Industrial Design of Experiments: A Case Study Approach for Design and Process Optimization

Other Books by the Author
Preface
About the Book´s Organization
Acknowledgments
Contents
List of Figures
List of Tables
About the Author
Chapter 1: Data Presentations, Statistical Distributions, Quality Tools, and Relationship to DoE
1.1 Graphical Presentation of Data
1.2 Probability Distributions and Their Use in Modern Quality Systems
1.2.1 The Binomial Distribution
Example of Calculating Binomial Distribution Function
Examples of Using the Binomial Distribution
1.2.2 The Poisson Distribution
Examples of Using the Poisson Distribution
1.2.3 Continuous Distributions and Reliability
Weibull Distribution Example
1.2.4 The Use of the Standard Normal Distribution in Quality Methods
Examples of Using SND for Determining Defect Rates
1.3 Six Sigma and Its Relationship with DoE
1.3.1 Converting Defect Rate to Cp and Sigma Design with No Mean Shift from Nominal
1.3.2 The Implied Mean Shift to Nominal of Cp and Six Sigma
1.3.3 Process Capability Studies for Quality Enhancement
1.3.4 Process Capability for Prototype and Early Production Parts
1.3.5 Corrective Action for Process Capability Problems
1.3.6 DoE Effects on Six Sigma
1.4 Control Charts and Their Relationship with DoE
1.4.1 Selection of Control Charts
1.4.2 Variable Control Charts
1.4.3 Relationship of Control Limits to Specification Limits
and R Variable Control Chart Calculation Example
1.4.4 Variable Control Chart Usage Guidelines
Example of Variable Control Charts´ Relationship to Process Capability
Variable Control Chart Solution
1.4.5 Control Charting and Process Capability for Low-Volume Production
Examples of Moving Range Calculations
1.4.6 Attribute Control Charts
Example of Attribute Control Chart Calculations and Relationship with Process Capability
Example of Variable Control Chart Solution
1.4.7 Use of Control Charts in Factories with High Process Capability
1.4.8 DoE Effects on Control Charts
1.5 Conclusions
References
Additional Reading Material
Chapter 2: Samples and Populations: Statistical Tests for Significance of Mean and Variability
2.1 Sample and Population Statistics
2.1.1 Standard Deviation Estimation Methodologies and Data Collection
2.1.2 Measurement System Error (GR&R) and Its Impact on Statistical Measurements
GR&R Study Example
Answer of GR&R Sample Study
2.2 Tests for Sample and Population
2.3 Tests for Means
2.3.1 z-Test for Population Means
2.3.2 The Wilcoxon Test for Non-normal Population Mean Test
Wilcoxon Test Example
2.3.3 t-Tests for Sample Means: Single Sample
2.3.4 t-Tests for Comparing Two Sample Means with Unknown Variance
2.3.5 d- or Paired t-Test
2.3.6 Confidence Interval (CI) of the Mean
2.3.7 Determination of Sample Size Based on Error
Sample Size for Specified Error Example
2.4 Tests for Variability
2.4.1 X2 (Chi-Square) Significance Test
X2 (Chi-Square) Test Example
2.4.2 X2 Goodness of Fit Test and Checking for Normality
2.4.3 F-Test
2.5 Conclusions
References
Additional Reading Material
Chapter 3: Regression, Treatments, DoE Design, and Modelling Tools
3.1 Regression Analysis
3.1.1 Least Squares Regression
3.1.2 Linear Regression Analysis Using Model Coefficients Estimates
Linear Regression Analysis Example
3.1.3 Linear Regression Analysis Using ANOVA
Linear Regression Analysis Example Using ANOVA
3.1.4 R2 and Accuracy of Model Estimate
3.1.5 Using Linear Regression for Normality Checking
3.2 Treatment Design and Analysis
3.2.1 Treatment Design and Analysis Example
3.2.2 Significance Determination Techniques and p% Contribution
3.3 Full Factorial DoE Design and Analysis
3.3.1 Limiting DoE Scope with Design Space and Process Map
3.3.2 Full Factorial DoE Design Analysis Using Interactions
3.4 Full Factorial DoE Design and Analysis Case Study: Green Electronics Manufacturing
3.4.1 Summary of Phase I Green Electronics DoE Case Study
3.4.2 Phase II of Green Electronics DoE Case Study
3.4.3 Analysis of Phase II DoE
3.5 Conclusions
Additional Reading Material
Chapter 4: Two-Level Factorial Design and Analysis Techniques
4.1 DoE Definitions, Expectations, and Processes
4.1.1 DoE Lifecycle Process
4.1.2 DoE Project Timing and Error Source
4.2 Two-Level Factorial DoE Design
4.2.1 Commonly Used Two-Level Orthogonal Arrays
4.2.2 Types of Uses for Two-Level OA
4.2.3 Interaction, Confounding, and Interconnecting Graphs
4.3 Two-Level OA Analysis and Model Reductions
4.4 Two-Level DoE Case Studies
4.4.1 Full Factorial L8 Case Study, Hipot DoE: Selecting Best Alternative Among Equally Performing Designs
4.4.2 Partial Factorial L8 Case Study, Underfill Voids DoE: Selecting Process Parameters for Zero Defects
4.4.3 Partial Factorial L16 Example Case Study, Rivet Design DoE: Selecting Part Dimension Design for Best Product Performance
4.4.4 Partial Factorial L32 Case Study, APOS for Robotics DoE: Selecting Process Parameters for Multiple Adjustment Production...
