Purpose: To determine appropriate levels of three factors to optimize two responses using the aspects of response surfaces, blocking and split plots. Summary: In this article, an experiment that involved a processing step that utilizes an oven in the production of a component for a manufactured product. In fact, it is planned to determine optimal settings of oven time and temperature so that the responses measures as tensile strength and diameter meet the customer requirements. By describing the process setup, it is observed that there exists a need for combining the aspects of response surface, blocking and split plot for the experiment. An equivalent model is considered useful for the experiment as it needs to take on o account the split- plot structure with two whole-plot factors and one sub-plot factor within blocks. Now, using the experimental data, the analyses with split-plot and whole plot structure are performed. Some of the model assumptions are also checked for the appropriateness of normality of error terms. Further, the current analysis including the residual analysis and multiple response optimization is illustrated.
展开▼