Scholarlink

<p>In this study, a machine learning-based framework was developed to predict surface roughness and identify energy-efficient process parameters in sustainable additive manufacturing while maintaining acceptable part quality. The study focused on AlSi10Mg aluminum alloy powder. Key process parameters, including laser power, scanning speed, and hatch distance, were used to calculate the volumetric energy density (VED), which was then used as input to the machine learning model. The model predicts surface roughness, a critical indicator of manufacturing quality. Four different model types—linear regression, random forest regression, Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN)—were evaluated sequentially. The results demonstrated that classical machine learning models, particularly the random forest model implemented using the Scikit-learn library, achieved the highest predictive performance, with an R² value of 0.876, and provided reliable parameter optimization. Furthermore, the optimized parameters enabled a reduction in VED-related process energy input of approximately 10.6% without compromising surface quality, with an average deviation of approximately 1.7% between optimized and original surface roughness values. This study highlights the potential of machine learning to enhance energy efficiency in additive manufacturing processes, contributing to advancements in sustainable manufacturing and offering insights applicable to other manufacturing technologies and materials. Consequently, the proposed model demonstrates adaptability to different manufacturing conditions and materials and, with further research, can be developed into an effective design tool for VED-based process energy optimization in various additive manufacturing applications. The primary focus of this study is to predict surface roughness using machine learning architectures to identify optimal SLM process parameters that minimize energy input without compromising part quality. By establishing a direct link between volumetric energy density and surface finish, the proposed framework actively guides parameters toward eco-efficient additive manufacturing. The proposed framework supports sustainable manufacturing objectives by improving resource efficiency and reducing VED-related process energy input.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning based surface roughness prediction and volumetric energy density based process optimization for sustainable selective laser melting of AlSi10Mg

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月2日

查看全文
  • Cem Batur,
  • İnayet Burcu Toprak

摘要

In this study, a machine learning-based framework was developed to predict surface roughness and identify energy-efficient process parameters in sustainable additive manufacturing while maintaining acceptable part quality. The study focused on AlSi10Mg aluminum alloy powder. Key process parameters, including laser power, scanning speed, and hatch distance, were used to calculate the volumetric energy density (VED), which was then used as input to the machine learning model. The model predicts surface roughness, a critical indicator of manufacturing quality. Four different model types—linear regression, random forest regression, Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN)—were evaluated sequentially. The results demonstrated that classical machine learning models, particularly the random forest model implemented using the Scikit-learn library, achieved the highest predictive performance, with an R² value of 0.876, and provided reliable parameter optimization. Furthermore, the optimized parameters enabled a reduction in VED-related process energy input of approximately 10.6% without compromising surface quality, with an average deviation of approximately 1.7% between optimized and original surface roughness values. This study highlights the potential of machine learning to enhance energy efficiency in additive manufacturing processes, contributing to advancements in sustainable manufacturing and offering insights applicable to other manufacturing technologies and materials. Consequently, the proposed model demonstrates adaptability to different manufacturing conditions and materials and, with further research, can be developed into an effective design tool for VED-based process energy optimization in various additive manufacturing applications. The primary focus of this study is to predict surface roughness using machine learning architectures to identify optimal SLM process parameters that minimize energy input without compromising part quality. By establishing a direct link between volumetric energy density and surface finish, the proposed framework actively guides parameters toward eco-efficient additive manufacturing. The proposed framework supports sustainable manufacturing objectives by improving resource efficiency and reducing VED-related process energy input.