M. Girish Prasad, Y. P. Ravitej, Rayappa Shrinivas Mahale, B. V. N. Ramakumar, Jayatirtha M. Patil, Krantikumar Kshaurad, Rajeev Gupta, Manish Kumar Mishra, N Abhijith, Balachandra Halemani, B. Nityanand
Nanostructured reinforcements such as graphene nanoplatelets (GNPs) offer exceptional strength and stiffness, but their use in polymer nanocomposites is constrained by dispersion challenges and property trade-offs. This study investigates the effect of GNP loading (0.1–0.5 wt.%) on epoxy nanocomposites fabricated using hybrid magnetic stirring and ultrasonication. Tensile strength improved by ∼50%, rising from 36 MPa at 0.1 wt.% to 55 MPa at 0.3 wt.%, before decreasing sharply at higher concentrations owing to agglomeration. Rockwell hardness declined progressively from L89 to L42, while SEM confirmed uniform dispersion at low loadings and clustering at higher ones. Predictive modelling supported the experimental findings: polynomial regression identified an optimum threshold of 0.27–0.30 wt.% (∼56 MPa), and a Random Forest model effectively captured nonlinear property variations. Overall, 0.3 wt.% emerges as the optimal reinforcement level, achieving significant strength enhancement while preserving structural stability. The novelty of this work lies in quantitatively defining the reinforcement threshold by integrating hybrid processing, characterisation, and machine learning. This predictive experimental framework provides practical guidance for the rational design of lightweight epoxy nanocomposites for aerospace, automotive, and protective coating applications.