Scholarlink

The increasing adoption of additive manufacturing (AM), commonly known as 3D printing, across various industries like aerospace, automotive, energy, and healthcare has brought numerous advantages. However, it also presents certain challenges, and one of the significant limitations of AM is related to the surface integrity of printed parts. Fused deposition modeling (FDM) is a widely used additive manufacturing (3D printing) method where material is deposited layer by layer to form a three-dimensional object, following the design specified in a digital 3D model created using computer-aided design (CAD) software. In the present work, to address the limitation of surface integrity, a data-driven predictive modeling approach for forecasting surface roughness in AM components in FDM has been introduced. This approach aims to enhance the quality of printed components by predicting and controlling their surface roughness. Various machine learning techniques were employed in a systematic study to explore the impacts of different printing parameters on the surface roughness of 3D printed parts. An ensemble of machine learning models was then utilized to develop a predictive model for surface roughness, aiming to attain the desired surface quality. The promising outcome is the validation of the predictive model using data obtained from a series of additive manufacturing tests carried out on an FDM printer. This validation demonstrates the model’s capability to accurately predict the surface roughness of 3D printed components.

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

Application of Machine Learning Ensemble Methods for Prediction of Surface Roughness for Fused Deposition Modeling Processed Parts

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • P. Kiranmayi,
  • Taj,
  • M. Hymavathi,
  • V. Vijaya babu,
  • P. V. Vinay,
  • Ch. Himagireesh

摘要

The increasing adoption of additive manufacturing (AM), commonly known as 3D printing, across various industries like aerospace, automotive, energy, and healthcare has brought numerous advantages. However, it also presents certain challenges, and one of the significant limitations of AM is related to the surface integrity of printed parts. Fused deposition modeling (FDM) is a widely used additive manufacturing (3D printing) method where material is deposited layer by layer to form a three-dimensional object, following the design specified in a digital 3D model created using computer-aided design (CAD) software. In the present work, to address the limitation of surface integrity, a data-driven predictive modeling approach for forecasting surface roughness in AM components in FDM has been introduced. This approach aims to enhance the quality of printed components by predicting and controlling their surface roughness. Various machine learning techniques were employed in a systematic study to explore the impacts of different printing parameters on the surface roughness of 3D printed parts. An ensemble of machine learning models was then utilized to develop a predictive model for surface roughness, aiming to attain the desired surface quality. The promising outcome is the validation of the predictive model using data obtained from a series of additive manufacturing tests carried out on an FDM printer. This validation demonstrates the model’s capability to accurately predict the surface roughness of 3D printed components.