Session: 08-02: Micro/Nanoscale Heat Conduction

Paper Number: 131028

131028 - Nanoscale Mechanisms of Heat Transfer in Carbon Fibers: Insights From Large-Scale Atomistic Simulations 

Abstract: 

Carbon fiber is a lightweight, ultra-strong material widely used in aerospace, automotive, defense, sports equipment, and wind energy sectors. The increasing demand for carbon fibers, largely driven by the quest for energy efficiency in transportation and the rapid growth of wind energy industry, are stimulating the exploration of a broader range of molecular precursors, structure-templating agents, and fiber manufacturing procedures aimed at producing high-performance fibers at a reduced cost. This exploration is guided by investigations focused on revealing the relationships between carbon fiber nanostructure and properties.

In this presentation, the dependence of the thermal transport properties on the nanoscale and mesoscopic structure of carbon fibers is investigated in large-scale nonequilibrium molecular dynamics simulations. The simulations are performed for computational samples produced by a computational procedure that involves self-assembly and reactive fusion of idealized hydrogenated carbon sheets or blocks of ladders of different sizes and orientations [1,2]. By applying a sequence of equilibration, compression, dehydrogenation, high-temperature annealing and relaxation steps, the structural characteristics of computational samples, such as the relative fractions of turbostratic, graphitic, and amorphous domains, the void size distribution, as well as the degree of orientational ordering, can be controlled and tuned to match those of real fibers.

The investigation of the thermal transport properties of carbon fibers is based on the calculation of the atomically-resolved heat flux, which helps to establish and visualize the channels of the heat flow through a carbon fiber nanostructure. The results of the simulations suggest that the heat preferentially flows through interconnected graphitic and turbostratic carbon regions, while the contribution of amorphous regions is low. At the quantitative level, the values of the longitudinal (along the fiber axis) and transversal (perpendicular to the fiber axis) thermal conductivity predicted in the atomistic simulations are found to be consistent with the values obtained in recent time-domain thermoreflectance measurements performed for polyacrylonitrile (PAN)-based carbon fibers [3]. The analysis of the dominant nanoscale channels of the heat transfer in the simulations reveals the factors that control the anisotropic thermal conductivity of carbon fibers.

An additional set of simulations is performed for a core-skin carbon nanofiber featuring an idealized graphitic multilayer skin region covering a more disordered core nanostructure [4]. The results of the simulations demonstrate that the heat flux through the graphitic shell is significantly higher than that through the core of the nanofiber. The chemical cross-links between the graphitic layers reduce the heat flux through the shell, while the atomically-resolved heat flux is particularly reduced in the vicinity of the cross-links. The analysis of the sample size dependence of the thermal conductivity suggests that the heat transfer along the core-shell nanofiber contains a ballistic contribution up to the length of 54 nm studied in the simulations.

Overall, the results of the simulations demonstrate the power of the atomically-resolved heat flux calculation in revealing the nanoscale channels of the heat transfer in heterogeneous systems and suggest the possibility of tuning the anisotropic thermal conductivity of carbon fibers through the computationally-assisted material design.

 

[1]  K. Joshi, M. I. Arefev, and L. V. Zhigilei, Generation and characterization of carbon fiber microstructure in atomistic simulations, Carbon 152, 396-408 (2019).

[2]  M. He, M. I. Arefev, K. Joshi, and L. V. Zhigilei, Atomistic modeling of tensile deformation and fracture of carbon fibers: Nanoscale stress redistribution, effect of local structural characteristics and nanovoids, Carbon 202, 528-546 (2023).

[3]  X. Ji, S. Matsuo, N. R. Sottos, and D. G. Cahill, Anisotropic thermal and electrical conductivities of individual polyacrylonitrile-based carbon fibers, Carbon 197, 1-9 (2022).

[4]  M. He, K. Joshi, and L. V. Zhigilei, Computational study of the effect of core-skin structure on the mechanical properties of carbon nanofibers, J. Mater. Sci. 56, 14598-14610 (2021).

Presenting Author: Antonios S. Valavanis University of Virginia

Presenting Author Biography: Antonios Stylianos Valavanis is a Ph.D. student in the Materials Science and Engineering program at the University of Virginia. Over the past 4 years, he has been actively engaged in the field of molecular dynamics (MD) simulations. In 2019-2020, he was working on continuum-level modeling of short-pulse laser interactions with metals in the group of Dr. Emmanuel Stratakis at FORTH Institute, Greece as an undergraduate student. In 2020, he joined the Computational Materials Group at the University of Virginia, as a Ph.D. student. In Virginia, Mr. Valavanis continued his computational research using MD to unravel the mechanisms of heat transfer in composite materials as well as studying the mechanisms of laser induced structural and phase transformations in metals. In July of 2022, his work on surface morphology modification in short-pulse laser processing was recognized by the Roger Kelly award at the 7th International School on Lasers in Materials Science in Venice, Italy.

Authors:

Antonios S. Valavanis University of Virginia
Leonid V. Zhigilei University of Virginia