| Abstract Scope |
Realizing Integrated Computational Materials Engineering (ICME) from a design perspective involves co-considering the interactions between stakeholders from manufacturing, materials, and product disciplines. Effectively realizing the product-material-manufacturing system requires a systems-based, top-down, inverse design approach that supports formulating the multidisciplinary problem and coordinating the couplings and interactions using simulation models. This necessitates the capability to carry out inverse robust co-design, which involves simultaneously exploring a range of robust design solutions that satisfy stakeholders' preferences across multiple disciplines while managing uncertainties.
In this paper, we present a multidisciplinary, inverse, robust co-design framework for design space exploration involving multiple disciplines and the management of uncertainty. The framework integrates (i) data-driven Gaussian Process (GP) models for establishing the Processing-microStructure-Property-Performance (PSPP) linkages, (ii) Bayesian inference methods together with robust design and decision support problem constructs for inverse robust co-design problem formulation, and (iii) interpretable Self-Organizing Maps (iSOM) to simultaneously visualize and explore the multidisciplinary design space. The efficacy of the framework is demonstrated using an industry-inspired problem involving the hot rod rolling process, which considers the interactions between material, microstructure, and manufacturing conditions. The framework generalizes to a broad range of manufacturing processes and materials and supports the co-design exploration of systems characterized by multidisciplinary interactions and the presence of uncertainties. |