A PDE Framework for Sensing and Control of Metal Additive Manufacturing

该奖项将允许开发新的分析框架，以改善对金属增材制造过程的控制。增材制造技术已经彻底改变了设计和制造，通过从数字模型创建三维物体，可以快速定制生产和产品开发。尽管在过去的二十年里做出了巨大的努力，但在制造过程中经常出现诸如几何不精确、裂缝和微观结构不均匀等制造缺陷。这些问题主要是由多种因素引起的，包括激光束和粉末床之间的相互作用，这些因素会影响制造过程中的热历史。本研究旨在通过理论和实验研究改善功能性金属最终产品的制造，从而更好地调节和控制温度、冷却速度和熔池宽度，这两者都是打印复杂几何形状的关键因素。该项目的成功完成将对金属增材制造在航空航天和国防等高价值行业中制造功能部件的使用产生积极影响。通过全面的推广和教育计划，本科生和研究生将参与将先进制造与系统理论相结合的跨学科学习活动。现有的金属增材制造过程（如选择性激光熔化）控制方法是建立在近似常微分方程（ODE）模型或集总参数假设上的。这些简化的方法不能精确地捕捉温度场的空间分布性质，并且完全忽略了相变动力学、层几何形状的影响以及先前的激光通道。相反，具有相变动力学的预测偏微分方程（PDE）模型对于实时控制来说过于复杂。通过这项研究，我们提供了基于基本守恒定律的面向控制的PDE-ODE模型，该模型包含相变动力学，并能够实时估计和控制过程特征。具体来说，该项目旨在创建并实验验证基于PDE-ODE模型的方法，用于设计仅依赖熔面测量来估计温度分布的估计器和使用这些估计来稳健调节熔池大小和冷却速率的反馈控制器，以适应不同的层几何形状和局部热性能。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 award will allow the development of new analytical frameworks to improve the control of metal additive manufacturing processes. Additive manufacturing technologies have revolutionized design and manufacturing, allowing rapid customization of production and product development by creating three-dimensional objects from digital models. Despite substantial efforts over the past two decades, manufacturing flaws such as geometric inaccuracy, cracks, and micro-structural inhomogeneity often occur during the build. These issues primarily arise from a variety of factors, including the interaction between the laser beam and powder bed, that affect the thermal history during the manufacturing process. This research aims to improve the manufacturing of functional metallic final products through theoretical and experimental investigations that will enable better regulation and control of the temperature cooling rate and the width of the melt-pool, both are key factors in printing complex geometries. The successful completion of this project will positively impact the use of metal additive manufacturing for fabrication of functional parts in high-value industries such as aerospace and defense. Through a thorough outreach and education plan, undergraduate and graduate students will participate in interdisciplinary learning activities that couple advanced manufacturing with system theory. Existing control approaches for metal additive manufacturing processes such as selective laser melting are built on approximate ordinary differential equation (ODE) models or lumped parameter assumptions. These simplified approaches do not precisely capture the spatially distributed nature of the temperature field and completely ignore phase change dynamics, the effect of layer geometry, as well as prior laser passes. Conversely, predictive partial differential equation (PDE) models with phase change dynamics are too complex for real-time control. Through this research effort, we offer control-oriented PDE-ODE models derived from fundamental conservation laws that incorporate phase-change dynamics and enable real-time estimation and control of process signatures. Specifically, this project aims to create and experimentally validate a PDE-ODE model-based approach for the design of estimators that rely on melt-front measurements alone to estimate temperature profiles and feedback controllers that use these estimates to robustly regulate melt pool size and cooling rates for varying layer geometry and local thermal properties.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.