The History About Taguchis Contributions

Published Date: 02 Nov 2017

Planning the Experiment

Conducting the Experiment

Analyzing the Response [8]

11.1 Planning the Experiments:

Everyone involved in the experiment should have a clear idea in advance of exactly what is to be studied, the objectives of the experiment, the questions one hopes to answer and the results anticipated

Select a response/dependent variable (variables) that will provide information about the problem under study and the proposed measurement method for this response variable, including an understanding of the measurement system variability

Select the independent variables/factors (quantitative or qualitative) to be investigated in the experiment, the number of levels for each factor, and the levels of each factor chosen either specifically (fixed effects model) or randomly (random effects model).

Choose an appropriate experimental design (relatively simple design and analysis methods are almost always best) that will allow your experimental questions to be answered once the data is collected and analyzed, keeping in mind tradeoffs between statistical power and economic efficiency. At this point in time it is generally useful to simulate the study by generating and analyzing artificial data to insure that experimental questions can be answered as a result of conducting your experiment.

Orthogonal Arrays &

Linear Graphs [8]

Orthogonal Arrays (OA):

A set of column and numbers1, 2 or 1, 2, 3 etc.

Columns are orthogonal to each other

Specially developed for assigning or allocating factors during experiments

Advantages of Orthogonal Arrays:

Time to design or plan an experiment is minimized [18]

Orthogonal Arrays (OA): Example

Figure : Orthogonal Array example [16]

Linear Graphs:

Simple Graphs like Rectangle, Hexagon etc. with nodes and lines joining the nodes.

Pictorial representation of Orthogonal Arrays.

Nodes represent Factor Columns of OA.

Lines indicate Interaction Columns of OA.

Linear Graphs:Example

Linear Graph of L8 OA

Figure : Linear Graph of L8 OA [16]

11.2 Conducting the Experiment: Taguchi’s Contributions

Perform the experiment (collect data) paying particular attention such things as randomization and measurement system accuracy, while maintaining as uniform an experimental environment as possible. How the data are to be collected is a critical stage in DOE. Based on the results of the analysis, draw conclusions/inferences about the results, interpret the physical meaning of these results, determine the practical significance of the findings, and make recommendations for a course of action including further experiments

While conducting the experiment

Simulate the real life situation.

Replicate the experiments in best as well as worst real life situation. [8]

Example 1: Conducting an Experiment to optimize a machining process.

Best Situation: New Tool + Clean Coolant

Worst Situation: Worn out Tool + Contaminated Coolant

Example 2: Conducting an Experiment to optimize the performance of electrical equipment.

Best Situation: Nominal or Desired Voltage

Worst Situation: Low Voltage or High Voltage

Noise: The real life variation is not under manufacture’s control.

Example: Voltage fluctuations, Tool Wear, Coolant Contamination, etc.

Noise Level 1: Best Situation

Noise Level 2: Worst Situation

Experimental Layout: Taguchi’s Suggestion

Figure : Comparison with Noise factors example [16]

11.3 Analysis of Experiments: Taguchi’s Contributions

Analyze the data using the appropriate statistical model insuring that attention is paid to checking the model accuracy by validating underlying assumptions associated with the model. Be liberal in the utilization of all tools, including graphical techniques, available in the statistical software package to insure that a maximum amount of information is generated

Example:

Figure : Taguchi’s Method example [16]

Target: 3.0

Example:

Figure : Level Graph [16]

To get this insight:

Do not analyze the response directly

Calculate MSD =  (Response - Target)2 / n

Analyze S / N Ratio = -10 log MSD

Classification of Data:

Smaller the better Type

Target: 0

E.g.: Surface Finish, Run out, Contamination etc.

Larger the better Type

Target: Infinity

E.g.: Hardness, Tensile strength etc.

Nominal the better Type

Target: Neither 0.0 nor 

E.g.: Diameter, Length etc.

Smaller the better Type:

MSD =  (Response - Target)2 / n

Target = 0.0

MSD =  (Response - 0)2 / n =  (Yi - 0)2 / n

=  (Yi)2 / n

S / N Ratio = -10 log MSD

Example: Smaller the better Type

Figure : Smaller the better example [16]

Target: 0.0

Larger the better Type:

MSD =  (Response - Target)2 / n

Target = 

MSD =  (Response - )2 / n =  (Yi-  )2 / n

=  (1/Yi- 1/) 2 / n =  (1/Yi)2 / n

S / N Ratio = -10 log MSD

Example: Larger the better Type

Figure : Larger the better example [16]

Target: Infinity

Nominal the Better Type

MSD =  (Response - Target)2 / n

Target = T (Neither 0 or infinity)

MSD =  (Yi - T)2 / n

= S2 = (Ybar - T)2

S: Standard Deviation of Responses Yi

Y bar: Average of Responses Yi

Analyze -10 Log S2 and Y bar separately

Example: Nominal the better Type

Figure : Nominal the better example [16]

Target: 3.0

Example: Nominal the better Type Selection of Optimum Combination:

For Factors affecting -10Log S2 Levels which maximizes -10 Log S2

For Factors affecting Y bar Levels with minimum deviation from Target [8,10]

12. ROLES & BENEFITS

Why Design Of Experiments? 

