Numerical Classification of Microstructures

这项研究工作的目的是提高从基材的微观结构和生产中使用的加工路径预测制成品性能的能力。为了做到这一点，有必要用定量的术语来描述微观结构；因此，图像分析系统和获得的显微结构信息的解释进行了研究，以数值描述显微结构组件。大多数金属合金的显微组织，如那些由凝固和铸造而来的，不容易用欧几里得几何来描述。对于这些情况，分形分析可以用来对微观结构进行数值描述，无论它是多么的混乱。这种方法的可行性已经在石墨铸铁中的石墨分布中得到了证明，并且应该适用于许多其他复杂的微观组织，这些组织目前还无法进行定量的数值评估。这项研究涉及合金模型系统和实际的工业产品，从铸铁和铝基发动机铸件开始。福特汽车公司和通用汽车公司将这项研究视为竞争前技术的一个重要领域，并为该计划做出了重大贡献。虽然该计划在学术上是刺激的，但它在制造过程的预测自动化方面具有相当大的工业潜力。＊＊＊

9520173 Hellawell This research effort is directed towards improved capability for predicting the properties of manufactured items from the microstructure of the base material and the processing path that was used in production. To accomplish this it is necessary to describe the microstructure in quantitative terms; accordingly, image analysis systems and the interpretation of acquired microstructural information are examined to numerically describe microstructural assemblies. A majority of metal alloy microstructures, such as those derived from solidification and casting, are not easily described in terms of Euclidean geometry. For these cases fractal analysis can be employed to make numerical statements about the microstructure, however chaotic it may be. The feasibility of this approach has been shown for distributions of graphite in graphitic cast irons and should apply to many other complex microstructures which presently defy quantitative numerical assessment. The research involves model alloy systems and actual industrial products, starting with cast irons and aluminum-based engine castings. Ford Motor Company and General Motors Corporation view this research as an important area of pre-competitive technology and are providing significant contributions to the program. %%% While this program is academically stimulating, it has considerable industrial potential for predictive automation of manufacturing processes. ***