A statistical experimental design, with specific information about the experiments, needs to be provided in advance to reach an optimized result.
This process is called DoE – Design of Experiments.
In order to save resources, statistical methods are used to reduce the effort between carrying out tests in relation to the accuracy of the obtained results.
For this one or more aims are set. In injection molding, this might be cycle time reduction, lower warpage, …
The changeable factors (injection time, melt temperature, packing pressure…) and their allowed range will be set as well.
On this base, an experimental design is created. It describes which experiments are to be carried out in which order.
The results of each experiment are determined for a final evaluation.
After the DoE, the information about how much each factor influences the aims is displayed. Furthermore, the combination of all the factors which comes closest to the requested aim is shown.
Today, software is used to carry out DoE, as these provide more cost-effective results than experiments in practice.