Abstract Feeder selection impacts the performance of bagging machinery throughout its life cycle, and yet it is usually based on qualitative assessments of flowability. We propose a data analysis methodology aimed at verifying the feeder-type classification of powders and grains by cluster analysis on their material properties. Results for a first data set of conventional properties show the granular materials clustered into as many groups as main bulk feeding systems. Mismatch between feeder classes and flowability-based clusters is explained by common industrial practice and incomplete material characterisation. For this reason, we introduce a set of specialised properties measured with the granular flow tester we have recently developed. Results for principal component analysis on a second extended property data set show that similarly flowing granular materials are better detected considering the specialised properties. This research contributes to objectify the decision-making process of bulk feeder selection from the quantitative description of granular flow.

中文翻译：

---

通过基于流动性的聚类对颗粒材料进行分类并应用于散装喂料

摘要 喂料器的选择会影响装袋机械整个生命周期的性能，但它通常基于流动性的定性评估。我们提出了一种数据分析方法，旨在通过对其材料特性的聚类分析来验证粉末和颗粒的进料器类型分类。常规特性的第一个数据集的结果显示，颗粒材料聚集成与主要散装进料系统一样多的组。馈线类别和基于流动性的集群之间的不匹配是由常见的工业实践和不完整的材料表征来解释的。出于这个原因，我们介绍了一组使用我们最近开发的颗粒流动测试仪测量的专业属性。对第二个扩展属性数据集进行主成分分析的结果表明，考虑到特殊属性，可以更好地检测类似流动的颗粒材料。本研究有助于从颗粒流的定量描述中客观化散装给料机选择的决策过程。