Taguchi Method

BY
PROF. AMIT KUMAR
PEC CHANDIGARH
Robust Design 2

Dr. Genichi Taguchi, a robust design is one that is created with a system of design
tools to reduce variability in product or process, while simultaneously guiding the
performance towards an optimal setting.

A product that is robustly designed will provide customer satisfaction even when
subjected to extreme conditions on the manufacturing floor or in the service
environment.
Taguchi Method – Robust Design 3

Taguchi method is based on performing evaluation or experiments to test the

sensitivity of a set of response variables to a set of control parameters (or independent
variables)by considering experiments in “orthogonal array” with an aim to attain the
optimum setting of the control parameters.
Orthogonal Arrays 4

• Orthogonal arrays provide a best set of well balanced (minimum) experiments.

• The number of rows of an orthogonal array represents the requisite number of
experiments.
• The number of rows must be at least equal to the degrees of the freedom
associated with the factors i.e. the control variables.
• In general, the number of degrees of freedom associated with a factor (control
variable) is equal to the number of levels for that factor minus one.
OA Representation 5

𝐿𝑁 (𝑆 𝑘 )

Where,
N – Total number of trials during experiment
S – Number of levels for each factor
k – Maximum number of factors whose affect can be estimated without any interactions.
Standard Orthogonal Arrays
6
7
Signal to Noise Ratio (S/N) 8

The signal to noise ratios (S/N), which are log functions of desired
output, serve as the objective functions for optimization, help in data
analysis and the prediction of the optimum results.

The Taguchi method treats the optimization problems in two

categories: static problems and dynamic problems.
Types of S/N : Static Problems 9

1. Smaller the better

𝑛 = −10 𝐿𝑜𝑔10 [𝑚𝑒𝑎𝑛 𝑜𝑓 𝑠𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 𝑜𝑓 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 𝑑𝑎𝑡𝑎]

This is usually the chosen S/N ratio for all the undesirable characteristics like “defects” for which
the ideal value is zero. When an ideal value is finite and its maximum or minimum value is
defined (like the maximum purity is 100% or the maximum temperature is 92K or the minimum
time for making a telephone connection is 1 sec) then the difference between the measured data
and the ideal value is expected to be as small as possible. Thus, the generic form of S/N ratio
becomes,

𝑛 = −10 𝐿𝑜𝑔10 [𝑚𝑒𝑎𝑛 𝑜𝑓 𝑠𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 𝑜𝑓 (𝑚𝑒𝑎𝑠𝑢𝑒𝑟𝑑 − 𝑖𝑑𝑒𝑎𝑙)]

Types of S/N : Static Problems 10

2. Larger the better

𝑛 = −10 𝐿𝑜𝑔10 [𝑚𝑒𝑎𝑛 𝑜𝑓 𝑠𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 𝑜𝑓 𝑟𝑒𝑐𝑖𝑝𝑟𝑜𝑐𝑎𝑙 𝑜𝑓 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 𝑑𝑎𝑡𝑎]

This is often converted to smaller-the-better by taking the reciprocal of the measured data and
next, taking the S/N ratio as in the smaller-the-better case.
3. Nominal-the-best
This case arises when a specified value is the most desired, meaning that neither a smaller nor a
larger value is desired.
𝑠𝑞𝑢𝑎𝑟𝑒 𝑜𝑓 𝑚𝑒𝑎𝑛
𝑛 = −10𝐿𝑜𝑔10
𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Full Factorial Design vs Fractional Factorial 11
Design
In Full Factorial Design the total number of experiments required to run all possible
combinations of all the levels for each factor.
In Fractional Factorial Design a portion of total combinations is studied in such a way so that
there is no effect on the output.
Steps in Taguchi method 12

1. Decide the important input process parameters and there levels (by pilot experimentation or
literature survey), response parameter and its characteristics.
2. Select the appropriate OA and assign the parameters to its various columns.
3. Conduct the experiments for the levels given in each row RANDOMLY and note down the
values of the response parameter. Conduct each experiment THREE times.
4. Study factor effect and find out the optimum combination of input parameters. Calculate the
best value of response parameter.
5. Calculate the range within which the experiment value should lie and conduct confirmation
experiment (if required) to verify the same.
6. Perform Analysis of Variance (ANOVA) to find out the significance of the various factors
and their relative contributions.
Orthogonal Array 13
Example - 1
14
In the present study, the goal is to evaluate the effects of process parameters on
the performance measure and the optimum combination of control factors that
would maximize the compressive strength, of the fly ash brick (prepared in the
laboratory), which is chosen as the quality characteristic.
Selection of control factors and their levels are made on the basis of some preliminary trial
15
experiments conducted in the laboratory and also from literature review on the subject.
Four control factors such as water/binder–ratio, fly ash, coarse sand, and stone dust are
selected for the study. Each of the four control factors is treated at three levels, as shown in
Table 1. The choice of three levels has been made because the effect of these factors on the
performance characteristic may vary nonlinearly.
An orthogonal array is a fractional factorial design with pair wise balancing
property. Using orthogonal array design the effects of multiple process variables on
16
the performance characteristic can be estimated simultaneously while minimizing
the number of test runs. An 𝐿9 (34 ) standard orthogonal array as shown in Table 2
was employed for the present investigation. This array is most suitable to provide
the minimum degrees of freedom as 9 [= 1 + 4 x (3–1)] required for the
experimental exploration.
Actual factors based on OA 17
Results and Analysis 18
S/N used in Example-1 19

