GOALI: Online Defect Detection and Mitigation Method for Incipient Anomalies in Additive Manufacturing Processes

与传统制造相比，加法制造具有显著的优势，有可能从根本上改变各种行业的最先进水平。尽管当前添加剂制造技术取得了巨大进步，但某些棘手的质量问题仍然存在。这导致大量返工和高废品率，从而对添加剂制造的可持续性构成重大障碍。因此，迫切需要改进添加剂制造工艺中缺陷检测的在线方法，以便能够在制造的早期阶段识别并可能防止初期工艺异常。这一学术联系机会(GOALI)研究项目预计将显著推进添加剂制造中的过程监测和控制技术，从而改善产品质量、提高过程生产率和更高的盈利能力。因此，这项研究的结果将产生重大的社会经济影响。这项研究的科学发现可以推广到许多其他先进的制造工艺。此外，该项目还包括许多教育组成部分，如新的课程模块，以及为本科生提供的研究经验。让学生接触到这种多学科的研究将培养出一支拥有先进制造技术的多样化和合格的劳动力队伍。本研究的目标是通过将新颖的多现象传感技术与先进的传感器数据融合分析方法相结合来解决添加剂制造中的关键质量问题：(1)各种过程属性的动态行为是什么，以及这种行为如何导致添加剂制造过程异常的发生？以及(2)加法制造质量表现和过程变量之间的因果联系是什么？研究的目的包括：(1)定量地阐明添加制造中工艺异常与从在线时空传感器信号中提取的特征之间的基本关系，即实现传感器特征与添加制造中不断演变的零件缺陷之间的映射，从而能够及早发现与表面形貌相关的缺陷；(2)建立添加制造过程中工艺条件/设置与连续和属性产品质量变量之间的关联模型，并识别对产品质量有显著影响的工艺变量，为未来添加制造系统中的缺陷缓解和实施闭环控制提供有价值的策略。这项研究将利用基于预测的过程监控来处理添加剂制造过程中潜在的复杂性和不确定性。

Additive manufacturing offers significant advantages over conventional manufacturing, with potential to fundamentally transform the state-of-the-art in a variety of industries. Notwithstanding the enormous progress in current additive manufacturing technologies, certain intractable quality issues persist. This leads to considerable rework and high scrap rates, and thus poses significant impediments for sustainability of additive manufacturing. Consequently, there is a vital need to advance online methods for defect detection in additive manufacturing processes, so that incipient process anomalies can be identified, and possibly prevented, at an early stage during manufacture. This Grant Opportunity for Academic Liaison with Industry (GOALI) research project is anticipated to significantly advance the process monitoring and control technology in additive manufacturing, leading to improved product quality, enhanced process productivity, and higher profitability. Thus, outcomes from this research will have substantial socioeconomic impacts. The scientific findings from this research are extensible to many other advanced manufacturing processes. Furthermore, this project also includes many educational components, such as new course modules, and research experience for undergraduate students. Exposing students to this multidisciplinary research will cultivate a diverse and qualified workforce possessing the state-of-the-art technologies in advanced manufacturing.The goal of this research is to resolve critical quality issues in additive manufacturing by addressing two fundamental research questions based on an integration of novel multi-phenomena sensing techniques with advanced analytical approaches for sensor data fusion: (1) what is the dynamic behavior of various process attributes, and how does this behavior cause the onset of additive manufacturing process anomalies? and (2) what are the causal linkages between additive manufacturing quality performance and process variables? The objectives of the research include: (1) quantitatively elucidate the fundamental relationships that connect process abnormalities in additive manufacturing with features extracted from online spatiotemporal sensor signals, i.e., achieve a mapping between the sensor features with the evolving part defects in additive manufacturing, enabling early detection of surface topography related defects; and (2) establish a model correlating process conditions/settings with both continuous and attribute product quality variables in additive manufacturing processes, and identify the process variables which have significant effects on product quality, providing valuable strategies for defect mitigation and implementation of close loop control in future additive manufacturing systems. This research will utilize prediction based process monitoring to tackle underlying complexity and uncertainty in additive manufacturing processes.