Design of Experiments (DOE) with Minitab – Advanced

Overview of the Course
This course explores how experimental design, an important off-line improvement tool, can be applied to the product and process optimization phase of development. Specifically, techniques are presented that can be used to achieve products’ and processes’ optimal performance.   This optimality is accomplished by using a strategy known as Response Surface Methods.

Participants will learn how to use their Design of Experiments knowledge to build models allowing the estimation of curvature and how to find a theoretical optimum based on this modeled curvature.  They will also learn how to simultaneously optimize multiple responses.  Desirability functions are employed to strike a balance between multiple responses through a suitable choice of input parameters.

Participants will have an opportunity to use his/her newly acquired knowledge to actually design, run and analyze a series of experiments via computer simulations.

This course does not teach participant on just how to point and click. The basic concepts underlying each tool are discussed before the use of the software is demonstrated.

Participants     Managers and Engineers from all engineering disciplines including Process, Research and Development, Quality, Maintenance and Reliability.

Duration – 2 days

Prerequisites   Knowledge of Full Factorial and Fractional Factorial (Screening) Designs.

Deliverables  

  • Understanding Response Surface Methods Strategy for Process Optimization.
  • Ability to develop statistical models to describe the relationship between Response Variables and Critical Process/Product Factors
  • Ability to simultaneously optimize Several Response Variables.

Course Outline
Experimental Design and Regression Analysis Fundamentals Review

  • Factorial Design and Fractional Factorial Designs Concepts
  • Blocked Designs
  • Regression Analysis

Introduction to Response Surface Methods (RSM)

  • Experimentation as an Iterative Process
  • Goals and Strategy of RSM
  • Case Study

First Order Designs

  • Experimental Designs for Fitting First Order Models
  • Use of Center Points to Test Curvature
  • Path of Steepest Ascent (POA)

Second Order Designs

  • Experimental Designs for Fitting Second Order Models: Central Composite
  • Designs, Box-Behnken Designs
  • Blocking in RSM Designs

Multiple Response Optimization

  • Desirability Function
  • Optimization of Process/Product involving Several Response Variables

Live Experimentation

 

Always Keep Improving!