Course Description
When formulations are developed there is one fact that can change the way standard DOEs are performed. This fact is the simple statement that the sum of the component proportions equals 100%. Because of this built-in dependency of the components, the design space will need to change, the model forms will need to change, and the analysis methods will need to reflect these new realities in design space and model forms.
Participants are prepared to understand the concept of a mixture and basic statistical models for fitting and analyzing mixture experiments. The focus is on application of mixture experiments to manufacturing or research operations involving chemical processes, powders, or other types of mixtures. Methods are developed for screening, characterizing, and optimizing mixture experiments involving basic designs such as simplex centroids, simplex lattice, and extreme vertices, along with process constraints.
Who Should Attend
This course is designed for engineers from manufacturing industries, chemists, or researchers dealing with processes involving chemical or mixture of components who need screen variables and develop optimal new formulations.
Learning Objectives
Through training, participants will:
- Be able to plan, conduct, and analyze a mixture experiment.
- Be able to understand the concept of mixtures and a simple mixture model
- Know how to select an appropriate mixture experimental design
- Know how to draw conclusions from mixture experiments
- Know how to use mixture designs for screening out important variables or for process optimization
Course Outline
Introduction to Mixture Experiments
- Review of Basic DOE Principles
- What is a Mixture Design?
- Types of Mixtures
- Fundamental Constraint and Consequences
- Factor Space
- Ternary Plots
- Mixture Models
- Interpretation of Model Coefficients
Simplex Lattice and Simplex Centroid Designs
- Simplex Lattice Designs
- Properties
- Cox Trace Plot
- Lack of Fit test
- Simplex Centroid Designs
- Properties
Complex Region Extreme Vertices Designs
- Complex Region Designs
- Actual vs. Real vs. Pseudo Components
- Piepels Trace Plot
- D-Optimality
Mixture Process Variable Experiments
- Why Mixture Process Variable Experiments are Necessary
- Types of Mixture Process Variable Experiments
- Design and Analysis
Prerequisite
Basic SPC or the equivalent
Course Format
8 hours
Instructor-led class training, with opportunities to practice learned skills using prepared data
Minitab, JMP, or Design Expert