Machine Learning-Assisted Processing Maps for Selective Laser Melted AlSi10Mg Alloy Optimization

Abstract This study investigates the application of machine learning (ML) models to optimize the selective laser melting (SLM) process for AlSi10Mg alloy by establishing processing windows that simultaneously achieve high relative density, low surface roughness, and high hardness. Gaussian Process Regression (GPR), Artificial Neural Network (ANN), and Random Forest (RF) models were trained using experimental data to predict the relationships between laser power, scan speed, and the resulting material properties. The trained models were used to construct processing parameter maps, three-dimensional property surfaces, and contour maps for identifying suitable processing conditions. The results revealed that laser power and scan speed significantly influence all three properties, with scan speed exhibiting a pronounced effect on the predicted responses. For the combined targets of relative density above 99%, surface roughness below 12 µm, and hardness above 120 HV, the predicted processing windows were 225–275 W and 500–1600 mm s⁻¹ for GPR, 175–275 W and 800–1600 mm s⁻¹ for ANN, and 150–300 W and 900–1600 mm s⁻¹ for RF. The experimentally identified processing window was 175–300 W and 700–1500 mm s⁻¹, demonstrating good agreement with the model predictions. The corresponding volumetric energy density ranges were 57–167 J mm⁻³ for GPR, 33–104 J mm⁻³ for ANN, 28–101 J mm⁻³ for RF, and 35–129 J mm⁻³ for the experimental window. The results demonstrate the potential of ML-assisted processing maps for identifying multi-property optimization windows, guiding experimental parameter selection, and reducing reliance on extensive trial-and-error testing.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-13
DOI
https://doi.org/10.21203/rs.3.rs-11389448/v1
Primary Topic
Additive Manufacturing Materials and Processes
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Machine Learning-Assisted Processing Maps for Selective Laser Melted AlSi10Mg Alloy Optimization

Yallappa Shiralgi, Sana Ben Moussa, Abdullah Yahya Abdullah Alzahrani, Joseph Kavitha et al.
1 links
Research Square
Additive Manufacturing Materials and Processes
preprint

Machine Learning-Assisted Processing Maps for Selective Laser Melted AlSi10Mg Alloy Optimization

Yallappa Shiralgi, Sana Ben Moussa, Abdullah Yahya Abdullah Alzahrani, Joseph Kavitha, Nur Insyirah Mohd Nadzri, Sudha Joseph, Brajesh Parashar
preprint en

Abstract

Abstract This study investigates the application of machine learning (ML) models to optimize the selective laser melting (SLM) process for AlSi10Mg alloy by establishing processing windows that simultaneously achieve high relative density, low surface roughness, and high hardness. Gaussian Process Regression (GPR), Artificial Neural Network (ANN), and Random Forest (RF) models were trained using experimental data to predict the relationships between laser power, scan speed, and the resulting material properties. The trained models were used to construct processing parameter maps, three-dimensional property surfaces, and contour maps for identifying suitable processing conditions. The results revealed that laser power and scan speed significantly influence all three properties, with scan speed exhibiting a pronounced effect on the predicted responses. For the combined targets of relative density above 99%, surface roughness below 12 µm, and hardness above 120 HV, the predicted processing windows were 225–275 W and 500–1600 mm s⁻¹ for GPR, 175–275 W and 800–1600 mm s⁻¹ for ANN, and 150–300 W and 900–1600 mm s⁻¹ for RF. The experimentally identified processing window was 175–300 W and 700–1500 mm s⁻¹, demonstrating good agreement with the model predictions. The corresponding volumetric energy density ranges were 57–167 J mm⁻³ for GPR, 33–104 J mm⁻³ for ANN, 28–101 J mm⁻³ for RF, and 35–129 J mm⁻³ for the experimental window. The results demonstrate the potential of ML-assisted processing maps for identifying multi-property optimization windows, guiding experimental parameter selection, and reducing reliance on extensive trial-and-error testing.

Research Square
Universiti Malaysia Perlis (MY), King Khalid University (SA)
Additive Manufacturing Materials and Processes
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