Experimental Design for Optimization in Manufacturing

Minimizing cost and data request runs for expensive trials by utilizing full factorial experiment to identify significance, centerpoints for investigating significance of curvature, and central composite design to measure curvature. In depth investigation of effects, significance, transformations, Analysis of variance, etc. Summary shown here. Details are in the report.

Step 1: Run a 2k factorial design

Run 2k factorial design to get following points:

Isolate the significant features:

Track results of regression model:

Step 2: Run center points design

Run centerpoints design on signficiant variables to get following points:

Track results of regression model:

Step 3: Run a central composite design

Run centerpoints design on signficiant variables to get following points:

Track results of regression model: