Why Taguchi Method Tip

1
Of Course It is tip 13 as well as It is Friday
the 13th
Friday, 13th July 2001
2
When Taguchi Method is NOT Appropriate
Tip 13
Friday, 13th July 2001
3
No NoIsE
When Taguchi Method is NOT Appropriate

• No NoIsE
• When you can not think of NoIsE that can be
  included during the experiment
• NoIsE can be included during experiments
• but can not think of Control Factors that have
  strong correlation to the NoIsE
• but can not effectively capture the effects of
  NoIsE

4
When Taguchi Method is NOT Appropriate in each of
the 8-STEPS

• IDENTIFY THE MAIN FUNCTION, SIDE EFFECTS,
  AND FAILURE MODE
• IDENTIFY THE NOISE FACTORS, TESTING
  CONDITIONS, AND QUALITY CHARACTERISTICS
• IDENTIFY THE OBJECTIVE FUNCTION TO BE
  OPTIMIZED
• IDENTIFY THE CONTROL FACTORS AND THEIR
  LEVELS
• SELECT THE ORTHOGONAL ARRAY MATRIX
  EXPERIMENT
• CONDUCT THE MATRIX EXPERIMENT
• ANALYZE THE DATA, PREDICT THE OPTIMUM LEVELS
  AND PERFORMANCE
• PERFORM THE VERIFICATION EXPERIMENT AND PLAN
  THE FUTURE ACTION

5
Step 1 Main Function / Side Effects
When Taguchi Method is NOT Appropriate

• Main Function
• When you have no clue as to what is the Ideal
  Final Result (the distance between the
  current result and IFR gives the necessary
  boldness to vary Control factor levels widely
  enough to exploit non-linearities)
• Side Effects
• When you can not think of Side-Effects
• you can not think of NoIsE that can cause such
  side-effects

6
Step 2 Including NoIsE
When Taguchi Method is NOT Appropriate

• Include NoIsE ???
• When you can not think of NoIsE that can be
  included during the experiment
• When you can not think of NoIsE that can be
  included during the measurements
• When you can not think of NoIsE that is
  analogous to aging or slow degradation
• during the experiment or measurements

7
Step 2 Capturing effects of NoIsE
When Taguchi Method is NOT Appropriate

• NoIsE can be included during experiments
• but can not think of Control Factors that have
  strong correlation to the NoIsE
• but can not effectively capture the effects of
  NoIsE
• External NoIsE (in explicitly added NoIsE
  Factors)
• Internal NoIsE (in Control Factors)
• but do not wish to increase either the
• experimental effort
• experimental resources

8
Step 3 Quality Characteristics / Objective
Function
When Taguchi Method is NOT Appropriate

• When you can not think of Quality
  Characteristics that closely represents the
  energy transfer mechanism in the main function
• When the Quality Characteristics can not be
  quantitatively measured
• When the Quality Characteristics is not
  monotonous (and has phase-transitions or
  represents a multiple valued function)

9
Step 3 Quality Characteristics / Objective
Function
When Taguchi Method is NOT Appropriate

• When you can not think of Variations in
  quality Characteristics as being important.
• In other words, you are able to give importance
  only to the mean value
• When you are interested only in improving the
  mean, or even worse, you are interested only
  in studying the factor effects (on mean)
• When you are not interested in identifying
  Control Factors
• which help reduce the Variance
• which help adjust the mean

10
Step 3 Quality Characteristics / Objective
Function
When Taguchi Method is NOT Appropriate

• When you can think of only one (desirable)
  Quality Characteristics and can not think of
  another (desirable or undesirable)
• When you can not think of two contradictory
  requirements i.e. Quality Characteristics
• (While Taguchi Method is capable of improving
  both)
• When you are not able to give priority to
• Tomorrows Problem (reducing Variance)
• and end up giving priority to
• Todays Problem (improving Mean)

11
Step 3 No Need to determine an Adjustment
Factor
When Taguchi Method is NOT Appropriate

• When you can think of Quality Characteristics
  that have more to do with mean like
  smaller-the-better or Larger-the-Better and can
  not think of any other Quality Characteristics
  that has to do with variance like
  Nominal-the-best
• When there is no need or scope of finding an
  adjustment factor (defined as the control factor
  that has negligible effect on variance and large
  effect on mean)

12
Step 4 Number of Control Factors and NoIsE
Factors
When Taguchi Method is NOT Appropriate

• When you can not think of Control Factors that
  are strongly correlated to NoIsE Factors
• When the number of Control Factors is not even
  twice the number of NoIsE Factors (This is a
  thumb rule originating from the assumption
  that at least one of the two control factors will
  have a favorable and strong nonlinearity that
  will help reduce the effect of NoIsE on the
  Quality Characteristics)

