Solvent‐Assisted Dispersion of Multiwalled Carbon Nanotubes in Epoxy‐Based Glass Fiber‐Reinforced Polymer Composites: Experimental Investigation, Micromechanical Modeling, and Machine Learning Prediction

ABSTRACT

The investigation examines the improvement of mechanical behavior of epoxy and glass fiber (GF)‐reinforced nanocomposites containing multiwalled carbon nanotubes (MWCNT) with weight fractions of 0.25, 0.50, and 0.75 respectively using ultrasonication‐assisted dispersion in organic solvents namely, ethanol and dimethyl ketone. A systematic approach to solvent‐mediated dispersion was investigated and assessed its impact on macro‐scale mechanical response, interfacial interaction, and microstructural morphology. Epoxy nanocomposite with 0.5 wt.% ethanol‐dispersed MWCNT demonstrated a notable 23% increase in ultimate tensile strength (48 MPa), as compared to pure epoxy. In MWCNT/glass fiber‐reinforced polymer (GFRP) nanocomposite, 0.5 wt.% MWCNT improves the Young's modulus by 5.3%, whereas 0.25 wt.% MWCNT results in a significant 48.54% increase in flexural strength (339.4 MPa) and a remarkable 142.22% rise in maximum flexural extension (15.26 mm) as compared to pristine GFRP. The Halpin–Tsai micromechanical model and the Mori–Tanaka mean field homogenization approach are used to provide a mechanistic basis for the experimental observations. This allows for a quantitative evaluation of the impact of filler aspect ratio, volume fraction, and orientation on composite elastic response. Agglomeration effects and solvent residuals phenomena are quantitatively modeled using an agglomeration efficiency parameter, and the empirically obtained moduli are contextualized against Halpin–Tsai predictions. Moreover, machine learning models accurately predict mechanical characteristics, demonstrating R ² values ranging from 0.976 to 0.983. This research establishes a dual analytical‐predictive framework with high translational potential for advanced structural applications.