Optimization of alkali treatment severity and laminate architecture in ramie/epoxy composites via D-optimal response surface methodology

This study presents a systematic multi-response optimisation of ramie bidirectional woven mat-reinforced epoxy composites through a D-Optimal Response Surface Methodology (RSM) design. NaOH concentration (1, 3, 5, 7 wt%), fibre immersion time (1, 3, 5, 7 h), and ply orientation angle (0°, 15°, 30°, 45°) were evaluated in 20 experimental cycles at a constant 30 wt% fibre loading in a five-layer laminate structure. It was modelled with six responses: tensile strength (TS), tensile modulus (TM), flexural strength (FS), flexural modulus (FM), impact strength (IS), and Brinell hardness (BHN). All quadratic models showed exceptional fidelity (R 2 ≥ 0.9979; all p < 0.0001) with non-significant lack-of-fit ( p > 0.05). One of the key and convergent observations is the domination of the quadratic term of NaOH concentration (A 2 ) in four of the six responses: tensile strength (F = 1725.99), flexural strength (F = 8743.68), impact strength (F = 1647.55), and flexural modulus (F = 14114.83), which made the non-linearity of the concentration of NaOH the primary controlling phenomenon. Conditions that gave the best results were 5 wt% NaOH, 3 h immersion, 0° orientation, giving the best tensile strength 134.37 MPa, tensile modulus 6.14 GPa, flexural strength 100.81 MPa, impact strength 6.65 kJ/m 2 , and hardness 93.72 BHN. The worst performance degradation measured in this study, namely over-treatment (7 wt%, 7 h), was found specifically at Run 16 (TS = 70.36 MPa, TM = 1.76 GPa). The thermogravimetric analysis and the SEM fractography supported the treatment optimum identified with RSM. All RSM predictions were confirmed in experiments with an error of less than five percent (±5%).