Page 58 - MSAM-4-3

Materials Science in

Additive Manufacturing

ORIGINAL RESEARCH ARTICLE
Explainable prediction of bead geometry in

laser-arc hybrid additive manufacturing of Al–Cu
alloy using a particle swarm optimization-based

ensemble model

Runsheng Li 1 , Hui Ma 1 , Xingwang Bai 2 , Boce Xue 1 , Changze Li 1 ,
1
4
Kui Zeng 1 , Youheng Fu 3 , Yonghui Liu * , and Yanzhen Zhang *
1 College of Mechanical and Electronic Engineering, China University of Petroleum (East China),
Qingdao, Shandong, China
2 School of Mechanical Engineering, University of South China, Hengyang, Hunan, China
3 School of Materials Science and Engineering, Huazhong University of Science and Technology,
Wuhan, Hubei, China
4 Shandong CharmRay Laser Technology Co., Ltd, Yantai, Shandong, China

Abstract

The weld bead is the basic structural unit in metal additive manufacturing, yet the
*Corresponding authors:
Yonghui Liu multiphysics coupling inherent to hybrid laser-arc processing greatly complicates
(78925411@qq.com) the prediction of bead dimensions. Despite the exploration of numerous predictive
Yanzhen Zhang methods, research on explainable prediction of weld-bead dimensions remains
(zhangyanzhen@upc.edu.cn)
limited. In this work, we developed a particle swarm optimization (PSO)-based
Citation: Li R, Ma H, Bai X, et al. ensemble prediction model (PSO-EP) for laser-arc hybrid additive manufacturing,
Explainable prediction of bead
geometry in laser-arc hybrid and through SHapley Additive exPlanations (SHAP) analysis, comprehensively
additive manufacturing of Al–Cu uncovered the underlying links between process variables and bead geometry.
alloy using a particle swarm Experimental evidence indicated that our PSO-EP outperformed individual models
optimization-based ensemble
model. Mater Sci Add Manuf. and alternative ensembles, delivering superior accuracy, reflected by an R-squared
2025;4(3):025220036. value of 0.9567 for bead width and an R-squared value of 0.9492 for bead height,
doi: 10.36922/MSAM025220036 and markedly lowering prediction errors. The SHAP findings indicated that weld
Received: May 26, 2025 speed is the dominant determinant of bead width, while laser power plays a pivotal
Revised: June 16, 2025 role in bead height. Subsequent single-factor dependence analysis showed that
different process variables had significantly different impacts on bead size across
Accepted: June 17, 2025
their respective value intervals. This study provides important theoretical support for
Published online: July 17, 2025 the intelligent development of the laser-arc hybrid additive manufacturing process.
Copyright: © 2025 Author(s).
This is an Open-Access article
distributed under the terms of the Keywords: Additive manufacturing; Ensemble learning; Laser-arc hybrid; Geometry
Creative Commons Attribution prediction; Aluminum–copper alloys; Explainable analysis
License, permitting distribution,
and reproduction in any medium,
provided the original work is
properly cited.
Publisher’s Note: AccScience 1. Introduction
Publishing remains neutral with Metal additive manufacturing (MAM), commonly referred to as metal 3D printing,
regard to jurisdictional claims in
published maps and institutional is known for its design versatility and high efficiency, and is therefore extensively
1
affiliations. employed in aerospace, automotive, shipbuilding, and energy industries. Through

Volume 4 Issue 3 (2025) 1 doi: 10.36922/MSAM025220036