<p>The effective properties of CNT-polymer nanocomposites, which constitute a random heterogeneous material system, depend on the constituent phase properties and on microstructural information. In this work, the microstructure of computationally generated CNT-polymer nanocomposite realizations were assessed for the effective elastic stiffness, electrical conductivity, and piezoresistive response via finite element analysis. &nbsp;In addition, each microstructure considered was statistically characterized via spatial two-point correlation functions. Several types of two-point correlation functions were considered, namely pair correlation functions, ensemble and volume averaged two-point correlation functions, lineal path functions and two-point cluster function. Each of these capture specific features of the microstructure, such as the volume fraction, preferred orientation, shape, size, clustering, and dispersion of the phases. These features and combinations thereof are expected to drive the effective properties of interest including elastic stiffness, conductivity, and piezoresistivity. Since the two-point correlation functions contain high dimensional information about the microstructure (of the order of 10<sup>6</sup>&nbsp;dimensions), dimensionality reduction was performed via Principal Component Aanalysis (PCA) which reduces the dimensionality to the order of 10 dimensions. Principal component regression (linear regression using the principal scores of the correlation functions), or PCR, was performed to predict the effective properties of interest. It was found that the principal scores of the volume averaged two-point correlation functions were sufficient for predicting the effective stiffness with an accuracy greater than 90%. However, the accuracy of PCR predictions for conductivity and piezoresistivity coefficients was less than 50%, which indicates that higher order information about the microstructure such as phase connectedness and clustering needs to be included in the prediction process. To this end, lineal path functions and 2-point cluster functions were constructed for the microstructures. These are planned to be used in conjunction with the volume averaged two-point correlation functions as inputs to PCR to predict effective properties. Information about the barrier potential, which is a parameter that controls the extent of electron tunneling between CNTs, is also planned to be included in the prediction since electron tunneling is the key mechanism for composite conductivity and piezoresistivity in these material systems. Such reduced-order structure-property linkages can enable computationally affordable multiscale modelling of composites.</p>

Reduced-Order Structure-Property Linkage for Multifunctional Cnt-Polymer Nanocomposites via Principal Component Regression

Paper Proceedings: https://asmedigitalcollection.asme.org/SMASIS/SMASIS2021/volume/85499

ASME 2021 Smart Materials, Adaptive Structures and Intelligent Systems (SMASIS2021)

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.