CDS&E/Collaborative Research: In-Situ Monitoring-Enabled Multiscale Modeling and Optimization for Environmental and Mechanical Performance of Advanced Manufactured Materials

这个由计算和数据驱动的科学与工程（CDS&amp；E）资助的项目将支持有助于实时控制和优化增材制造（AM）金属部件的研究，以改善其环境和机械性能。金属增材制造由于其在制造复杂几何形状方面的优异性能，已逐渐在生产高价值部件的工业中获得认可。然而，缺乏有效的过程-结构-性能（PSP）模型，特别是环境辅助失效性能，阻碍了金属增材制造的广泛应用。质量保证在很大程度上依赖于试错，这是昂贵、耗时且容易出错的。该奖项将建立一个物理约束的人工智能（PCAI）框架，以促进对增材制造过程引入的独特特征和缺陷如何影响已建成部件的环境辅助性能的基本理解。开发的工具将提供给学术界和工业界。此外，将为本科生和研究生开设新的先进制造PCAI课程，培养具有人工智能、物理模拟和先进制造技能的未来劳动力。该项目将建立一个原位加工数据驱动的框架，可以有效地将制造过程与部分规模激光粉末床熔合（L-PBF）的环境相关性能联系起来，并实现工艺优化，以改善环境辅助故障性能。技术途径包括：1)建立基于pcai的替代模型，结合原位监测数据预测局部残余应力和微观结构；2)建立基于物理的降阶模型，定量地将残余应力和微观组织与成品件的腐蚀疲劳性能联系起来；3)建立工艺优化方法，实现局部L-PBF的目标腐蚀疲劳性能。这项工作也可能适用于其他制造工艺，如直接能量沉积、生物制造和纳米制造。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This Computational and Data-Enabled Science and Engineering (CDS&E) funded project will support research that contributes to the real-time control and optimization of additive manufactured (AM) metal components to improve their environmental and mechanical performance. Metal AM has gradually gained acceptance in industries for producing high-value components, thanks to its excellent performance in fabricating complex geometries. However, the lack of efficient process-structure-performance (PSP) models, particularly for environmentally assisted failure performance, hinders broad application of metal AM. Quality assurance heavily depends on trial-and-error, which is expensive, time-consuming, and mistake-prone. This award will establish a physics-constrained artificial intelligence (PCAI) framework to promote the fundamental understanding of how the unique features and defects introduced by the AM process affect the environmentally-assisted performances of as-built parts. The developed tools will be made available to the academic and industrial communities. Furthermore, new courses of the PCAI for advanced manufacturing will be created for both undergraduate and graduate students, cultivating future workforce with skills in AI, physical simulation, and advanced manufacturing.This project will establish an in-situ processing data-driven framework that can effectively link manufacturing process to environmentally-related performance for part-scale laser powder bed fusion (L-PBF) and enable process optimization for improved environmentally-assisted failure performance. The technical approaches involve 1) Establish a PCAI-based surrogate model that can incorporate in-situ monitoring data to predict part-scale residual stress and microstructures; 2) Build a physics-based reduced-order model that can quantitatively correlate the residual stress and microstructures to the corrosion fatigue properties of as-built parts; 3) Establish a process optimization method to achieve the targeted corrosion fatigue properties for part-scale L-PBF. This work may also be applicable to other manufacturing processes such as direct energy deposition, biomanufacturing, and nanomanufacturing.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.