Machine learning-based performance prediction and Pareto optimization in 3D printing: a case study of PLA/Wood composites

Purpose This study uses machine learning to predict the performance of PLA/wood composites and identify suitable 3D printing parameters. The aim is to balance mechanical strength, material consumption, and processing requirements. Design/methodology/approach Experimental tensile and compressive strength data were used to develop second-order Ridge regression models. The models were validated using leave-one-out cross-validation (LOOCV) and then applied to predict mechanical performance across 3,125 combinations of five printing parameters: nozzle temperature, printing speed, layer height, infill density, and raster angle. Nominal material mass and relative processing burden were estimated for each configuration. Multi-objective Pareto optimization was used to identify trade-offs between mechanical performance and resource requirements. Findings The machine learning models achieved R2 values of 0.9954 for tensile strength and 0.9787 for compressive strength. Among the 3,125 configurations, 35 were identified as Pareto-optimal (1.12%). The selected best-compromise configuration had a nozzle temperature of 210 °C, a printing speed of 40 mm/s, a layer height of 0.15 mm, an infill density of 45%, and a raster angle of 0°. For this configuration, the models predicted a tensile strength of 22.38 MPa and a compressive strength of 33.75 MPa. Originality/value This study uses machine learning to predict the mechanical strength of PLA/wood composites and Pareto optimization to select suitable printing parameters. By evaluating more printing configurations than those tested experimentally, the method helps find a balance between mechanical strength, material use, and processing requirements.

Authors

Institutions

Publication Details

Journal
Multidiscipline Modeling in Materials and Structures
Published
2026-11-19
DOI
https://doi.org/10.1108/mmms-09-2026-0437
Primary Topic
Additive Manufacturing and 3D Printing Technologies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Machine learning-based performance prediction and Pareto optimization in 3D printing: a case study of PLA/Wood composites

Monire Norouzi
6 links
Multidiscipline Modeling in Materials and Structures
Additive Manufacturing and 3D Printing Technologies
article

Machine learning-based performance prediction and Pareto optimization in 3D printing: a case study of PLA/Wood composites

Monire Norouzi
article en

Abstract

Purpose This study uses machine learning to predict the performance of PLA/wood composites and identify suitable 3D printing parameters. The aim is to balance mechanical strength, material consumption, and processing requirements. Design/methodology/approach Experimental tensile and compressive strength data were used to develop second-order Ridge regression models. The models were validated using leave-one-out cross-validation (LOOCV) and then applied to predict mechanical performance across 3,125 combinations of five printing parameters: nozzle temperature, printing speed, layer height, infill density, and raster angle. Nominal material mass and relative processing burden were estimated for each configuration. Multi-objective Pareto optimization was used to identify trade-offs between mechanical performance and resource requirements. Findings The machine learning models achieved R2 values of 0.9954 for tensile strength and 0.9787 for compressive strength. Among the 3,125 configurations, 35 were identified as Pareto-optimal (1.12%). The selected best-compromise configuration had a nozzle temperature of 210 °C, a printing speed of 40 mm/s, a layer height of 0.15 mm, an infill density of 45%, and a raster angle of 0°. For this configuration, the models predicted a tensile strength of 22.38 MPa and a compressive strength of 33.75 MPa. Originality/value This study uses machine learning to predict the mechanical strength of PLA/wood composites and Pareto optimization to select suitable printing parameters. By evaluating more printing configurations than those tested experimentally, the method helps find a balance between mechanical strength, material use, and processing requirements.

Multidiscipline Modeling in Materials and Structures
Haliç University (TR)
Industry, innovation and infrastructure
Additive Manufacturing and 3D Printing Technologies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon