Design of Experiments for Engineers and Scientists

Description

This third edition of Design of Experiments for Engineers and Scientists adds to the tried and trusted tools that were successful in so many engineering organizations with new coverage of design of experiments (DoE) in the service sector. Case studies are updated throughout, and new ones are added on dentistry, higher education, and utilities. Although many books have been written on DoE for statisticians, this book overcomes the challenges a wider audience faces in using statistics by using easy-to-read graphical tools. Readers will find the concepts in this book both familiar and easy to understand, and users will soon be able to apply them in their work or research. This classic book is essential reading for engineers and scientists from all disciplines tackling all kinds of product and process quality problems and will be an ideal resource for students of this topic. Written in nonstatistical language, the book is an essential and accessible text for scientists and engineers who want to learn how to use DoEExplains why teaching DoE techniques in the improvement phase of Six Sigma is an important part of problem-solving methodologyNew edition includes two new chapters on DoE for services as well as case studies illustrating its wider application in the service industry Front Cover Design of Experiments for Engineers and Scientists Copyright Page Dedication Contents About the author Preface Acknowledgements 1 Introduction to industrial experimentation 1.1 Introduction 1.2 Some fundamental and practical issues in industrial experimentation 1.3 Statistical thinking and its role within DOE Exercises References 2 Fundamentals of design of experiments 2.1 Introduction 2.2 Basic principles of DOE 2.2.1 Randomisation 2.2.2 Replication 2.2.3 Blocking 2.3 Degrees of freedom 2.4 Confounding 2.4.1 Design resolution 2.4.2 Metrology considerations for industrial designed experiments 2.4.3 Measurement system capability 2.4.4 Some tips for the development of a measurement system 2.5 Selection of quality characteristics for industrial experiments Exercises References 3 Understanding key interactions in processes 3.1 Introduction 3.2 Alternative method for calculating the two-order interaction effect 3.3 Synergistic interaction versus antagonistic interaction 3.4 Scenario 1 3.5 Scenario 2 3.6 Scenario 3 Exercises References 4 A systematic methodology for design of experiments 4.1 Introduction 4.2 Barriers in the successful application of DOE 4.3 A practical methodology for DOE 4.3.1 Planning phase 4.3.1.1 Problem recognition and formulation 4.3.1.2 Selection of response or quality characteristic 4.3.1.3 Selection of process variables or design parameters 4.3.1.4 Classification of process variables 4.3.1.5 Determining the levels of process variables 4.3.1.6 List all the interactions of interest 4.3.2 Designing phase 4.3.3 Conducting phase 4.3.4 Analysing phase 4.4 Analytical tools of DOE 4.4.1 Main effects plot 4.4.2 Interactions plots 4.4.3 Cube plots 4.4.4 Pareto plot of factor effects 4.4.5 NPP of factor effects 4.4.6 NPP of residuals 4.4.7 Response surface plots and regression models 4.5 Model building for predicting response function 4.6 Confidence interval for the mean response 4.7 Statistical, technical and sociological dimensions of DOE 4.7.1 Statistical dimension of DOE 4.7.2 Technical dimension of DOE 4.7.3 Sociological and managerial dimensions of DOE Exercises References 5 Screening designs 5.1 Introduction 5.2 Geometric and non-geometric P–B designs Exercises References 6 Full factorial designs 6.1 Introduction 6.2 Example of a 22 full factorial design 6.2.1 Objective 1: Determination of main/interaction effects that influence mean plating thickness 6.2.2 Objective 2: Determination of main/interaction effects that influence variability in plating thickness 6.2.3 Objective 4: How to achieve a target plating thickness of 120 units? 6.2.3.1 Effect of plating time on plating thickness 6.2.3.2 Interaction effect between plating time and plating solution temperature (AB) 6.3 Example of a 23 full factorial design 6.3.1 Objective 1: To identify the significant main/interaction effects that affect the process yield 6.3.2 Objective 2: To identify the significant main/interaction effects that affect the variability in process yield 6.3.3 Objective 3: What is the optimal process condition? 