Open AccessPhysics-Informed Network Models: a Data Science Approach to Metal Design
Author(s) -
Amit K. Verma,
Roger H. French,
Jennifer L. W. Carter
Publication year - 2017
Publication title -
integrating materials and manufacturing innovation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.878
H-Index - 22
eISSN - 2193-9772
pISSN - 2193-9764
DOI - 10.1007/s40192-017-0104-5
Subject(s) - interdependence , a priori and a posteriori , interpretation (philosophy) , function (biology) , computer science , control (management) , work (physics) , experimental data , mechanical engineering , engineering , artificial intelligence , mathematics , programming language , philosophy , epistemology , evolutionary biology , political science , law , biology , statistics
Functional graded materials (FGM) allow for reconciliation of conflicting design constraints at different locations in the material. This optimization requires a priori knowledge of how different architectural measures are interdependent and combine to control material performance. In this work, an aluminum FGM was used as a model system to present a new network modeling approach that captures the relationship between design parameters and allows an easy interpretation. The approach, in an un-biased manner, successfully captured the expected relationships and was capable of predicting the hardness as a function of composition.