synchronising the improvement on quality and productivity.
References
Antony, J., et al. (1999) ‘Reducing manufacturing process variability using experimental design
technique: a case study’, Integrated Manufacturing Systems, Vol. 10, No. 3, pp.162–169.
Antony, J. and Banuelas, R. (2002) ‘Key ingredients for the effective implementation of Six Sigma
program’, Measuring Business Excellence, Vol. 6, No. 4, pp.20–27.
Antony, J. (2001) ‘Improving the manufacturing process quality using design of experiments:
a case study’, International Journal of Operations and Production Management, Vol. 21,
Nos. 5/6, pp.812–822.
Banuelas, R., Antony, J. and Brace, M. (2005) ‘An application of Six Sigma to reduce waste’,
Quality and Reliability Engineering International, Vol. 21, pp.553–570.
Defect control analysis for improving quality and productivity 281
Behara, R.S., Fontenot, G.F. and Gresham, A. (1995) ‘Customer satisfaction measurement and
analysis using Six Sigma’, International Journal of Quality and Reliability Management,
Vol. 12, No. 3, pp.9–18.
Bessant, J. and Francis, D. (1999) ‘Developing strategic continuous improvement capability’,
International Journal of Operations & Production Management, Vol. 19, No. 11,
pp.1106–1119.
Breyfogle, F.W. (1999a) Implementing Six Sigma – Smarter Solutions Using Statistical Methods,
Wiley, New York, NY.
Breyfogle, F.W. (1999b) Implementing Six Sigma, John Wiley & Sons, Inc., New York, NY.
Bhote, K.R. (2002) The Ultimate Six Sigma: Beyond Quality Excellence to Total Business
Excellence, 1st ed., AMACOM/American Management Association, New York, NY.
Caulcutt, R. (2001) ‘Why is Six Sigma so successful?’, Journal of Applied Statistics, Vol. 28,
Nos. 3–4, pp.301–306.
Cox, J.F., Blackstone, J.H. and Schleier, J.G. (2003) Managing Operations: A Focus on Excellence,
Two Volume Set, North River Press, Great Barrington, MA.
Deming, W.E. (1986) Out of the Crisis, pp.23–24, 97–98, MIT Centre for Advanced Engineering
Study, Cambridge, MA.
Elshennawy, A.K., Maytubby, V.J. and Aly, N.A. (1991) ‘Concepts and attributes of total quality
management’, Total Quality Management, Vol. 2, No. 1, pp.75–97.
Feigenbaum, A.V. (1991) Total Quality Control, 3rd ed., McGraw-Hill, London.
Goh, T.N. and Xie, M. (2004) ‘Improving on the Six Sigma paradigm’, The TQM Magazine,
Vol. 16, No. 4, pp.235–240.
Hahn, G.J. and Doganaksoy, N. (2000) ‘The evolution of Six Sigma’, Quality Engineering,
Vol. 12, No. 3, pp.317–326.
Harry, M.J. (1998) ‘Six Sigma: a breakthrough strategy for profitability’, Quality Progress,
Vol. 31, No. 5, pp.60–64.
Harry, M.J. and Schroeder, R. (1999) Six Sigma: The Breakthrough Management Strategy
Revolutionizing the Worlds Top Corporations, Double Day, New York, NY.
Hobbs, D.P. (2004) Lean Manufacturing Implementation, 1st ed., Ross Publishing, Inc., Boca
Raton, FL.
Jugulum, R. and Dichter, A. (2001) ‘Taguchi methods in American universities and corporations’,
Quality Engineering, Vol. 13, No. 4, pp.607–621
Juran, J.M. (1988) Juran on Planning for Quality, The free press, New York, NY.
Kackar, R. (1985) ‘Off-line quality control, parameter design and the Taguchi method with
discussion’, Journal of Quality Technology, Vol. 17, No. 4, pp.176–209.
