The experiment, a collaborative effort by MAZIN Co., Ltd. and Hitachi High-Tech Corporation, aims to stabilize the quality of molded products using recycled materials. Currently, with the global trend of sustainable development, there is growing interest in environmentally friendly materials. This has led to an increase in the movement to replace virgin plastics with recycled plastics. Particularly in Europe, there are initiatives to legally mandate the use of recycled plastics, compelling manufacturers to adapt to new regulatory requirements. There are two types of recycled plastics: Pre-Industrial Recycled (PIR) plastics, which are reused from factory scraps generated before reaching the market, and Post-Consumer Recycled (PCR) plastics, which are collected after consumer use.
Due to heightened environmental considerations, the use of Post-Consumer Recycled (PCR) plastics is recommended. However, PCR plastics, recovered from a variety of products, exhibit variations in properties depending on their usage duration, manufacturing company, and the type of original product. This leads to issues of quality inconsistency. Additionally, the presence of undesired plastics and impurities poses a concern, making it challenging to collect PCR plastics with uniform properties. To mitigate this variability, a method of blending a few to several tens of percent of virgin plastic with PCR plastic has been adopted. However, considering future regulatory compliance and the difficulty in obtaining high-quality PCR plastic with less variability, it is necessary to increase the usage ratio of PCR plastics. Under these circumstances, the variability in properties across different batches poses a significant problem, and even with consistent molding conditions, changing batches could result in defective products. To address this issue, an experiment was conducted to analyze the variability in properties of PCR plastic across different batches.
This experiment, conducted with the cooperation of Horizon Corporation, involved molding experiments for printer paper guides. Three different batches of recycled ABS (PCR) were used, and each batch was molded 100 times under the same conditions. The resin pressure inside the mold was measured using an internal pressure sensor, and characteristic features were extracted from the obtained data. Based on these features, the variability in properties between the batches was quantitatively evaluated.
From the data obtained by the mold internal pressure sensor, two features thought to contribute significantly to the quality of the molded product (hereafter referred to as Feature 1 and Feature 2) were identified, and a scatter plot (Figure 3) was created.
This plot revealed that each batch formed distinct clusters, indicating variations in the quality of the molded product depending on the batch. In particular, focusing on the distribution of the cluster's central points clarifies the differences in quality between batches. Therefore, it can be said that there is a high likelihood of inconsistency in the quality of the molded products with each batch change.
This experiment confirmed that there is variability in the quality of PCR plastic from batch to batch. Given these results, it is clear that continuous production under the same molding conditions may not adequately address batch-to-batch variability, increasing the risk of defects. As a solution to this problem, the use of MAZIN's developing injection molding AI allows for automatic adjustment of molding conditions according to the variability of each batch, aligning the centers of the clusters. This enables stable production with minimal quality fluctuations, even when using recycled plastics. Currently, MAZIN is advancing the development of this AI system and working towards its practical application.
