Optimal Experiment Designs Combining Mixture and Process Inputs

This guide presents practical aspects for combining mixture, process, and categorical variables into optimal experiment designs. It outlines how to minimize experimental runs using a clever 'KCV' combined model and details options for structuring these designs into a split plot, enhancing their feasibility. The document discusses modeling mixture-process experiments, contrasting crossed models with KCV approaches, and includes case studies such as the optimization of fried fish patties and chocolate chip cookies. The guide explains the significance of split-plot designs, which originated in agriculture, and how they can be applied to experimental designs where some factors are more challenging or costly to vary. The KCV model is highlighted as an efficient method that reduces the number of terms needed while still capturing essential interactions between mixture components and process factors. Overall, this document serves as a comprehensive resource for deploying design of experiments (DOE) to optimize product and process improvements effectively.