DOE are widely used to optimize the process or product in very first step by minimizing the process variation by maximizing the signal of noise factors ratio of the controllable factors that affect variation and select the levels of the factor(s) that affect the mean to adjust in the desired direction.[20]

12.1. The Power of DOE

DOE is a powerful analytical method that can be learnt by technical professionals from basic level training program me, which provides a cost-effective and organized approach to conduct industrial experiments with multiple process variable study at the same time with these efficient designs, instead of using trial and error approach, DOE provides very reproducible results due to the statistical approach. This kind of tool not only saves experimental costs, but it greatly increases the accuracy in identifying the real and odd factors which are normally hard-to-find by other Quality tools and method for quality problems. [7]

12.2. How Does DOE Fit in the SIX SIGMA Tool Box?

DOEs are performed in the "Analyze" & "Improve" phases of the "DMAIC" in that phase breakthrough Strategy DOEs determine the impact that variables have on a product or process characteristic which can monitored by Process Maps are by doing FMEAs to identify the inputs (x) to be investigated and or validated by Process Mapping & FMEAs to find identify what outputs (y) need to be measured. MSEs are performed on all important measurement systems before the DOE. [20]

Figure : DMAIC CYCLE. [4]

12.2.1 Six Sigma and Total Quality Management:

Six-Sigma is a combination of business philosophies to gain more profit and to fulfill customer requirements. Six-Sigma is acting as a strategic and cultural change in a company which implements new vision, develop the competencies, build activated teams and great infrastructure for the company to keep on sustainable improvement. The Six Sigma activities by applying proper tools and techniques such as TRIZ or QFD will make use of resources more effective and efficient. It increases customer satisfaction by eliminating causes of dissatisfaction, improving performance and introducing unexpected features and finally enhances the value of the product. This method can be very effective for a company but it is also connected with a lot of effort to implement the basic principles and to convince every employee to take part at this strategy. Six sigma has two main concepts. The first is to reduce the defects to 3.4 defects per Million opportunities as a statistical approach. The other one is a business perspective, which focuses on increasing the profit and customer satisfaction. Total quality management established well before Six sigma methodology and six sigma carries some of the TQM theories. TQM focuses more on improving thing by collaborative and cultural approaches while Six-Sigma is more of a data drive system. Total Quality Management is a comprehensive quality management system to achieve the market leadership for a company. This approach extends on all different areas of the organization as Six Sigma does. However, in contrast to Six Sigma, TQM is included in the daily work of the employees. The projects are not selected systematically and it is more an improvement step by step then as a radical change, as Six Sigma does. Also the main focus of total quality management system customer’s satisfaction & meeting customer demands and if there is no customer there is no company but opposite to this in six sigma the main focus is to generate large profits and meeting customer demand is seen as one way to generate more profit.

The theory of Six Sigma is strongly connected to different approaches of quality management systems. In their implementation it uses different tools of the given theories, but is still an own philosophy.

12.2.2 Implementation of TQM by implementing DOE.

Total quality management as a measure or tool for improving the organizational performance has captured the attention of many authors and researchers. It is agreed by many researchers that long term success of total quality management in an organization relies on its effective implementation. The process and tools like DOE through which total quality management is implemented in the organization plays major role for success. Thus, the success of total quality management owes a great deal to the effective implementation. For the effective implementation of the total quality management, organizational people should be very careful in selecting the Quality tools for correcting the problems and solving the errors without any kind of wastes.

Implementation of total quality management in the organizations is very important because of many reasons. Central to theses is that effective implementation of total quality management must have positive influence on the behaviors, attitudes and values of employees. Initial phase of implementation by selecting proper tools and techniques for the company’s total quality management has long lasting impact on its success.

13. APPLICATION OF DOE AND ITS ROLE:

This pictorial format shows us the various application and roles of DOE in the global. [3]

Figure : General applications of DOE [19]

14. ADVANTAGES AND DISADVANTAGES OF DOE:

Taguchi method is that it emphasizes a mean performance characteristic value close to the target value rather than specification limits. Additionally, Taguchi's method for experimental design is straightforward and easy to apply to many engineering situations, making it a powerful yet simple tool. It is mainly used to quickly narrow down the scope of a research project or to identify problems in a manufacturing process from data already in existence. For example, a process with 8 variables, each with 3 states, would require 6561 experiments to test all variables. However using Taguchi's orthogonal arrays, only 18 experiments are necessary, or less than .3% of the original number of experiments. In this way the parameters that have little effect can be ignored. [9, 15]

14.1. Disadvantages of DOE:

Taguchi method gives only relative and do not exactly indicate what parameter has the highest effect on the performance characteristic value. Also, since orthogonal arrays do not test all variable combinations, The Taguchi method has been criticized in the literature for difficulty in accounting for interactions between parameters. Another limitation is that the Taguchi methods are offline, Furthermore, since Taguchi methods deal with designing quality in rather than correcting for poor quality, they are applied most effectively at early stages of process development. They require a lot of human energy and resources and it takes a lot of ingenuity, cleverness, and experience to design experiments well. This most often lead to have a considerable artificiality for practice. [9,13,15,19]