Polymer nanocomposites are a designer class of materials where nanoscale particles, functional chemistry and polymer matrix combine to yield unprecedented combinations of superior physical properties. While well-defined data representation and quantitative models have been developed in material science domains such as metallic alloys, polymer nanocomposites have remained a largely untouched space in terms of development of data-driven methods and data resources. Using nanodielectrics as a testbed and experimental dataset, a viscoelasticity-inspired finite element model is first developed to predict composite properties with local morphology and explicit inclusion of the interphase. Then, we demonstrate an online data-driven platform dedicated for polymer nanocomposite materials research that hosts materials data, analysis tools, and property simulation packages. To represent the high dimensional nanocomposite data, a non-relational XML schema is constructed to curate and store nanocomposite data, which is compatible with other materials data representation paradigm including knowledge graph theory and ontology. A test case is investigated leveraging existing tools to process image and materials and examine influence of interphase on viscoelasticity. The data-driven framework demonstrates promising potential for a large scale, comprehensive, quantitative exploration into the process-structure-property space for nanocomposite materials.

Creator

• Zhao, He

DOI

• https://doi.org/10.21985/n2-ktt8-7a23

Subject

• Mechanical engineering

Language

• en

Alternate Identifier

• etdadmin_upload_639974
• http://dissertations.umi.com/northwestern:14530

Keyword

• nanocomposite
• dielectrics
• materials informatics

Date created

• 2019-01-01

Resource type

• Dissertation

Rights statement

• In Copyright