Scholarlink

<p>Laser powder bed fusion of metals using a laser beam (PBF-LB/M) is a complex process that requires a deep understanding of the temperature history and its sensitivities to uncover structure–property–processing relationships. This challenge is addressed by integrating the HYPercomplex-based automatic differentiation finite element method (HYPAD–FEM) with key characteristics of the PBF-LB/M process for modeling and fast sensitivity analysis. This integration facilitated the efficient computation of temperature sensitivities with respect to laser parameters, initial and boundary conditions, material properties, and geometry, enabling the generation of ranked sensitivity figures. These figures were used to identify the most influential parameters affecting the thermal history of a part build, such as powder layer thickness, the laser’s initial location, power, and speed, providing insights for refining the PBF-LB/M process. Using this model, HYPAD–FEM calculated 30 sensitivities in only an additional 1.09 × compute time relative to the time required to compute the temperature history alone. This means a speedup of 24X computation time compared to the corresponding sensitivities obtained with the finite difference method. Furthermore, HYPAD–FEM enabled the efficient development of a reduced order model, which provided guidelines for optimizing process parameters, thus enhancing the quality of the simulated part build. Consequently, HYPAD–FEM emerges as an effective tool to study complex phenomena in PBF-LB/M, facilitating the identification of influential parameters, creating models to identify mechanisms for quality improvement in build parts, and helping to uncover structure–property–processing relationships.</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

Sensitivity analysis for the study and enhancement of powder bed fusion of metals via computational models and hypercomplex automatic differentiation

  • 期刊论文

  • 发表日期：

    2025年5月5日

查看全文
  • Juan-Sebastian Rincon-Tabares,
  • Mauricio Aristizabal,
  • Arturo Montoya,
  • Harry Millwater,
  • David Restrepo

摘要

Laser powder bed fusion of metals using a laser beam (PBF-LB/M) is a complex process that requires a deep understanding of the temperature history and its sensitivities to uncover structure–property–processing relationships. This challenge is addressed by integrating the HYPercomplex-based automatic differentiation finite element method (HYPAD–FEM) with key characteristics of the PBF-LB/M process for modeling and fast sensitivity analysis. This integration facilitated the efficient computation of temperature sensitivities with respect to laser parameters, initial and boundary conditions, material properties, and geometry, enabling the generation of ranked sensitivity figures. These figures were used to identify the most influential parameters affecting the thermal history of a part build, such as powder layer thickness, the laser’s initial location, power, and speed, providing insights for refining the PBF-LB/M process. Using this model, HYPAD–FEM calculated 30 sensitivities in only an additional 1.09 × compute time relative to the time required to compute the temperature history alone. This means a speedup of 24X computation time compared to the corresponding sensitivities obtained with the finite difference method. Furthermore, HYPAD–FEM enabled the efficient development of a reduced order model, which provided guidelines for optimizing process parameters, thus enhancing the quality of the simulated part build. Consequently, HYPAD–FEM emerges as an effective tool to study complex phenomena in PBF-LB/M, facilitating the identification of influential parameters, creating models to identify mechanisms for quality improvement in build parts, and helping to uncover structure–property–processing relationships.