Informatics for mapping engineering data

Since their inception nearly forty years ago, Ashby maps have demonstrated their value in providing guidelines for materials selection. Through the mapping of structure and property data through the use of dimensional analysis tied to phenomenological relationships, these maps have been shown to be a useful tool for classifying materials behavior. In this paper, we demonstrate how computational tools based on data mining and informatics methods can provide a new dimension to such maps by permitting the prediction of new materials and expected properties. The value of such methods rest in the fact that it can effectively fill in new information that is at present is missing in Ashby maps. It is proposed that informatics based methods provides the critical link to integrate data and information for computation and experiment in materials engineering.