Dimensional metrology with X-ray CT: A comparison with CMM measurements on internal features and compliant structures

X-ray computed tomography (CT) is uniquely suited for dimensional measurement of components having internal geometry, difficult-to-reach part features, and easy-to-deform or flexible structures. Through the development of standards it is also becoming accepted as a metrology tool. The objectives of this study are to compare dimensional measurements on artifacts containing internal structures and mechanically compliant features using current state-of-the-art CT and coordinate measure machine (CMM) techniques. It also aims to discuss some of the issues with recently developed standards for CT measurement uncertainty estimation. To illustrate the challenges of CMM and the potential of X-ray CT for reliable dimensional inspection, two different problems are presented: 1) characterization of internal geometry in a metallic artifact that has features inaccessible to tactile or vision-based measurement techniques; and 2) evaluation of dimensional measurement of diameters, form, and relative distances in the components of a flexible object with stems that bend due to mechanical contact by tactile instruments, e.g., CMM. In addition, some issues with the direct application of the ISO 15530 series for the estimation of CT measurement uncertainties are discussed, and a generalized formalism for uncertainty budgeting in CT metrology (still based on the ISO 15530 guidelines but with some differences) is applied to the measurements presented throughout this paper. For dimensions of geometric features ranging from 0.6 mm to 65 mm, a comparison between CT and CMM measurements typically resulted in differences of approximately 5 mu m or less for most of the measurements, while expanded uncertainties computed for the CT measurements ranged from 1 to 20 mu m. CT measurements on the internal geometries resulted in uncertainties comparable to CMM data that were obtained with the metallic artifact disassembled. However, in the particular case of measuring compliant structures, the conventional CMM strategies led to large systematic errors in the measurements. It is demonstrated that using extrapolation methods, measurements from these two systems converged to within +/- 2 mu m.