DOE DECOUPLED MOLDING PART I: PROCESS CENTERING AND VALIDATION FROM THE PLASTIC'S POINT OF VIEW

摘要

Frequently, injection molders approach experimental designs from the perspective of the settings on the molding machine, sometimes resulting in sub-optimal or misleading results. Further, when the resulting process is validated, it is typically done under short term sources of variation. This frequently proves to be a poor evaluation of the true long-term capability of the process. A new approach to centering and validating processes is proposed and a model study is presented for illustration. Here, the experiment is conducted around the fundamental variables that affect the part inside the mold, using a "physical model" that acknowledges the unique physics of the injection molding process. Next, the data is correlated with fundamental process variables, allowing causal relationships to be established. Finally, during validation, primary sources of long-term variation are introduced into the capability study.