Mixture Design

Design for recipes whose components sum to 100 % — proportions cannot vary independently.

Also known as: mixture design, Simplex-Lattice, Simplex-Centroid

A mixture design is used when the factors are component proportions of a recipe (wt %, vol %, mole fractions) which by definition sum to 100 %. The usual full factorial or CCD thinking fails here because the components cannot vary independently.

  • Simplex-Lattice design — grid on the simplex; combinations like (1, 0, 0), (½, ½, 0), (⅓, ⅓, ⅓)
  • Simplex-Centroid design — pure components plus all binary, ternary, …, blends at equal share
  • Extreme-Vertices design — with lower/upper limits per component, for constrained recipes

See also

Used in

Sources

  • Cornell, J. A. — Experiments with Mixtures: Designs, Models, and the Analysis of Mixture Data, 3rd Edition, Wiley
  • Scheffé, H. — Experiments with Mixtures, Journal of the Royal Statistical Society B, 20(2), 1958
  • Montgomery, D. C. — Design and Analysis of Experiments, 10th Edition, Chapter 11