Computer-aided multi-scale materials and product design

The success of computer-aided design (CAD) tools in engineering design has been witnessed in the past four decades. Advanced CAD tools have become a critical enabling technology in product design since they help build virtual prototypes of complex products (e.g. automobiles, airplanes, and electronic appliances). In today’s CAD-enabled design processes, design engineers select availablematerials from databases to realize desired product functionality. Material databases are constructed by materials scientists and engineers based on the experimentally tested physical properties of existing materials. Therefore, the available ‘degrees of freedom’ for design engineers to control and optimize the performance of products are restricted to geometry and topology only. To offer design engineers more degrees of freedom, the additional dimension of material properties in the design solution space is highly valuable. Customizable materials can provide extra flexibility to realize the increasingly intricate product functionality. In other words, materials selection should be replaced by materials design to better meet stakeholders’ requirements in modern product realization. It can be envisioned that design engineers will perform product and materials design aided by multi-scale CAD software tools in the near future. They can not only perform the traditional geometry construction and structural analysis in such a multiscale CAD environment, but also customize thematerial properties for some local region of the design by simply zooming into the specific region to specify material compositions or crystalline configurations. Design occurs concurrently at multiple length scales, including nano-,meso-,micro-, andmacro-scales. Extensive problem-specific materials customization is thus achievable. Design decisions made at the microscopic levels determine the physical properties of newmaterials and the behavior of products. Materials design is integrated with geometry and topology design at the macroscopic level. Multi-scale CAD software tools help engineers to set functional objectives at different scales, construct material-inclusive product models, simulate and optimize design, and guide laboratory efforts to implement materials and target behaviors of final products. The value of a multi-scale CAD environment lies in its seamless knowledge integration across disciplinary boundaries. The conventional material selection approach that design engineers usually take is based on the isolated databases that were built without the input of problem-specific needs. Such a one-directional bottomup approach to discover, devise, and deploy new materials has a long development cycle and is not cost-effective. Even the materials science and engineering community has realized that the ‘lack of design’ limits the rapid advancement of engineering materials. Therefore, a systems engineering approach to integrate processing, structure, property, and performance for materials design has been adopted (see: Olson, G. B., ‘‘Computational design of hierarchically structured materials’’, Science, 1997, Vol. 277, No. 5330, pp. 1237–1242). A new term of Integrated Computational Materials Engineering (ICME) was coined to capture the top-down and target-oriented design process for materials that employs modern computational power to represent fine-grainedmicrostructures, to simulatemulti-level properties, and to devise fabrication processes of newmaterials. Most recently a newMaterials Genome Initiative was kicked off in the U.S.A. to develop a new materials innovation infrastructure of computational and experimental tools along with digital data to accelerate materials development. Various technical challenges to realize the new computational infrastructure for materials design also provide research opportunities, such as (i) how to rigorously quantify microstructure-property-processing relationships and establish highlevel abstraction of materials knowledge, (ii) how to effectively extract useful information and knowledge out of abundant characterization data, (iii) how to integrate information and knowledge gained from simulation at individual scales for multi-scale systemsdesign, (iv) how to improve efficiency and accuracy in such multi-scale simulation schemes, (v) how to extend available design methodologies such as data mining, analysis, and optimization in materials design, and (vi) how to quantify uncertainties in the design process, specifically variability caused by inherent randomness and incertitude due to the lack of perfect knowledge, in order to improve the robustness of design. In this special issue, the paper entitled ‘‘Key computational modeling issues in Integrated Computational Materials Engineering ’’, authored by Panchal, J.H., Kalidindi, S.R., and McDowell, D.L. provides an extended overview of the major ongoing research elements in ICME, includingmaterials databases for information collected from characterization, microstructure knowledge systems that represent and visualize high-level knowledge about materials, multi-scale modeling for materials design, uncertainty quantification andmanagement. Similar to the conventional CAD at bulk scale as the first tool for virtual prototyping, the primary function of multi-scale CAD is to allow for the efficient construction and interactive modification of geometric models for microstructures at small scales. Existing boundary-representation based parametric modeling approaches have become inefficient in model construction at nanoandmesoscales where geometry and topology are highly complex, because a great number of control points are required. New modeling and representation techniques are thus needed. For example, an implicit modeling approach can efficiently represent superporousmicrostructures with intricate topology. To designmaterial properties, the local control of shapes in microstructures is no longer important. It is more critical to allow design engineers to modify the overall structure globally in a representative volume