Deep Dive Courses


EXPERIMENT DESIGN

This workshop will cover the efficient design and analysis of experiments the purpose of which is to either 1) make comparisons or 2) establish cause-and-effect relationships.
Topics covered include:


  1. Good design practice: randomization, replication, blocking
  2. The analysis of variance for comparing means
  3. Factorial and fractional factorial designs
  4. Robust product design: incorporating noise factors and examining variation as a response

Nominal Outline for a three-day workshop:


  1. Conjecture
    1. Setting objectives
    2. Model
    3. Operational definitions
    4. Assumptions
  2. Design
    1. General topics
      1. Replication
      2. Repeated measures
      3. Randomization
      4. Blocking
      5. Covariates
    2. Specific designs
      1. Comparison
        1. One factor, two levels
        2. One factor, three or more levels
      2. Cause/effect
        1. Fractional factorial
          1. Aliasing
          2. Resolution
        2. Full factorial
          1. Main effects
          2. Interactions
          3. Probability plots of effects
        3. Introduction to Response surface
        4. Evolutionary operation
  3. Observation
    1. In control
    2. Out of control
  4. Analysis
    1. Residual plots
      1. Normal probability
      2. Time effects
      3. Stability of variance
    2. Statistics
      1. Half-normal probability plots of effects
      2. p-values
      3. Multiple comparisons
      4. Attribute data
    3. Interpretation
      1. Main effect plots
      2. Two-way interaction plots
      3. Practical significance


--Up--

PROCESS MODELING AND OPTIMIZATION

This workshop covers the design and analysis of experiments the purpose of which is to build empirical models that can be useful to model processes or optimize the process for one or more quality characteristics.


Nominal outline for a three-day workshop:

  1. Introduction
    1. The role of RSM
    2. Example of contour plots
    3. Review of Experimentation Concepts and Practice
  2. Statistical Underpinnings
    1. Correlation
    2. Regression
    3. Simple linear regression
    4. Diagnostics
      1. Overall measures or regression adequacy
        1. Standard error, s
        2. R2, adjusted R2 , and predicted R2
        3. Lack-of-fit
        4. Statistical significance
      2. Regression diagnostics from individual points
        1. Leverage
        2. Residuals
        3. Influence
    5. Multiple linear regression
    6. Comparing coefficients
  3. Response Surface Models and Designs – First Order
      1. Conjecture
      2. Design
      3. Observation
      4. Analysis
        1. Statistics
        2. Method of steepest ascent
  4. Response Surface Models and Designs – Second Order
    1. Conjecture
      1. Region of interest versus region of operability
      2. Second-order models
      3. Third-order and non-polynomial models
    2. Design
      1. 3k
      2. Central composite
      3. Box-Behnkin
      4. Choosing between central composite and B-B designs
      5. D-optimal
    3. Observation
    4. Analysis
      1. Estimation of analytical statistics
      2. Verification of assumptions
      3. Interpretation: contour plots
      4. Summary
  5. Drawing Contours and Surfaces
    1. Introduction
    2. Two factors - yield
    3. Three factors - adhesion
    4. Three factors - telescoping (messy data)
  6. Optimization
    1. Graphical optimization
    2. Numerical optimization
      1. Squared loss
      2. Desirability function
    3. Examples
      1. Two factors, one response - yield
      2. Two factors, two responses
      3. Three factors, four responses – tire tread
      4. Two factors - abrasion test method development



    STATISTICAL PROCESS CONTROL


    This workshop covers the fundamentals of the basic but critical techniques to successfully implement a program of statistical process control.

