Mechanical characterization and predictive modeling of activated carbon-reinforced PLA thermoplastic composites: A comparative study of ANN, RSM, and Bezier curve models

PLA-based composites can offer environmental advantages over composites based on conventional petroleum-derived polymers, owing to the renewable origin and potential biodegradability of PLA. The objective of this research is to compare the capability of predictive models such as ANN, RSM, and Bezier curve on the mechanical performance of polylactic acid (PLA) composites reinforced with activated carbon (AC) derived from coconut shell at different filler loadings from 0 to 10 wt.%. The composites were manufactured using melt blending, extrusion, palletizing, and compression moulding. The mechanical characterizations, like tensile behavior, surface hardness, and impact strength, were studied and scanning electron microscopy (SEM) was utilized to study the morphology of the developed composites. The experimental results showed that the tensile and impact strengths of the composites reduced with the increasing AC content. Among the investigated predictive models, the Bezier-based model provided close agreement with the experimental data, with R 2 values of 0.9995, 0.9450, and 0.9657 for tensile strength, hardness, and impact strength. The results show the potential of smooth curve-based modelling for representing nonlinear composition–property relationships in small experimental datasets. The Bezier model is intended for interpolation within the investigated experimental range only and should not be used for extrapolation beyond the available experimental data. The combined experimental and predictive framework provides insight into the mechanical performance of PLA–AC thermoplastic composites and supports the development of data-driven approaches for the design of sustainable particulate-filled polymer composites.