Select the type of blends to include in the design and how many different settings to use for each type of blend. The higher the degree, the more complex the model the design can fit. The following table shows simplex designs of varying degrees with four component designs. The number of blends become more complicated with extreme vertices designs.

For more information about the types of models up to degree 3, go to What is the degree of a lattice?.

Degree | Number of blends |
---|---|

1 | Contains only vertex points. |

2 | Contains double blends where the components are mixed equally. |

3 | Two double blends for each pair of components and triple blends that are mixed equally. |

4 | Three double blends for each pair of components and two triple blends for each trio of components. |

Select this option to include a point in the center of the design space. Center points have multiple uses. Designs with process variables or replicates increase the number of center points in the final design. For more information on augmentation, go to How Minitab augments mixtures designs by adding center points.

Select this option to include 1 axial point per component in the base design. An axial point is a blend where the component proportions are the average of the center point and a vertex. A design with these interior points provides more information about the inner portion of the response surface to model more complicated curvature. For diagrams that show examples of unaugmented and augmented simplex designs with three components, go to Choose a mixture design.

Specify how many times to perform each experimental run that is in the base design. Usually, you consider the available resources and the purpose of your design when you select the number of replicates. You can add replicates to your design later with . For more information on replicates, go to Replicates and repeats in designed experiments.

- Number of replicates for the whole design: Select the number of replicates for all of the points in the design. Replicates are multiple experimental runs with the same component settings. (If the design has process variables, these settings are the same too.) Usually, you consider the available resources and the purpose of your design when you choose the number of replicates.
- Number of replicates for the selected types of
points: Select this option to replicate some types of points in the design instead of all points. For example, you can replicate center points to estimate pure error when you do not have enough resources to replicate other types of points. The Point
Type column shows the value that identifies that type of point when you store the design in the worksheet. Enter the number of replicates for each point type in the Number column.
- vertex
- Points where one component is at its upper bound and all others are at their lower bound.
- blends
- Points with two or more components not at their lower bounds. For example, a double blend has two non-zero component proportions and 0 for any other component proportions. A triple blend has 3 non-zero component proportions and 0 for any other component proportions.
###### Note

For simplex lattice designs and extreme vertices designs, replication of blend points adds experimental runs only if the degree of the design includes the blend in the base design.

- center point
- Points whose component proportions are all equal. The base design has 1 center point.
- axial point
- Points whose component proportions are averages of the proportions of the center point and a vertex. Axial points have no component proportions set to their lower bound. In the base design, the number of axial points is the same as the number of vertex points. Replication of these points adds experimental runs only if you selected to augment the design with axial points.