In the field of injection molding, aside from shape defects like short shots and burrs, aesthetic flaws such as silver streaks and blisters can also occur. Silver streaks are a defect characterized by the appearance of white lines on the surface of the molded product, primarily caused by insufficient drying of the resin material or excessive dwell time inside the cylinder. Blisters are formed as bubble-like defects on the surface of the molded product, arising from shear heating or mold temperature effects that elevate the resin temperature beyond its optimal range. Traditional methods using internal pressure sensors have found it challenging to detect the occurrence of silver streaks and blisters.
The research team has successfully demonstrated the capability to detect aesthetic defects using their unique analytical techniques.
(1) Silver Streaks
As illustrated in Figure 1, internal pressure sensors were installed in the mold, and pressure data for each cycle during injection molding were collected. In the experiment, 24 shots of defect-free products were molded, and the occurrence of silver streaks under the same molding conditions was confirmed. Furthermore, by intentionally prolonging the dwell time of the resin material in the cylinder, silver streaks were induced. Under these conditions, a total of 37 shots worth of internal pressure data were collected.
(2) Blisters
Similarly, as depicted in Figure 1, internal pressure sensors and resin temperature sensors were installed in the mold, and data for each cycle were obtained. Under identical molding conditions, data for 25 shots of good quality and 18 shots where blisters occurred were gathered.
(1) Silver Streaks
Based on the data obtained from the internal pressure sensors, feature extraction and dimension reduction were performed to determine the centroid of the good quality data set of 24 shots. Figure 2 represents the plot of dimension-reduced feature values of each shot against the distance from the centroid of the good quality data set. Blue dots represent good quality shots, and orange dots indicate shots where silver streaks occurred. It was observed that shots with silver streaks showed a relatively greater distance from the centroid of the good quality data set. By setting the red line in the figure as a threshold, shots exceeding this threshold could be identified as silver streaks.
(2) Blisters
Based on the data from the pressure and resin temperature sensors, feature extraction and dimension reduction were carried out to find the centroid of the good quality data set of 25 shots. The subsequent figure shows the plot of dimension-reduced feature values of each shot against the distance from the centroid of the good quality data set. Here, blue dots represent good quality shots, and orange dots indicate shots where blisters occurred. Shots with blisters were found to have a relatively greater distance from the centroid of the good quality data set. Similarly, by setting the dotted line in the figure as a threshold, shots exceeding this threshold could be identified as blisters.
(1) Silver Streaks
Traditional methods using internal pressure sensor data, which utilized parameters such as maximum pressure waveform value, pressure waveform area, and time to maximum pressure, struggled to accurately detect silver streaks. However, in this research, we established a method to effectively detect silver streaks from internal pressure sensor data by extracting newly defined feature values and applying a unique algorithm. This advancement enhances the precision of quality control in the injection molding process.
(2) Blisters
Blisters, being temperature-caused molding defects, can now be detected by extracting specific feature values from resin temperature sensor data. This approach allows for a deeper understanding of the causes of blister formation and the introduction of new methods for their prevention.
These technological innovations mark the first time in the injection molding industry that both silver streaks and blisters can be detected using mold sensor data. Consequently, in addition to detecting shape defects like short shots and burrs, we can now identify aesthetic defects such as silver streaks and blisters, offering a more comprehensive solution for defect detection in molding processes.
