Session: 12-20-01: Composite Materials and Mechanics

Paper Number: 144779

144779 - Modeling of Nanoscale Characteristics of Carbon Nanotube Nanocomposites Through Experimental and Statistical Methods 

Polymer matrix fiber-reinforced composites have a wide range of applications in industry due to their low density and high strength; consequently, materials and methods for manufacturing such composites are a large focus in current research. However, delamination in laminated composites is a concern and influences the integrity of the composite systems. Common solutions include nanoparticles and interleaving membranes, which act as crack arrestors and improve anti-delamination properties. The authors examined the use of carbon nanotubes (CNTs) in laminated composites, as mixing a low-weight percentage of dispersed CNTs into a polymer improves the interlaminar fracture toughness of composites; the weight percentage of CNTs is a crucial metric, as at a higher weight percentage, nanoparticles tend to agglomerate and reduce mechanical properties.

Specifically, this paper focuses on the use of Buckypaper (BP)—a composite consisting of a crystalline CNT network, dispersed CNTs, and resin-rich zones—in layered composites. Previous studies have shown that integrating a well-infiltrated thin CNT membrane improved the mode I fracture toughness of carbon fiber reinforced polymer (CFRP) composites by 60% via diverting the crack front above and below the membrane. Conversely, poor impregnation of BP reduced interlaminar fracture toughness in composites. As such, characterization of BP is essential for its effective use in layered polymer nanocomposites (PNCs).

To examine the nanoscale material properties of BP, Atomic Force Microscopy Peak Force Quantitative Nanomechanics Mapping (AFM PFQNM) was used. AFM is a robust characterization technique capable of obtaining material data on a scale from a few nanometers to tens of nanometers, and AFM PFQNM is based on peak force tapping technology that eliminates adhesion-based artifacts while tapping. AFM PFQNM allows for detailed characterization of CNT nanocomposites and the influence of CNTs on composite material properties. Length and diameter of CNT, type of CNT, functionalization and weight percentage, and dispersion technique affect agglomeration vs. network creation, the microstructure size and shape, and ultimately, the macroscale properties of the CNT nanocomposites. Therefore, the distribution of the nanoscale and macroscale properties is stochastic.

The stochastic nature of the data distribution of interphase size and properties, nanoparticle networks, and macroscale fracture properties impedes analysis and modelling of PNC material properties. While the macroscale fracture properties of nanocomposites have been studied for years, further understanding of interphase size, properties, and particle network morphology is needed. Traditional approaches for modelling this data have not been reliable; as an alternative, this paper explores a data-dependent model of PNC properties. The standard Weibull model was tested to simulate the material properties of polymers, composites, and nanocomposites. AFM PFQNM was used to collect 500 and 180 data sets of CNT network properties for low- and high-weight CNT PNCs, respectively. The authors considered all Weibull distribution functions, and two-, three-, and four-parameter Weibull models were used to simulate the nano- and macro-scale material properties of CNT PNCs to find the most accurate and efficient functions. Results indicate that a four-parameter Weibull model is the most efficient model to simulate data.

Presenting Author: Tyler Norkus Arizona State University

Presenting Author Biography: Tyler Norkus is pursing a master's degree in mechanical engineering at Arizona State University, with a focus on carbon nanotube composites, nanoscale mechanics, atomic force microscopy, and finite element analysis. He has published in two previous conference proceedings and plans on publishing a journal paper before graduating with his masters.

Authors:

Tyler Norkus Arizona State University
Masoud Yekani Fard California Polytechnic State University