Professional Documents
Culture Documents
Process
. Type of process: chemical process. Role of Taguchi methods in total quality
. Nature of the problem: low process yield. management
. Size of the experiment: eight trials.
There are many definitions of quality but the
. Benefits: process yield was improved by
definition of quality proposed by Dr Taguchi
over 10 per cent.
(Kolarik, 1995) is more relevant in terms of
Electronics and semi-conductors working towards target performance of
. Type of process: wire bonding process. product/process. Indeed working towards the
. Nature of the problem: low wire pull target performance reflects the continuous
strength and therefore large customer improvement attitude. Although not explicitly
80
An application of experimental design for process optimization The TQM Magazine
Hefin Rowlands, Jiju Antony and Graeme Knowles Volume 12 . Number 2 . 2000 . 78±83
stated, Deming's definition implies that the categorised as control or noise factors as
needs of the customer may change. Also with shown below:
the view that new customers may be added in . Control factors: die temperature, hold
the future, this suggests a dynamic definition time, batch number, maximum force
of quality. Quality issues do not stay still and during hot forming and force application
therefore we must be prepared to adapt our rate.
ideas and views on quality in order to meet . Noise factors: dimensional variation of
the ever-changing needs of our customers. valve body, depth of stake, material
This idea is consistent with the approach of variation and rate of hot forming.
continuous improvement. . Control/noise factor: strain rate during pull-
The ethos of total quality management is out.
continuous improvement. In order to improve All five control factors were used for the
our product/process quality, we need to experiment. The strain rate pull-out was used
measure appropriate quality characteristic(s), as a noise factor and controlled at two levels
which are most critical to our customers. In for the experiment to simulate a varying load
otherwords, we cannot simply manage any on the product when in use. Due to the
process without measurement. The role of difficulties and expense of manufacturing
Taguchi methods in TQM is to identify and valve bodies to different dimensions and with
optimise the critical quality characteristics variations in material, it was decided not to
which affect the final product/process include these in the experiment.
performance. The levels for each factor were selected
systematically by the engineers based on their
knowledge and experience of the process. The
Taguchi case study levels chosen are shown in Table I. As a
reference for the results of the experiment,
This section details one of the Taguchi based on the experience of the engineers, the
experiments that the students have following levels were considered to give the
implemented at their workplace. In many best performance: A2, B1, C1, D1, E1.
cases this would have been the first The experiment was carried out using a
application of such a technique in the modified L8 orthogonal array (Taguchi and
company. The results of the study have Konishi, 1987) to account for the four levels
provided a greater stimulus in terms of the of factor A. The S/N calculations are based on
wider application of the Taguchi methods in a larger the better S/N ratio.
other core processes. The objective of the Table II illustrates the average pull-out
study was to optimise the production process values and average S/N ratio values
of retaining a metal ring in a plastic body in a corresponding to each level of the chosen
braking system by a hot forming method. factors.
The production process consisted of a Figure 1 illustrates the mean response
heated die, which was forced down by air graph which simply plots the average pull-out
pressure onto a valve body forming a plastic strength values at each level of the selected
lip into which a retaining metal ring was factors. Figure 2 illustrates the S/N ratio plot
inserted. Although the process was fairly which basically shows the average S/N ratio
straightforward it was felt that the maximum values at each factor level.
strength of the product was not being In order to determine which of the effects
achieved. A test rig was designed to simulate are statistically significant, it was decided to
the production process and to enable a series perform the Analysis of Variance (ANOVA).
of experiments to be performed. A Taguchi ANOVA is a powerful technique which
style experiment was carried out to identify
the process variables that would provide the Table I Factor levels
greatest and consistent pull-out strength. This Factor Level 1 Level 2 Level 3 Level 4
was tested on a standard tensometer.
A Die temperature (deg. C) 180 200 220 240
A brainstorming session consisting of chief
B Hold time (sec) 5 15
development engineers, senior design
C Batch no. 1 2
engineers, plastic engineers, identified a list of
D Maximum force (KN) 6 7
factors which were thought to affect the pull-
E Force application rate (KN/sec) 5 1
out strength. These factors were then
81
An application of experimental design for process optimization The TQM Magazine
Hefin Rowlands, Jiju Antony and Graeme Knowles Volume 12 . Number 2 . 2000 . 78±83
Table II Response table ± mean pull-out strength and mean S/N ratio tables, F0.05,1,2 = 18.51, F0.05,3,2 = 19.16,
Factor levels Average (KN) Average S/N ratio F0.10,1,2 = 8.53 and F0.10,3,2 = 9.16. This
indicates that only factor B is statistically
A1 2.72 8.44
significant at the 90 per cent confidence level.
A2 2.93 9.27
A3 3.29 10.30 For the mean analysis, F0.05,1,17 = 4.45,
A4 4.12 12.26 F0.05,3,17 = 3.20, F0.10,1,17 = 3.03 and
B1 2.92 9.09 F0.10,3,17 = 2.44. This clearly indicates that
B2 3.60 11.04 factors A, B, C and D have significant effect
C1 3.15 9.65 on the mean pull-out strength.
C2 3.38 10.49
D1 3.15 9.68 Determination of optimal factor
D2 3.38 10.45 settings
E1 3.26 10.02
E2 3.27 10.12 The optimal settings are those which provide
the best process/product performance based
on the obtained data from the experiment. In
this case, the objective was to maximise the
Figure 1 Mean reponse plot for the experiment
pull-out strength with minimum variation. As
only factor B has significant effect on the S/N
ratio, we have chosen level 2 of factor B, as it
provides a higher S/N ratio. It is important to
note that the higher the S/N ratio, the greater
the process robustness and product
performance. Having analysed the S/N ratio,
the next step was to determine the factor
settings that yielded the maximum pull-out
strength. In this case, we have again chosen
Figure 2 Mean S/N ratio plot for the experiment
those factor settings with higher mean pull-
out strength values. The analysis of mean
pull-out strength values has provided the
following factor settings:
. Factor A ± level 4;
. Factor B ± level 2;
. Factor C ± level 2; and
. Factor D ± level 2.
The predicted average pull-out strength
sub-divides the total variation in the data into obtained based on the optimal factor settings
useful and meaningful components of is approximately 4.5KN. A confirmation run
variation. The results of the ANOVA are gave results close to the prediction. This
shown in Tables III and IV respectively. shows a significant improvement on the pull-
In order to determine the significant effects, out strength compared to the average pull-out
the calculated F-ratios are compared with the strength value of 3.26, i.e. an increase of
tabled F-ratios. For the S/N analysis, from F about 38 per cent.
82
An application of experimental design for process optimization The TQM Magazine
Hefin Rowlands, Jiju Antony and Graeme Knowles Volume 12 . Number 2 . 2000 . 78±83
Commentary
A detailed exploration of the Taguchi method and its role in process optimization.
83