Statistical modeling and optimization of the mechanical performance of natural fiber-reinforced polyolefin composites by applying response surface methodology

Natural fiber-reinforced polyolefin composites have emerged as promising sustainable materials; however, their mechanical performance is strongly governed by the chemical composition and moisture characteristics of the reinforcing fibers. This research establishes an interpretable Response Surface Methodology (RSM)-based framework to forecast and enhance the tensile strength (TS) and tensile modulus (TM) of natural fiber/HDPE composites as functions of cellulose content, hemicellulose content, moisture content, and filler loading. A complete dataset from the literature was put together, cleaned up, normalized, and coded for regression analysis. Second-order polynomial models were formulated within a Central Composite Design (CCD)-based analytical framework and validated using analysis of variance (ANOVA), residual diagnostics, and cross-validation. The developed models were highly significant ( p < 0.0001), exhibited excellent goodness of fit (R 2 > 0.98), and achieved average external validation errors below 5%. The results indicated that TS and TM increased with cellulose content and filler loading up to an optimal region of approximately 65–70 wt.% cellulose and 25–30 wt.% filler, after which performance declined because of fiber agglomeration and reduced polymer wetting. In contrast, increasing hemicellulose and moisture content decreased mechanical performance because of their hygroscopic and amorphous characteristics. Multi-response optimization identified an optimal composition of approximately 66 wt.% cellulose, 19 wt.% hemicellulose, 8 wt.% moisture, and 27 wt.% filler, corresponding to maximum predicted values of 45.2 MPa for TS and 2.95 GPa for TM. The proposed composition-driven RSM framework provides a practical quantitative basis for designing high-performance sustainable polyolefin composites.