The aim of the mixture design is to find the optimal formulation. The mixture consists up of three components A, B, and C. The three components have lower and upper limits. However, it could be managed to keep the ranges of the components the same so a simplex centroid design with axial points could be defined. In addition, component C has three types C1, C2, and C3.
- Would it be recommended to use an extreme vertices design with the actual ranges instead of forcing the ranges to be equal to use the simplex design?
- Would it be possible to consider all the types of component C in one model: (A, B, C1, C2, C3)? Or one model per type i.e. three separate models would be necessary: (A, B, C1) ; (A, B, C2) ; (A, B, C3)?
