Design of Experiments

This course presents techniques for planning studies where the inputs to a system/process can be varied and the outputs observed. Efficient planning and analysis methods for determining which inputs have statistically significant effect on outputs are covered, including analysis of variance, full and fractional factorial experiments, blocking, randomization, and robust design. The course includes hands-on experience using statapults for classroom experiments. Other classroom projects will use computer simulations set-up by the students and run by the instructor. Students may use a DOE statistical software package that is included with their course text if they are able to bring a laptop and power cord to class. The instructor will also demonstrate DOE solutions using the software package MINITAB. This course will cover lean six-sigma principles, output variance control, output mean control, and transfer function derivation. Students will be introduced to simulation techniques that accommodate inputs with varying quality distributions. Upon completion of this course, students will have the knowledge necessary to efficiently conduct experimental studies for comparative evaluation, input-output characterization, and process optimization. The course text, including a single output DOE software package, and the hardcopy course notes in a binder will be provided to each student as part of the course fee.

Course Outline

  1. The need for designed experiments
    • Basic principles
    • Hypothesis test and inference
    • Experimental error and models
    • Precision of estimates
    • Randomization
    • Replication and sample size
    • Blocking to improve precision
  2. Comparing several treatment means
    • One-way analysis of variance
    • Fixed model
    • Multiple comparisons
    • Model diagnostic checking
    • Variance components
    • The random model
  3. Randomized block and Latin square design
    • Randomized blocks
    • Multiple comparisons
    • Model diagnostics checking
    • The Latin square
  4. Factorial experiments
    • Main effects and interactions
    • The advantage of factorials
    • Two factors, fixed model
    • Random and mixed models
    • More than two factors
    • Special case
    • The 2^k factorial design
  5. Blocking in the 2 design
    • Two or more blocks
    • Guidelines for the choice of a design
  6. Fractional replication of the 2^k design
    • The one-half fraction
    • The one-quarter fraction
    • Other fractions
    • Guidelines for the choice of a design
  7. Nested designs
    • The two-stage nested design
    • More than two stages
    • A design with nested and factorial factors
  8. Robust design
    • Control and noise variables
    • Quality loss and signal to noise ratios
    • Orthogonal array and interaction table
    • Variation reduction and process optimization
  9. Process control
    • Six sigma strategies
    • Variance control
    • Control of output means

Instructor

Dr. Steven (Flash) Gordon is the director of TEREC and the Orlando Field Office Manager for Georgia Tech Research Institute. He served over 26 years in uniform in the United States Air Force (USAF) in various assignments, including as an aircrew member and instructor in the F-111, Director of Air Operations Centers, Division Chief on the Air Staff, and Professor and Head of the Department of Mathematical Sciences at the USAF Academy. At the Academy, he taught pre-calculus, calculus I-III, linear algebra, probability and statistics, systems of differential equations, and design of experiments. He also served over 4 years as the Technical Director for the Air Force Agency for Modeling and Simulation. He has a BS in Mathematics from Marymount College of Kansas, MS in Human Development Counseling/Education from Peabody College of Vanderbilt University, MS in Industrial Engineering/Operations Research from Purdue University, and a PhD in Aeronautical and Astronautical Engineering from Purdue University. His PhD research topic was orbit determination and optimal control of Earth-Moon-Sun libration point spacecraft orbits, and this research was extended to include experimental design for libration orbit control cost excursions.

Continuing Education Credit

This program meets the criteria for the nationally accepted Continuing Education Unit (CEU). Each participant successfully completing this 3 day course will earn 1.65 CEUs. This course is a required course for the TEREC Test & Evaluation Certificate.

Schedule and Fees

This course is currently scheduled for October 20-22, 2009 in Atlanta, GA. The course fee of $1250 covers all course materials.

Course Location and Accommodations

The course will be held at the The Georgia Tech Global Learning Center, located in Technology Square at 84 Fifth Street. This facility is directly adjacent to the Georgia Tech campus in downtown Atlanta.

The nearest accommodations are available at the Georgia Tech Hotel, located in Technology Square at 800 Spring Street, Atlanta, Georgia, 30308. Mention that you are attending a Georgia Tech program for a special room rate. For hotel reservations, call 1-800-706-BUZZ (2899). Special rate and rooms are based upon availability.

Registration

Registration for this course is done through the Georgia Tech Continuing Education Department. To register on-line, use the Continuing Education Information/Registration Page for this course. Once there, click on "Register Now." A phone number is provided for those who do not wish to register and pay on-line.

Course Last Offered

October 7-9, 2008, Atlanta, GA

Last Updated May 1, 2009