Dynamical Systems Diagnostics for Intelligent Machine Tools

这笔赠款将支持有助于获得与制造过程相关的新知识的研究，既促进科学进步，又促进国家繁荣。减法制造工艺通过在小层中去除材料来制造三维零件和物体。机械加工是一种比较普遍的减法加工，包括铣削、钻孔、拉削和车削等材料去除方法。这些加工过程中的每一个都使用切割工具从散装材料中去除材料层，并留下所需的三维零件。这些材料去除工艺也是为汽车、航空航天、医疗器械行业制造金属、木材和塑料部件的最广泛使用的方法之一。然而，所有加工过程的精度、表面质量和生产率都受到切割过程中产生的振动的限制。尽管最近的进展揭示了机械加工过程中这些障碍背后的基本物理原理，但在最好的学术实验室和生产环境中仍然存在着巨大的差距。该项目旨在开发诊断工具，使制造商能够利用最新的学术知识来解决振动问题，如稳定性限制、表面光洁度和表面位置误差。因此，这笔赠款的结果将使美国经济和社会受益。这项研究涉及并影响到多个学科，包括动力系统与控制、制造、机器学习和数据科学。预计这项研究以及补充的教育努力将有助于培训未来的劳动力，并扩大未被充分代表的群体在STEM学科中的参与。刀具振动对工业能力造成了严重限制，例如精度降低、表面光洁度差以及与不稳定相关的成本增加。尽管过去的研究已经发现了导致不稳定的基本机制，但由于需要重复、昂贵和手动密集的模态测试和单独的切削力测试，美国工业基础仍然几乎不可能应用这一知识。这项研究将开发两种数据驱动的方法，使美国工业基地能够将加工动力学与现代网络基础设施相结合。更具体地说，第一个研究目标开发了一种新的方法来自动识别预测加工动力学模型和分析工具所需的物理参数。这将使现代网络基础设施能够优化切割过程参数，从而做出更好的决定，现在包括振动施加的限制。第二个研究目标是开发一种数据驱动的方法来发现包含时滞的系统的控制方程。预计该框架将提供对加工动力学模型中包含的重要物理机制的更全面的了解；该方法还可以生成现代网络基础设施可用来获得最佳切割过程参数或监控切割过程以诊断模型参数变化的问题的模型。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This grant will support research that will contribute to new knowledge related to a manufacturing process, promoting both the progress of science and advancing national prosperity. Subtractive manufacturing processes make three-dimensional parts and objects by removing material in small layers. Machining is one of the more pervasive subtractive processes and includes material removal methods known as milling, drilling, broaching, and turning. Each of these machining processes uses a cutting tool to remove material layers from a bulk material and to leave behind a desired three-dimensional part. These material removal processes are also one of the most widely used approaches to create metal, wood, and plastic parts for the automotive, aerospace, medical device industries. However, the accuracy, surface quality, and productivity of all machining processes are limited by the vibrations caused in the cutting process. Although recent advances have unveiled the fundamentals physics behind these barriers for machining processes, there still exist a large gap between what is possible in the best academic lab and in a production setting. This project seeks to develop diagnostic tools that will enable manufacturers to take advantage of the latest academic knowledge for vibration problems, such as stability limitations, surface finish, and surface location error. Therefore, the results of this grant will benefit the U.S. economy and society. This research involves and impacts several disciplines which include dynamical systems and control, manufacturing, machine learning, and data science. It is expected that this research, along with the complementary educational efforts, will help train the future workforce and broaden the participation of underrepresented groups in STEM disciplines. Tool vibrations impose severe limitations on industrial capability, such as reduced accuracy, a poor surface finish, and increased costs which are linked to instability. Although past research has uncovered the fundamental mechanism that leads to instability, it is still nearly impossible for the U.S. industrial base to apply this knowledge due to the need for repetitive, costly, and manually intensive modal tests and separate cutting force tests. This research will develop two data-driven approaches that will enable the U.S. industrial base to integrate machining dynamics with modern cyber-infrastructure. More specifically, the first research objective develops a new approach to automate the identification of the physical parameters required by predictive machining dynamics models and analysis tools. This will enable modern cyber infrastructure to optimize cutting process parameters and thus allow better decision to be made that now include the limitations imposed by vibrations. The second research objective develops a data-driven approach to discover the governing equations of systems that include time delays. It is expected that this framework will provide a more comprehensive understanding of the important physical mechanisms to include in machining dynamics models; this method could also generate models that modern cyber infrastructure could use to obtain optimal cutting process parameters or monitor the cutting process to diagnose problems from model parameter changes.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.