Larger the better

𝑛 = −10 𝐿𝑜𝑔10 [𝑚𝑒𝑎𝑛 𝑜𝑓 𝑠𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑠 𝑜𝑓 𝑟𝑒𝑐𝑖𝑝𝑟𝑜𝑐𝑎𝑙 𝑜𝑓 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 𝑑𝑎𝑡𝑎]

𝑛
1 1
𝑛 = −10 𝐿𝑜𝑔10 ෍
𝑛 𝑌𝑖𝑗
𝑗=1

Where 𝑌𝑖𝑗 is the value of quality characteristics for 𝑖𝑡ℎ trial and 𝑗𝑡ℎ experiment.
20
 21
ANOVA for 95% Confidence Level
Example-2 22

Determine the effect of four process parameters: temperature (A), pressure (B),
setting time (C), and cleaning method (D) on the formation of surface defects in a
chemical vapor deposition (CVD) process to produce silicon wafers. Also estimate the
optimum setting of the above process parameters for minimum defects. Table 5.4.3
depicts the factors and their levels.
Step 1: Select the design matrix and perform the experiments 23
Since 4 factors and 3 levels so a L9 matrix can be chosen for experimentation.

Step 2: Perform the experiments

Conduct the experiments and observe the surface defect counts per unit area at three
locations each on three silicon wafers (thin disks of silicon used for making VLSI circuits)
so that there are nine observations in total for each experiment. The summary statistic, 𝜂𝑖 ,
for an experiment, i, is given by

𝜂𝑖 = −10𝐿𝑜𝑔10 𝐶𝑖

where Ci refers to mean squared effect count for experiment i and the mean square refers
to the average of the squares of the nine observations in the experiment i.
The L9 Orthogonal Array 24
Step 3: Calculation of factor effects 25
The effect of a factor level is defined as the deviation it causes from the overall mean.
Hence as a first step, calculate the overall mean value of η for the experimental region
defined by the factor levels in table as
ധ -41.67 dB
Overall Mean (𝑋)=
The effect of the temperature at level A1 (at experiments 1, 2 and 3) is calculated as the
difference of the average S/N ratio for these experiments (mA1) and the overall mean. The
same is given as
The effect of temperature at level A1 = 𝑋ത 𝐴1 – 𝑋ധ
Similarly,
The effect of temperature at level A2 = 𝑋ത 𝐴2 – 𝑋ധ
Also effect of A3 can be found.
Average 𝜂 for different factor levels 26
Using the S/N ratio data available in table, the average of each level of the four
factors is calculated and listed in table. These average values are shown below. They
are separate effect of each factor and are commonly called main effects.
Plots for factor effects 27
Step 4: Selecting Optimum factor Levels 28
Our goal in this experiment is to minimize the surface defect counts to improve the
quality of the silicon wafers produced through the chemical vapor deposition process.
Since –log depicts a monotonic decreasing function we should maximize η. Hence the
optimum level for a factor is the level that gives the highest value of η in the experimental
region. From plot and table, it is observed that the optimum settings of temperature,
pressure, settling time and cleaning method are A1, B1, C2 and D2 or D3. Hence we can
conclude that the settings A1B1C2D2and A1B1C2D3 can give the highest η or the lowest
surface defect count.
Step 5: Developing the additive model for factor effects 29
The relation between η and the process parameters A, B, C and D can be approximated
adequately by the following additive model:

where the term m refers to the overall mean (that is the mean of η for the experimental
region). The terms ai, bj, ck and dl refer to the deviations from μ caused by the setting Ai,
Bj, Ck, and Dl of factors A, B, C and D, respectively. The term e stands for the error. In
additive model the cross- product terms involving two or more factors are not allowed.
Equation (10) is utilized in predicting the S/N ratio at optimum factor levels.
Step 5: Analysis of Variance (ANOVA) 30
For ANOVA and more on this example following link can be followed.
 31

For more research on Taguchi method by PECIANS follow the link

32