4.5 Conclusions
Additional Reading Material
Chapter 5: Three-Level Factorial Design and Analysis Techniques
5.1 Three-Level Factorial Design
5.1.1 Commonly Used Three-Level Orthogonal Arrays
Use of Three-Level OA: Full and Partial Factorial and Screening Modes
5.1.2 Use of Three- Versus Two-Level DoE
5.2 Three-Level DoE Analysis and Model Reductions
5.3 Three-Level DoE Case Studies
5.3.1 Screening Design L9 Case Study: Bonding I DoE
5.3.2 Screening Design L9 Case Study: Zero-Defect Mixed Soldering DoE
5.3.3 Partial Factorial DoE L27 Case Study: Green Electronics Phase I DoE
5.3.4 Screening Design Software L27 Case Study: Minimizing Half-Adder Chip Delay Time DoE
5.4 Conclusions
Additional Reading Material
Chapter 6: DoE Error Handling, Significance, and Goal Setting
6.1 DoE Error Handling Techniques for Significance Testing
6.1.1 Regression Equation and Predicted Outcome with Interaction
6.1.2 DoE Error Handling Types
6.1.3 Error Handling and Significance Technique L8 Case Study: Plastics Injection Molding DoE
6.1.4 Error Handling and Significance for Single Repetition DoE Analysis
6.1.5 Error Handling and Significance for Multiple Replication DoE Analysis
6.1.6 Error Handling and Significance for Multiple CenterPoint Replications
6.1.7 Error Handling and Significance for Some Experiment Outcome Replications
6.2 Project Goal Setting and Design Space
6.2.1 Types of DoE Project Goals
6.2.2 Design Space and Level Selection
6.3 Experiment Blocking (Dividing)
6.4 Conclusions
Additional Reading Material
Chapter 7: DoE Reduction Using Confounding and Professional Experience
7.1 Design Resolution and Confounding
7.1.1 Techniques for Managing Confounding
7.2 Interactions and Confounding for L8 for Reduced Experiments
7.2.1 L8 Half Fraction Interaction and Confounding
7.2.2 L8 Partial Factorial Design and Confounding
7.2.3 L8 Screening Design Confounding
7.2.4 L8 Factor Conversion Tables for Labeling Numeric and Alphabetic Factors
7.3 Interactions and Confounding for L16 for Reduced Experiments
7.3.1 L16 Half Fraction Interaction and Confounding
7.3.2 Interactions and Confounding in L16 Rivet DoE Case Study
7.3.3 L16 Partial Factorial Design and Confounding
7.3.4 L16 Partial Factorial Design and Confounding with DoE Interaction Matrix Method
7.3.5 Interaction Matrix L32 Case Study: Solder Wave DoE Design
7.3.6 Resolving Confounded Interactions
7.3.7 L16 Partial Factorial Design and Confounding for Eight or More Factors
7.4 Interaction and Confounding for Large OA
7.5 Conclusions
Additional Reading Material
Chapter 8: Multiple-Level Factorial Design and DoE Sequencing Techniques
8.1 Multi-level OA Arrangements
8.1.1 Multi-level Arrangement DoE L8 Case Study: Machining I Pin Fin Heat Sinks
8.1.2 Multi-level Arrangement DoE L16 Case Study: Machining II Stencil Forming
8.2 DoE Sequencing Techniques
8.2.1 Foldover Sequencing Techniques: Folding on One Factor
8.2.2 Foldover Sequencing Techniques: Folding on All Factors
8.2.3 DoE Sequencing Technique Case Study: Printer Design DoE
8.3 Non-interacting Orthogonal Array Use in DoE
8.3.1 Non-interacting OA Case Study I: L18 Painting DoE
8.3.2 Non-interacting OA Case Study II: L36 Air Knife DoE
8.4 Conclusions
Additional Reading Material
Chapter 9: Variability Reduction Techniques and Combining with Mean Analysis
9.1 Controlled and Noise Factors in DoE
9.2 Variability Reduction Definitions and Analysis
9.3 Balancing Mean and Variability Outcomes
9.4 Conclusions
9.4.1 Controlled and Noise Factors in DoE
Reducing Noise Factor Repetitions
9.4.2 Variability Reduction Definitions and Analysis
Mean and Variability Analysis L8 Case Study I, Larger Is Better: Bonding II DoE of Neoprene to Steel
Average and Variability Analysis L18 Case Study II, Smaller Is Better: Painting DoE
Average and Variability Analysis L9 Case Study III, Target Is Best: Stencil Screening II DoE
9.4.3 Balancing Mean and Variability Outcomes
Balancing Mean and Variability Gold Plating Example
9.4.4 Conclusions
Additional Reading Material
Chapter 10: Strategies for Multiple Outcome Analysis and Summary of DoE Case Studies and Techniques
10.1 Summary of Previous Chapters
10.2 Combining Multiple Desired Outcomes with Mean and Variability Analysis
10.2.1 Interaction Matrix L32 Case Study: Solder Wave DoE Analysis
10.3 Summary of DoE Case Studies and Techniques
10.3.1 DoE Case Studies List by OA Size for Two and Three Levels
10.3.2 DoE Techniques Used and Demonstrated in Chapters and Case Studies
10.4 Conclusions
Index