A Software Architecture for Managing the Material Information Data Streams, Test Data and Model Predictions, Evident in Successful ICME Implementations

Working with the Materials Data Management Consortium and ASM’s National Center for Computational Materials Science and Engineering Data has shown how the right software tools are needed to maintain, visualize, and analyze the data, and to fully integrate that data with computer aided design and analysis packages. Underlying this, and crucial to providing secure, traceable, and accurate integration, is an information architecture with the flexibility to store and utilize the full range of materials information.

At the heart of an ICME system must be structures—flexible database schema—to handle the complexity of materials data.  Experimental details—including machine-types, processing parameters, temperatures, feeds and speeds, dates, names, etc.—must be maintained together with the experimental results. Computer modeling and simulation of the same situation can have just as many parameters, yet will require different structures to handle this data. Here, the metadata may relate to the model, algorithm, or analysis version. It will be linked to model verification and validation and will refer back to the input deck and pedigree of any input data. This paper will emphasize how, in both cases, the nature of these results will be different across the length scales, and yet the same structures must maintain all this information: microstructure, for example, must be stored alongside predicted properties of entire components.