Thermal and Mechanical Properties Enhancement of Epoxy Composites Using Graphene, MWCNTs, and Biochar: A Machine Learning-Assisted Approach
摘要
The development of lightweight, thermally efficient polymer composites is vital for advancing technologies in aerospace, electronics, and renewable energy sectors. This study investigates the thermal and mechanical performance of epoxy-based composites reinforced with graphene nanoplatelets, multi-walled carbon nanotubes (MWCNTs), banana peel-derived biochar (BB), and hybrid combinations of MWCNT and BB. These fillers were incorporated in varying weight percentages (0.2–0.8%) and processed using controlled magnetic stirring and ambient curing. The fabricated specimens were precisely cut using abrasive water jet technology to preserve structural integrity. Thermal conductivity, density, and Vickers hardness were measured following standard protocols. Among all fillers, graphene at 0.6 wt% exhibited the highest improvement in both thermal conductivity (0.237 W/m K) and hardness (17.75 HV), surpassing base epoxy values. MWCNTs showed consistent, though slightly lower, enhancements, while hybrid and BB fillers demonstrated modest gains. Additionally, a Random Forest regression model was developed to predict material properties based on filler type and composition. The model yielded strong performance in forecasting thermal conductivity and acceptable accuracy for hardness, validating its use as a predictive tool. Comparative analysis confirmed that graphene is the most effective filler for enhancing thermal and mechanical properties, while BB offers an eco-friendly, cost-effective alternative. The study highlights the potential of combining advanced nanofillers with sustainable bio-based materials for creating thermally functional, lightweight composites with machine learning–assisted optimization capabilities.