Praniewicz M, Ameta G, Fox J, Saldana C
Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30318, United States.
Siemens Corporate Research, Princeton, NJ 08540, United States.
Addit Manuf. 2022 Oct;35. doi: 10.1016/j.addma.2020.101292.
This work refines surface registration methods for metrological datasets to improve the multi-method qualification accuracy of additively manufactured (AM) lattices. Datasets acquired from X-ray computed tomography and a coordinate measurement machine of an AM lattice were aligned using derived geometry datum features based on a theoretical supplemental surface definition, which has been established in recent draft standards, but has had limited examination using complex AM structures. A refined sampling registration approach for lattice geometry based on spatially-dependent subsampling is derived and shown to statistically decrease variation between measurement sources. This importance of well-defined sampling practice and definition is highlighted. The applicability of this approach for multi-method qualification of complex AM parts is discussed. This work lays the foundation of utilizing specifications under consideration in a new standard with possible verification techniques that can be employed.
这项工作改进了计量数据集的表面配准方法,以提高增材制造(AM)晶格的多方法鉴定精度。基于在最近的标准草案中建立的理论补充表面定义,利用派生的几何基准特征对从X射线计算机断层扫描和AM晶格的坐标测量机获取的数据集进行对齐,不过使用复杂AM结构对其进行的检验有限。推导了一种基于空间相关子采样的晶格几何细化采样配准方法,并证明该方法能在统计上减少测量源之间的差异。强调了定义明确的采样实践和定义的重要性。讨论了该方法在复杂AM零件多方法鉴定中的适用性。这项工作奠定了利用新标准中正在考虑的规范以及可采用的可能验证技术的基础。