Data-driven methods have attracted increasingly more attention in materials research since the advent of the material genome initiative. The combination of materials science with computer science, statistics, and data-driven methods aims to expediate materials research and applications and can utilize both new and archived research data. In this paper, we present a data driven and deep learning approach that builds a portion of the structure–property relationship for polymer nanocomposites. Analysis of archived experimental data motivates development of a computational model which allows demonstration of the approach and gives flexibility to sufficiently explore a wide range of structures. Taking advantage of microstructure reconstruction methods and finite element simulations, we first explore qualitative relationships between microstructure descriptors and mechanical properties, resulting in new findings regarding the interplay of interphase, volume fraction and dispersion. Then we present a novel deep learning approach that combines convolutional neural networks with multi-task learning for building quantitative correlations between microstructures and property values. The performance of the model is compared with other state-of-the-art strategies including two-point statistics and structure descriptor-based approaches. Lastly, the interpretation of the deep learning model is investigated to show that the model is able to capture physical understandings while learning.
more »
« less
Materials Informatics and Data System for Polymer Nanocomposites Analysis and Design
The application of Materials Informatics to polymer nanocomposites would result in faster development and commercial implementation of these promising materials, particularly in applications requiring a unique combination of properties. This chapter focuses on a new data resource for nanocomposites — NanoMine — and the tools, models, and algorithms that support data-driven materials design. The chapter begins with a brief introduction to NanoMine, including the data structure and tools available.
Critical to the ability to design nanocomposites, however, is developing robust structure–property–processing (s–p–p) relationships. Central to this development is the choice of appropriate microstructure characterization and reconstruction (MCR) techniques that capture a complex morphology and ultimately build statistically equivalent reconstructed composites for accurate modeling of properties. A wide range of MCR techniques is reviewed followed by an introduction of feature selection and feature extraction techniques to identify the most significant microstructure features for dimension reduction. This is then followed by examples of using a descriptor-based representation to create processing–structure (p–s) and structure–property (s–p) relationships for use in design. To overcome the difficulty in modeling the interphase region surrounding nanofillers, an adaptive sampling approach is presented to inversely determine the inter-phase properties based on both FEM simulations and physical experiment data of bulk properties. Finally, a case study for nanodielectrics in a capacitor is introduced to demonstrate the integration of the p–s and s–p relationships to develop optimized materials for achieving multiple desired properties.
more »
« less
- Award ID(s):
- 1729452
- PAR ID:
- 10187213
- Date Published:
- Journal Name:
- Handbook on Big Data and Machine Learning in the Physical Sciences
- Volume:
- 1
- Page Range / eLocation ID:
- 65-126
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Abstract Accelerating the design and development of new advanced materials is one of the priorities in modern materials science. These efforts are critically dependent on the development of comprehensive materials cyberinfrastructures which enable efficient data storage, management, sharing, and collaboration as well as integration of computational tools that help establish processing–structure–property relationships. In this contribution, we present implementation of such computational tools into a cloud-based platform called BisQue (Kvilekval et al., Bioinformatics 26(4):554, 2010). We first describe the current state of BisQue as an open-source platform for multidisciplinary research in the cloud and its potential for 3D materials science. We then demonstrate how new computational tools, primarily aimed at processing–structure–property relationships, can be implemented into the system. Specifically, in this work, we develop a module for BisQue that enables microstructure-sensitive predictions of effective yield strength of two-phase materials. Towards this end, we present an implementation of a computationally efficient data-driven model into the BisQue platform. The new module is made available online (web address:
https://bisque.ece.ucsb.edu/module_service/Composite_Strength/ ) and can be used from a web browser without any special software and with minimal computational requirements on the user end. The capabilities of the module for rapid property screening are demonstrated in case studies with two different methodologies based on datasets containing 3D microstructure information from (i) synthetic generation and (ii) sampling large 3D volumes obtained in experiments. -
Interfaces such as grain boundaries (GBs) and heterointerfaces (HIs) are known to play a crucial role in structure-property relationships of polycrystalline materials. While several methods have been used to characterize such interfaces, advanced transmission electron microscopy (TEM) and scanning TEM (STEM) techniques have proven to be uniquely powerful tools, enabling quantification of atomic structure, electronic structure, chemistry, order/disorder, and point defect distributions below the atomic scale. This review focuses on recent progress in characterization of polycrystalline oxide interfaces using S/TEM techniques including imaging, analytical spectroscopies such as energy dispersive X-ray spectroscopy (EDXS) and electron energy-loss spectroscopy (EELS) and scanning diffraction methods such as precession electron nano diffraction (PEND) and 4D-STEM. First, a brief introduction to interfaces, GBs, HIs, and relevant techniques is given. Then, experimental studies which directly correlate GB/HI S/TEM characterization with measured properties of polycrystalline oxides are presented to both strengthen our understanding of these interfaces, and to demonstrate the instrumental capabilities available in the S/TEM. Finally, existing challenges and future development opportunities are discussed. In summary, this article is prepared as a guide for scientists and engineers interested in learning about, and/or using advanced S/TEM techniques to characterize interfaces in polycrystalline materials, particularly ceramic oxides.more » « less
-
Polymer-derived ceramic (PDC) nanocomposites enable access to a large library of functional properties starting from molecular design and incorporating nanofillers. Tailoring preceramic polymer (PCP) chemistry and nanofiller size and morphology can lead to usage of the nanocomposites in complex shapes and coatings with enhanced thermal and mechanical properties. A rational design of targeted nanocomposites requires an understanding of fundamental structure–property–performance relations. Thus, we tailor our discussions of PCP design and nanofiller integration into single source precursors as well as pyrolytic processing for functionalizing PDCs. We also discuss the promises and limitations of advanced characterization techniques such as 4D transmission electron microscopy and pair distribution functions to enable in situ mapping structural evolution. The feedback loop of in situ monitoring sets the foundation for enabling accelerated materials discovery with artificial intelligence. This perspective assesses the recent progress of PDC nanocomposite research nanocomposites and presents scientific and engineering challenges for synthesis, fabrication, processing, and advanced characterization of PDC nanocomposites for enhanced magnetic, electrical, and energy conversion and storage properties.more » « less
-
With an unprecedented combination of mechanical and electrical properties, polymer nanocomposites have the potential to be widely used across multiple industries. Tailoring nanocomposites to meet application specific requirements remains a challenging task, owing to the vast, mixed-variable design space that includes composition ( i.e. choice of polymer, nanoparticle, and surface modification) and microstructures ( i.e. dispersion and geometric arrangement of particles) of the nanocomposite material. Modeling properties of the interphase, the region surrounding a nanoparticle, introduces additional complexity to the design process and requires computationally expensive simulations. As a result, previous attempts at designing polymer nanocomposites have focused on finding the optimal microstructure for only a fixed combination of constituents. In this article, we propose a data centric design framework to concurrently identify optimal composition and microstructure using mixed-variable Bayesian optimization. This framework integrates experimental data with state-of-the-art techniques in interphase modeling, microstructure characterization and reconstructions and machine learning. Latent variable Gaussian processes (LVGPs) quantifies the lack-of-data uncertainty over the mixed-variable design space that consists of qualitative and quantitative material design variables. The design of electrically insulating nanocomposites is cast as a multicriteria optimization problem with the goal of maximizing dielectric breakdown strength while minimizing dielectric permittivity and dielectric loss. Within tens of simulations, our method identifies a diverse set of designs on the Pareto frontier indicating the tradeoff between dielectric properties. These findings project data centric design, effectively integrating experimental data with simulations for Bayesian Optimization, as an effective approach for design of engineered material systems.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1142/9789811204555_0003