13
Step 4 Control Factors Levels (t o o w i d e
or too narrow)
When Taguchi Method is NOT Appropriate

• When Control Factors are chosen correctly (in the
  sense that these are strongly correlated to NoIsE
  Factors as well as have strong effect on Quality
  Characteristics) but the levels are not wide
  apart, with the result that the nonlinearity is
  not fully exploited (ending up in getting only
  sensitivity)
• On the other hand,
• When the Levels of one of the Control Factors are
  so widely separated that only that control factor
  dominates (and other control factors show less
  than 5 effect)
• For example Temperature in a bio-culture growth
  has levels of 25ºC, 37º C and 50º C
• This will dominate over all other control factors

14
Step 5 Select the inner Orthogonal Array
When Taguchi Method is NOT Appropriate

• When you can not guarantee that all the Control
  Factors are indeed orthogonal to each other and
  you have chosen an orthogonal array that does not
  allow study of all suspected interactions
• When the number of Control Factors and the chosen
  OA is such that there are no degrees of freedom
  left for estimating error (this forces one to
  declare control factors with less than 15 effect
  to be pooled as error)

15
Step 5 Select the outer Orthogonal Array
When Taguchi Method is NOT Appropriate

• When the OA selected for NoIsE factors (also
  called the outer array) is bigger than the main
  OA (also called the inner array) for Control
  Factors. (The main idea behind using an outer
  OA is to reduce the number of testing conditions
  and a bigger array defeats this main purpose).
• While the outer array primarily gives the
  desired worst case conditions, it should not
  lead to failure of the experiment. (failure
  could be defined as not able to quantitatively
  measure Quality Characteristics or causing
  damage/breakdown of the process equipment)

16
Step 6 Conduct the Matrix experiment based on
inner and outer OAs
When Taguchi Method is NOT Appropriate

• When the experimental conditions (other than the
  combinations of control factors that appear in
  the inner or outer OAs) can not be
  maintained over the entire Matrix experiment
• When NoIsE can not be effectively captured on/in
  the samples or during the measurements
• When all experiments are not satisfactorily
  completed (even one less would give incorrect
  calculation of factor effects and predictions)

17
Step 6 and 7 Make the Measurements and
calculate the S/N Ratios
When Taguchi Method is NOT Appropriate

• Zero-Reading for a Larger-the-better type S/N
  Ratio or identical readings for Nominal-the-best
  type S/N Ratios (both give rise to division by
  zero when evaluating the above mentioned S/N
  ratios)
• If you get one measurement less than the
  detection sensitivity or multiple measurements
  within the measuring accuracy of the measuring
  apparatus
• In fact, including NoIsE helps here, the
  measurements becomes larger than the least-count
  or measuring accuracy

18
Step 8 Conduct the Confirmation /
Verification Experiments
When Taguchi Method is NOT Appropriate

• When the confirmation experiments give results
  that are not close to the predicted results
  (i.e. are not within the prediction error)
• Some important control factor is not chosen
• Some NoIsE factor that has a dominant effect
• NoIsE is not captured effectively
• There is no control factor that has strong
  correlation to NoIsE
• Interaction between Factors There is
  interaction between two dominant control factors
  and it has not been studied or the chosen OA does
  not allow this interaction to be studied

19
Earlier Tips Links below

• Friday, 27th July 2001
• Friday, 20th July 2001
• Friday, 13th July 2001

• Taguchi Method
• inner L9 array with outer L4 and L9 NoIsE
  arrays
• Taguchi Method
• inner L18 array with outer L4 and L9 NoIsE
  arrays
• Taguchi Method Why/When is Taguchi Method not
  Appropriate?

Tips 12, 11, 10 ?
20
Earlier Tips Links below

• Friday, 6th July 2001
• Friday, 29th June 2001
• Friday, 22nd June 2001

• Taguchi Method
• inner L8 array with outer L4 and L9 NoIsE
  arrays
• Taguchi Method
• Useful at ALL Life-stages of a Process or
  Product
• Taguchi Method
• Performs Process centering or fine tuning

Tips 9, 8, 7 ?
21
Earlier Tips Links below

• Taguchi Method Identifies the right NoIsE
  factor(s) for Tolerance Design
• Taguchi Method
• Finds best settings to optimize TWO quality
  characteristics Simultaneously
• 7. Taguchi Method
• When to select a Larger OA to perform
  Factorial Experiments

• Friday, 15th June 2001
• Friday, 8th June 2001
• Friday, 1st June 2001

Tips 6, 5, 4 ?
22
Earlier Tips Links below

• Friday, 25th May 2001
• Friday, 18th May 2001
• Friday, 11th May 2001

• Taguchi Method Using Orthogonal Arrays for
  Generating Balanced Combinations of NoIsE Factors
• Taguchi Method Signal-to-Noise Ratio for Quality
  Characteristics approaching IDEAL value
• 4. Taguchi Method improves " quality at
  all the life stages
• at the design stage itself

Tips 3, 2, 1 ?
23
Earlier Tips Links below

• Friday, 4th May 2001
• Friday, 27th April 2001
• Friday, 6th April 2001

• 3. Taguchi Method Appropriate for Concurrent
  Engineering
• 2. Taguchi Method can study Interaction
• between Noise Factors and Control
  Factors
• 1. Taguchis Signal-to-Noise Ratios are in Log
  form

24
end