    Nominal outline for a three-day workshop:

    1. FRAMEWORK
      1. Four activities of quality
      2. SPC: Benefits and Limitations
      3. Problem solving guides
        1. PDCA
      4. Quality Improvement Story
        1. 8D
    2. SNAPSHOT TOOLS
      1. Conceptual Tools:
        1. Brainstorming
        2. Data
      2. Application Tools (snapshot):
        1. Pareto Analysis
        2. Flow Charts
        3. Cause and Effect Analysis
        4. Histograms
        5. Scatter Diagrams
    3. CONTROL CHART THEORY and X-BAR AND  R CHART CONSTRUCTION
      1. Theory
      2. Two uses of charts
      3. Kinds of charts
      4. Central Limit Theorem
      5. Chart mechanics
      6. Statistical Process Control vs. Engineering Process Control
      7. What to chart
      8. Construction of X-bar and R charts
      9. Flow chart of X-bar and R chart construction
      10. Process capability
      11. Alpha and beta risks
      12. Pattern analysis
      13. Zone control
    4. OTHER VARIABLES CHARTS
      1. Individuals charts
      2. Charts for processes with space frame variations
      3. Processes with drift
      4. S and sigma charts
      5. Cumulative sum charts
      6. Exponentially weighted moving average charts
      7. PRE-control
    5. ATTRIBUTE CHARTS
      1. p and np charts
      2. c and u charts
    6. MISCELLANEOUS CHARTING TOPICS
      1. The control chart as an operation
      2. Flow chart for charts
      3. Regular reviews of charts
      4. Short production runs
      5. Set-up variation
      6. Chart diagnostics



      PROCESS CAPABILITY

      This workshop examines two broad approaches to evaluating our process capability in support of continuous improvement. With the first approach we estimate total capability and discuss various ways of expressing capability: natural tolerance, capability indexes, percent out of specifications and "6 sigma" capability.

      With the second approach, we break down the components of variance to examine with which sources of variation our common cause variation is associated: material, equipment, operator, measurement, etc.

      Nominal outline for a three-day workshop:

      1. Introduction
        1. Motivation
        2. Process
        3. Process control
        4. Process capability
        5. Process capability versus designed experiments
      2. Total Process Capability
        1. Objectives and assumptions
        2. Verifying control
        3. Verifying normality
        4. Non-normal distributions
        5. Estimating the standard deviation
        6. Expressing capability
        7. Motivating a components of variance study
      3. Components of Process Capability
        1. Conjecture step
        2. Design step
          1. Crossed versus nested arrangements of factors
          2. Balanced nested designs
          3. Unbalanced nested designs
        3. Analysis step
          1. The ANOVA
          2. Whither variation?
          3. R & R studies
          4. Allocating resources between samples and tests
          5. Unbalanced designs
      4. Diagnostics
        1. Multi-vari plots
        2. s-charts
      5. Mixed models
        1. Fixed versus random effects
        2. Mixed models defined and examples



      TEST METHOD DEVELOPMENT AND EVALUATION

      With this workshop we explore various criteria by which we evaluate the usefulness of a test whether it is under development or pre-existing


      Nominal outline for a two-day workshop:


      1. Motivation
      2. `
        1. Rubber bands
        2.                   
      3. Overview
        1. Measurements
          1. What is a measurement?
          2. Why take measurements?
          3. Problems with poor measurements
            1. Poor resolution
            2. Test instability
            3. Lack of trust
        2. The Measurement Process
        3. Exercise
        4. Useful Test Methods
      4. Criteria for TM Understanding
        1. Sampling
          1. For conformance
          2. For control
          3. For Internal Reference Material
        2. Feature
          1. Relevance
          2. Operationally defined 
        3. Protocol
          1. Practical
            1. Well-documented
            2. Easily mastered
            3. Safe
          2. Rugged/reproducible
          3. Economical
        4. Datum
          1. Of sufficient resolution
          2. Precise
            1. Pure error
            2. Homoscedastic
              1. across samples
              2. across operators, machines, etc.
            3. Gage R&Rs
          3. Unbiased
          4. Discriminating
            1. relative to customer
            2. relative to product variations
            3. relative to specifications
          5. From a stable measurement process
      5. Student Lab
        1. Gage R&Rs
        2. A Basic EMP
        3. Analysis via software
      6. Comparing Test Methods
        1. On discrimination
        2. On economy
        3. Equivalence

When experience counts..


"Scripps & Associates, PC is the answer to your product and process development and improvement needs"

Thomas A. Scripps 1202 Bergen Parkway, Suite 208, Evergreen, CO 80439
PH: (303) 674-2530 FAX: (303) 674-4205 tom@tascripps.com
Incognito