6.4 Example of a 24 full factorial design 6.4.1 Objective 1: Which of the main/interaction effects affect mean crack length? 6.4.2 Objective 2: Which of the main/interaction effects affect variability in crack length? 6.4.3 Objective 3: What is the optimal process condition to minimise mean crack length? 6.4.4 More examples of FFEs Exercises References 7 Fractional factorial designs 7.1 Introduction 7.2 Construction of half-fractional factorial designs 7.3 Example of a 2(7−4) factorial design 7.4 An application of 2-level fractional factorial design 7.4.1 Example of a 2(5−1) factorial design 7.4.2 Objective 1: To identify the factors which influence the mean free height 7.4.3 Objective 2: To identify the factors which affect variability in the free height of leaf springs 7.4.4 How do we select the optimal factor settings to minimise variability in free height? 7.4.5 Another example of a 2(5−1) factorial design 7.4.6 Example of a 2(7−4) factorial design 7.4.7 Another example of a 2(7−4) factorial design Exercises References Further reading 8 Some useful and practical tips for making your industrial experiments successful 8.1 Introduction 8.1.1 Get a clear understanding of the problem 8.1.2 Project selection 8.1.2.1 Management involvement and commitment 8.1.2.2 Return on investment 8.1.2.3 Project scope 8.1.2.4 Time required to complete the project 8.1.2.5 Value to your organisation 8.1.3 Conduct exhaustive and detailed brainstorming sessions 8.1.4 Teamwork and selection of a team for experimentation 8.1.5 Select the continuous measurable quality characteristics or responses for the experiment 8.1.6 Choice of an appropriate ED 8.1.7 Iterative experimentation 8.1.8 Randomise the experimental trial order 8.1.9 Replicate to dampen the effect of noise or uncontrolled variation 8.1.10 Improve the efficiency of experimentation using a blocking strategy 8.1.11 Understanding the confounding pattern of factor effects 8.1.12 Perform confirmatory runs/experiments Exercises References 9 Case studies 9.1 Introduction 9.2 Case studies 9.2.1 Optimisation of a radiographic quality welding of cast iron 9.2.1.1 Objective of the experiment 9.2.1.2 Selection of the response function 9.2.1.3 List of factors and interactions of interest for the experiment 9.2.1.4 Levels of parameters and their ranges 9.2.1.5 Choice of design and number of experimental trials 9.2.1.6 Design generators and the confounding structure of the design 9.2.1.7 Uncoded design matrix with response values 9.2.1.8 Analysis and interpretation of results 9.2.1.9 Confirmatory trials 9.2.2 Reducing process variability using experimental design technique 9.2.2.1 Objective of the experiment 9.2.2.2 Selection of the response 9.2.2.3 List of process parameters and their levels 9.2.2.4 Choice of design and number of experimental trials required for the experiment 9.2.2.5 Design generators and resolution 9.2.2.6 Coded and uncoded design matrix with response values 9.2.2.7 Analysis and interpretation of results 9.2.2.8 Determination of optimal settings to minimise variability 9.2.2.9 Confirmation trials 9.2.2.10 Significance of the work 9.2.3 Slashing scrap rate using fractional factorial experiments 9.2.3.1 Nature of the problem 9.2.3.2 Objective of the experiment 9.2.3.3 Selection of the response 9.2.3.4 List of process parameters and their levels 9.2.3.5 Coded design matrix with response values for the experiment 9.2.3.6 Analysis and interpretation of results 9.2.3.7 Confirmation runs 9.2.4 Optimising the time of flight of a paper helicopter 9.2.4.1 Objective of the experiment 9.2.4.2 Description of the experiment 9.2.4.3 Selection of the response 9.2.4.4 List of design parameters and their levels 9.2.4.5 Choice of design and design matrix for the experiment 9.2.4.6 Statistical analysis and interpretation of results 9.2.4.7 Determination of optimal design parameters 9.2.4.8 Predicted model for time of flight 9.2.4.9 Confirmatory runs 9.2.4.10 Significance of the work 9.2.5 Optimising a wire bonding process using DoE 9.2.5.1 Objective of the experiment 9.2.5.2 Description of the experiment 9.2.5.3 Selection of the response 9.2.5.4 Identification of Process Variables for Experimentation 9.2.5.5 Choice of design and experimental layout 9.2.5.6 Statistical analysis and interpretation 9.2.5.7 Model development based on the significant factor/interaction effects 9.2.5.8 Conclusion 9.2.6 Training for DoE using a catapult 