Next-generation computer-aided inspection technologies

在制造过程中，检验是一项至关重要的工作，不仅是为了验证制造零件的质量，也是为了为过程控制提供必要的反馈。如果不能快速准确地测量所制造的零件，就不可能保持生产精度以最大限度地减少废品。在今天的先进制造车间，检查过程是使用计算机辅助技术进行的，而不是由经验丰富的检查员进行传统的硬测量和目测评估。几十年前，计算机辅助检测(CAI)技术的引入(当时主要限于坐标测量机上的触发式探头)的引入，极大地提高了检测过程的速度和精度，为制造商打开了一个全新的世界。随着现代制造工艺的发展，对更先进的计算机辅助检测技术的需求越来越大，以跟上新的制造实践。今天，随着加法制造(AM)的发展，先进制造中出现了新的范式。具有极其复杂的几何形状和功能内部结构的部件可以通过传统的减法制造方法制造出永远无法实现的部件。然而，在航空航天等受监管的行业中，AM扩张的最严重障碍之一是部件资格问题。具有复杂几何形状的附加制造零件具有广泛的检测需求，而当前的测量系统和数据分析技术尚未满足这些需求。一旦零件完全制造完成，它的内部几何形状就无法接触，也不可能通过依赖表面测量的传统检测工具进行测量。我在这里提出了一个为期5年的研究计划，其主要理念是测量AM零件最有效和最高效的方法是过程测量，即在零件制造时对每一层进行检查。这种方法能够逐层检测缺陷以及评估内部几何误差。此外，这种逐层检查能够以纠正措施的形式在每一层进行现场过程控制。现场过程控制是将AM部件引入航空航天等受监管行业的关键。拟议的研究计划将开发用于在线形状测量的工具，以及新的算法，以实现逐层自动检测缺陷和评估内部几何误差。拟议的研究计划具有很高的兴趣，因为它将缩小金属AM工艺与受监管行业(如航空航天)的精度要求之间的差距。此外，该计划将培训下一代工程师，使AM在加拿大取得商业成功。**

Inspection in manufacturing is a crucial exercise, not only to verify the quality of the manufactured part, but also to provide the necessary feedback for process control. Without fast and accurate measurement of the manufactured part, maintaining production precision to minimize scrap parts is not possible. On today's advanced manufacturing shop floors, the inspection process is performed using computer-aided technologies, as opposed to traditional hard gauges and visual evaluation by experienced inspectors. A few decades ago, the introduction of computer-aided inspection (CAI) technologies (which at that time was mainly limited to the touch-trigger probes on coordinate measuring machines) opened a whole new world for manufacturers by greatly increasing the speed and accuracy of the inspection process. With the modern manufacturing processes, there is an ever-increasing need for more sophisticated computer-aided inspection technologies that can catch up with the new manufacturing practices. Today, with the advancement of additive manufacturing (AM), a process in which a part is made layer by layer, new paradigms are emerging in advanced manufacturing. Parts with extremely complex geometries and functional internal structures can be made that could never be achievable by the means of traditional subtractive manufacturing. However, in regulated industries such as aerospace, one of the most serious hurdles to the expansion of AM is the question of part qualification. The additively manufactured parts with complex geometries have a wide variety of inspection needs that are not yet addressed by current measurement systems and data analysis techniques. Once the part is completely manufactured, its internal geometry is inaccessible and impossible to be measured by conventional inspection tools relying on surface measurements. I propose here a 5-year research program with the main philosophy that the most effective and efficient way of measuring AM parts is in-process measurement in which each layer of the part is inspected as it is built. This approach enables defect detection as well as evaluation of internal geometric errors on a layer-by-layer basis. Moreover, such layer-wise inspection enables in-situ process control at each layer in the form of corrective actions. In-situ process control is the key to the introduction of AM parts in regulated industries such as aerospace. The proposed research program will develop tools for in-process shape measurement, as well as novel algorithms to enable automatic defect detection and evaluation of internal geometric errors on a layer-by-layer basis. The proposed research program is of high interest since it will close the gap between metal AM process and the precision requirements of the regulated industries such as aerospace. Furthermore, the program will train the next generation of engineers who enable the commercial success of AM in Canada.**