Linderman, K., Schroeder, R., Zaheer, S. and Choo, A. (2003) ‘Six sigma: a goal – theoretic
perspective’, Journal of Operations Management, Vol. 21, No. 2, pp.193–203.
Maghsoodloo, S. (1990) ‘The exact relation of Taguchi’s signal-to-noise ratio to his quality loss
functions’, Journal of Quality Technology, Vol. 22, pp.57–67.
McAdam, R. and Lafferty, B. (2004) ‘A multilevel case study critique of six sigma: statistical
control or strategic change?’, International Journal of Operations & Production Management,
Vol. 24, Nos. 5/6, p.530.
Montgomery, D.C. (1998) Introduction to Statistical Quality Control, John Wiley & Sons, Inc.,
New York.
Nair, V. (1992) ‘Taguchi’s parameter design: a panel discussion’, Techno Metrics, Vol. 34, No. 2,
pp.127–161.
Nataraj, M. and Arunachalam, V.P (2006) ‘Optimizing impellor geometry for performance
enhancement of a centrifugal pump using the Taguchi quality concept’, Part A: Proceedings of
Institute of Mechanical Engineers, Vol. 220, pp.765–782.
282 M. Shanmugaraja et al.
Nataraj, M., Arunachalam, V.P. and Suresh, K.G., (2006a) ‘Optimizing planer cam mechanism in
printing machine for quality improvement using Taguchi method: risk analysis with
concurrent engineering approach’, International Journal of Computer Applications in
Technology, Vol. 26, No. 3, pp.164–173.
Nataraj, M., Arunachalam, V.P. and Ranganathan, G., (2006b) ‘Using risk analysis and Taguchi’s
method to find optimal conditions of design parameters: a case study’, International Journal
of Advanced Manufacturing Technology, Vol. 27, pp.445–454.
Nataraj, M., Arunachalam, V.P. and Balaji, B. (2007)’A practical approach to optimize the coating
parameters to win customer confidence’, Part B: Proceedings of Institute of Mechanical
Engineers, Vol. 222, pp.495–506.
Nave, D. (2002) ‘How to compare six sigma, lean and the theory of constraints’, Quality Progress,
Vol. 35, No. 3, p.73.
Pande, P.S., Neuman, R.P. and Cavanagh, R.R. (2000) The Six Sigma Way, 1st ed., McGraw-Hill,
Inc., New York NY.
Pande, P.S., Neuman, R.P. and Cavanagh, R.R. (2002) The Six Sigma Way, Team Field book:
An Implementation Guide for Process Improvement Teams, McGraw-Hill, New York, NY.
Park, S. (2003) Six Sigma for Quality and Productivity Promotion, Asian Productivity
Organization, Japan.
Park, S.H., Hu, M., Barth, B. and Sears, R. (2005) ‘Leveraging Six Sigma disciplines to drive
improvement’, International Journal of Six Sigma and Competitive Advantage, Vol. 5,
Nos. 1 and 2, pp.121–133.
Pyzdek, T. (2001) The Six Sigma Handbook, McGraw-Hill, London.
Pyzdek, T. (2003) The Six Sigma Handbook: A Complete Guide for Green Belts, Black Belts, and
Managers at all Levels, McGraw-Hill, New York, NY.
Robinson, T.J., Borror, C.M. and Myers, R.H. (2003) ‘Robust parameter design: a review’, Quality
and Reliability Engineering International, Vol. 20, pp.81–101.
Ross, P. (1988) Taguchi Techniques for Quality Engineering, McGraw-Hill, London.
Snee, R.D. (2004) ‘Six sigma: the evolution of 100 years of business improvement methodology’,
International Journal of Six Sigma and Competitive Advantage, Vol. 1, No. 1, pp.4–20.
Taguchi, G. (1986) Introduction to Quality Engineering: Designing Quality into Products and
Processes, Asian Productivity Organization, Tokyo.
Tsui, K.L. (1992) ‘An overview of Taguchi method and newly developed statistical methods for
robust design’, IIE Transactions, Vol. 24, pp.44–57.