Answers are grounded in this paper and cite their sources.
Additive manufacturing, specifically Fused Deposition Modelling (FDM), enables the fabrication of complex geometries with high precision. Polylactic acid (PLA), a biodegradable polymer, is widely utilised in FDM, and its mechanical properties can be enhanced by reinforcing it with natural fillers. This study investigates a PLA biocomposite reinforced with 10% coconut shell powder, with a primary focus on optimising surface roughness as a critical indicator of overall product quality. Three machine learning regression models, Support Vector Regression (SVR), Random Forest Regression (RFR), and Gradient Boosting Regression (GBR), were developed to predict surface roughness, and their goodness of fit to the 27-run full-factorial dataset was compared descriptively. The effects of key process parameters, including infill density, nozzle temperature and layer thickness, were analysed using surface and contour plots. The fitted SVR model was subsequently used as a smooth surrogate objective within Bayesian optimisation to identify a model-predicted optimum. The optimisation predicted a surface roughness of 2.74 μm at a nozzle temperature of 203.55 °C, a layer thickness of 0.10 mm and an infill density of 60%. The surface roughness of the confirmatory specimen was 2.81 μm, comparable to the best experimental condition (Run 3, 2.77 μm). The result identifies a low-roughness region. This study applies machine learning and Bayesian optimisation to identify low-roughness printing conditions for a PLA composite containing a nominal 10 wt% agricultural waste-derived coconut shell powder.
No takes yet. Share an insight, caveat, or question.
Sarma et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: