Mixture designs are constrained - factors can only take on values between 0 and 1, and the factors (components in the mixture) must sum to 1 (100%). The design space is a triangular rather than cubic (as you'd see in a factorial design). I believe that, because of these constraints, adding axial points from the augment platform for mixture designs is not available.
mia.stephens
anything else where i can get mixture of three factors and axial point to get triangle?
There are different types of mixture designs that can be generated from the DOE > Mixture Design platform. The Simplex Centroid and Simplex Lattice both place points on the edge of the design space. The Simplex Lattice also places design points in the interior of the design space.
In the ternary plots below, I generated the default designs, and then used the Graph > Ternary Plot option - you can see where the design points are placed relative to the triangle (i.e. the design space). The optimal option allows you to create optimal custom mixture designs. - this gives you flexibility in the size of the experiment and the terms (parameters) you need to estimate.
There's a nice description of these designs and the mixture design platform in the JMP Book DOE Guide (under the Help Menu, Books - search for Mixture).
I hope this helps!
Mia
Default Simplex Centroid - 3 factors
Default Simplex Lattice
An optimal mixture design (using the default settings):
But there are some mixture designs including axial runs. In this case, axial runs are located in the interior of the design space, on the line connecting each vertex to the overall centroid.
