In injection molding, molds equipped with multiple cavities allow for the simultaneous molding of several items. However, this multi-cavity approach often leads to uneven resin filling across cavities, raising the likelihood of defects. This project aimed to verify whether our developed molding defect detection algorithm could be effectively applied to detect defects in high-volume injection molding.
We demonstrated the ability to detect defects occurring in individual cavities by measuring the in-mold pressure during molding using pressure sensors attached to a multi-cavity mold and analyzing the time-series waveforms of the data.
An experiment was conducted using a multi-cavity injection mold with four cavities, molding polyester resin, and analyzing the pressure waveforms during molding with our defect detection algorithm to detect short shots and scratches.
The analysis using our defect detection algorithm showed that when short shots and scratches occurred, the values of indicators representing the quality of the molded items significantly deviated from those of good products, confirming the ability to detect these defects.
This initiative has confirmed the effectiveness of our developed molding defect detection algorithm in high-volume injection molding. Moving forward, we will expand the application of defect detection across a wider variety of materials, molded items, and types of defects.
