Accelerated Evaluation of Properties of Polyphase Alloys

Simulations are utilized as partial substitutes for laboratory experiments.

A methodology for accelerated evaluation of mechanical properties of polyphase alloys is based on digital representations of the alloys. For a given alloy material system, this representation is utilized in concert with (1) software tools and probes that simulate traditional laboratory testing equipment and instrumentation, and (2) real laboratory mechanical testing by nontraditional methods.

Lattice Strains During Uniaxial Compression of a mechanically sintered, hot isostatically pressed alloy consisting of equal volume parts of copper and iron were measured experimentally and simulated.
The methodology was conceived under the presumption that the development of new polyphase alloys having mechanical properties suitable for loading conditions expected to be encountered in specific applications could be accelerated by accelerating the evaluation of the mechanical properties of such alloys by use of a combination of simulation, experimentation, and knowledge of quantitative associations between mechanical responses of the alloys and critical features of structures of the alloys. The digital representation of an alloy material system, denoted here simply as a "digital material," provides a design environment in which simulated and experimental data can be employed interchangeably in probing material-configuration space.

The representation is based on observable features of the internal geometry of an alloy material system, so that models of different forms and based on different size scales can be employed. The representation includes information on phases, grains within phases, dislocation structure, grain boundaries, and particles. At present, the representation is truncated at the size scale of constituent particles. Distributions of attributes are represented in a standardized manner by use of piecewise polynomial functions.

Finite-element alloy samples that simulate specimens machined from alloy stock can be created from data sampled from the aforementioned distributions of attributes. These simulated samples can be mechanically loaded and probed in a manner that mimics physical experimentation on real alloy samples. Mechanical and other physical responses of samples can be measured, and corresponding physical properties can be extracted from such virtual experiments, then stored in an appendix to a database that is part of the digital representation. By use of the information in this appendix, one can browse various estimates of critical material properties. Thus, the digital material system mimics a laboratory- based system for evaluation of mechanical properties.

Although the methodology is largely oriented toward accelerating the development of alloys by using multiscale software simulations as substitutes for some physical tests, experimentation lies at the core of the methodology in the sense that the high-fidelity experiments are used to validate and calibrate the simulations. The primary role of the experimental data within the digital material system is the establishment of reference states, which are direct points of contact between measured and calculated values of attributes for the same nominal material states. As such, reference states enable both the initialization and validation of computational models of alloys. For example, one attribute that has been employed extensively is lattice strain (see figure), which is regarded as a direct link to the micromechanical state and, therefore, invaluable for validation of models.

This work was done by Paul Dawson, Matthew Miller, Tong Seok Han, and Joel Bernier of Cornell Univeristy for the Air Force Research Laboratory. For further information, download the free white paper at www.defensetechbriefs.com  under the Materials category. AFRL-0001



This Brief includes a Technical Support Package (TSP).
Document cover
Accelerated Evaluation of Properties of Polyphase Alloys

(reference AFRL-0001) is currently available for download from the TSP library.

Don't have an account? Sign up here.