9.2.6.1 Objective of the experiment 9.2.6.2 Selection of response 9.2.6.3 List of factors and their levels used for the experiment 9.2.6.4 Choice of design and experimental layout for the experiment 9.2.6.5 Statistical analysis and interpretation of results 9.2.6.6 Determination of optimal factor settings 9.2.6.7 Confirmatory experiment 9.2.6.8 Significance of the work 9.2.7 Optimisation of core tube life using designed experiments 9.2.7.1 Company’s first attempt to experimental approach 9.2.7.2 Company’s second attempt to use designed experiments 9.2.7.3 Choice of experimental layout for the experiment 9.2.7.4 Statistical analysis and interpretation 9.2.7.5 Determination of the optimal process parameter settings 9.2.7.6 Confirmation trials 9.2.7.7 Significance of the study 9.2.8 Optimisation of a spot welding process using DoE 9.2.8.1 Interactions of interest 9.2.8.2 Statistical analysis of experimental results 9.2.8.3 Loss-function analysis for larger-the-better characteristics 9.2.8.4 Significance of the study 9.2.9 DoE applied to a fizz-flop experiment 9.2.9.1 Hypotheses 9.2.9.2 Experimental plan 9.2.9.3 Execution of experiment 9.2.9.4 Data collection, analysis and interpretation 9.2.9.4.1 Data collection 9.2.9.4.2 Analysis of data 9.2.9.4.3 Experimental conclusions 9.2.9.4.4 Key lessons learned 9.2.9.4.5 Significance of the study 9.2.10 DoE applied to a higher education context 9.2.10.1 Significance of the study 9.2.11 DoE applied to a transactional process 9.2.11.1 Data analysis 9.2.12 DoE applied to a banking operation 9.2.13 DoE applied to a transactional process 9.2.13.1 Significance of the study 9.2.14 Design of experiments in understanding and evaluating teaching effectiveness in UK higher education 9.2.14.1 Phase 1: Planning of the experiment 9.2.14.2 Phase 2: Designing the experimental layout 9.2.14.3 Phase 3: Conducting the experiment 9.2.14.4 Phase 4: Analysing the experiment 9.3 Discussion and limitations of the study References Further reading 10 Design of experiments and its applications in the service industry 10.1 Introduction to the service industry 10.2 Fundamental differences between the manufacturing and service organisations 10.3 DOE in the service industry: fundamental challenges 10.4 Benefits of DOE in service/non-manufacturing industry 10.5 DOE: case examples from the service industry 10.5.1 Data entry errors 10.5.2 Debt collection 10.5.3 Emergency department performance 10.6 Role of computer simulation models within DOE Exercises References 11 Design of experiments and its role within Six Sigma 11.1 What is Six Sigma? 11.2 How Six Sigma is different from other quality improvement initiatives of the past 11.3 Who makes Six Sigma work? 11.3.1 Six Sigma deployment champions 11.4 Six Sigma methodology (DMAIC methodology) 11.4.1 Define phase 11.4.2 Measure phase 11.4.3 Analyse phase 11.4.4 Improve phase 11.4.5 Control phase 11.5 DOE and its role within Six Sigma Exercises References 12 Design of Experiments in the service industry: a critical literature review and future research directions 12.1 Introduction 12.2 Methodology 12.3 Key findings 12.3.1 Experimentation environment and number of replications 12.3.2 Design of Experiments strategies and designs 12.3.3 Number of factors, levels and quality characteristics 12.3.4 Critical success factors 12.3.5 Essential skills required for professionals 12.3.6 Key lessons learned from designed experiments 12.4 Discussion and implications 12.5 Limitations and future directions of research References 13 Design of Experiments in the service industry: results from a global survey and directions for further research 13.1 Introduction 13.1.1 Research methodology 13.1.1.1 Development of survey instrument and piloting the instrument 13.1.1.2 Sampling strategy and data collection 13.1.2 Key findings 13.1.2.1 Demographical information 13.1.2.2 Education, training and experience in Design of Experiments 13.1.2.3 Challenges in applying Design of Experiments in the service industry 13.1.2.4 Critical success factors for applying Design of Experiments in the service industry 13.1.2.5 Essential skills for successful application of Design of Experiments in the service industry 13.1.3 Discussion and implications 13.1.4 Limitations and directions for future research Appendix A Statements related to the challenges in applying Design of Experiments (DoE) in the service industry References Index Back Cover