COMSOL simulation of CNT assembly in an epoxy matrix under static magnetic fields for polymer nanocomposite applications

Open Access
- Author:
- Oyama, Kohei
- Graduate Program:
- Aerospace Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- April 04, 2022
- Committee Members:
- Namiko Yamamoto, Thesis Advisor/Co-Advisor
Xin Ning, Committee Member
Amy Pritchett, Program Head/Chair
Charles E Bakis, Committee Member - Keywords:
- Carbon nanotube
Composites
Epoxy
FEM
COMSOL
MATLAB
Magnetic field
Nickel - Abstract:
- Carbon fiber reinforced plastics (CFRPs) have been used for many industries because of their light weight, and significant properties. However, CFRPs show low electrical and thermal conductivities, and short fatigue life because of the delamination of the interlaminar boundaries. The through-thickness properties of CFRPs can be reinforced by crack bridging of CNTs without affecting the in-plane properties of CFRPs, and can improve anisotropic electrical conduction. One of the challenges to implement CNTs into CFRPs is to control the dispersion and the alignment of CNTs and aggregations of CNTs. Effective organization of CNTs can be achieved using magnetic field application. In the previous work of our laboratory, multi-walled CNTs were fabricated and they were magnetized, diazotized, and magnetically aligned under a static magnetic field during the process of forming CNT-epoxy composites. While fracture toughness improved with small CNT addition in experiments, such improvement was not linear with increasing CNT volume fraction or increasing magnetic field strength, possibly due to CNT agglomeration. More specifically, the CNT-epoxy composites with lower volume fraction (0.1 vol.%) showed comparable fracture toughness improvement regardless of the magnetic field strength (180 G and 300 G), while those with higher volume fraction (0.5 vol.%) showed less fracture toughness improvement with the stronger magnetic field (300 G) than with the weaker magnetic field (180 G). Therefore, the factors of those results were needed to be studied to better control fracture toughness improvement. However, experimental observation of CNT within an epoxy is a challenge. Thus simulations of CNT assembly are necessary to understand the CNT structures in relation to the property change by CNTs. In this work, assembly behaviors of elliptic cylinder particles, which simulate nickel-coated CNT bundles, within an epoxy matrix were studied using analytical solutions and multi-physics simulations with a finite element method (FEM) software, COMSOL Multiphysics. The effects of various parameters on magnetic assembly of particles under static magnetic fields were studied, including the initial positions and angles of particles, the volume fraction of particles, the aspect ratio of particles, and the strength of the magnetic fields. First, the averaged length, width, and aspect ratio of 108 CNTs/CNTs agglomerations were obtained as 14 ± 4 μm, 3 ± 1 μm, and 9 ± 2 respectively using optical microscope as inputs for analytical solutions and COMSOL simulations. Second, the time required to complete particle rotation was calculated analytically using MATLAB. To simulate the nickel-coated CNTs in experiments, a particle was shaped as a cylinder (aspect ratio of 7.5) and assigned with material properties of nickel. The matrix was set to have the viscosity of 70 cP, to simulate the epoxy matrix at the working temperature. When the static field of 180 G was applied, the particle completed rotation in 9.7 × 10-3 s, which was comparable with the time calculated using COMSOL for an elliptic cylinder (<10-3 s). Third, COMSOL simulations were conducted to understand the effects of various parameters, and also to simulate the particle assembly that are comparable with the experiments. In simulation, the two particles assembled faster or assembled at all when one (or two) of the particles are more aligned with the applied magnetic field direction. The inter-particle distance after particle rotation made a larger effect on particle assembly behavior than the initial inter-particle distance. The COMSOL simulation results about the cases with 0.1 vol% of CNTs (four particles) indicated that magnetic assembly is mostly determined by the initial particle positions and angles, rather than magnetic field strength. This simulation result helped to interpret the experimental results where fracture toughness improvement was not affected by the magnetic field strength. The COMSOL simulation results about the cases with 0.5 vol% of CNTs (20 particles) indicated that their magnetic assembly was observed regardless of the magnetic field strength, as the particles were more densely distributed. This simulation result helped to understand the experimental results where fracture toughness was mitigated with the larger field strength of 300 G where more agglomerations were expected. In order to maximize fracture toughness of the sample, seeking the best volume fraction of CNTs between 0.1 vol.% and 0.5 vol.% under a weak magnetic field strength (180 G or smaller) in experiments is suggested to gain more aligned CNTs within a matrix than 0.1 vol.% cases, with less amount of their agglomerations than 0.5 vol.%. In future, simulation studies need to be modified to accommodate contact between particles so that simulation keeps going even after the first assembly of particles, providing better understandings regarding the behaviors of CNTs within an epoxy matrix to obtain more fracture toughness improvement.