Pulling at the digital thread: Exploring the tolerance stack up in scan to print processes

The rise of affordable rapid non-contact digitizers and rapid prototyping tools, such as 3D printers, is enabling the seamless integration of geometric reverse engineering into the early phases of engineering design. Scanning technology has been widely adopted in bio-reverse engineering and the use of high fidelity non-contact scanners, such as Computed Tomography devices, allows designers, doctors, and researchers to digitally model boney structures, design orthotic and prosthetic devices, and preemptively plan complex surgeries. While the combination of 3D scanning and printing processes holds much promise for the fields of reverse engineering, biodesign, and new product development, problems with repeatability, accuracy, and precision have limited the wider spread adoption of 3D scan to print processes. While some studies have explored the errors inherent in higher fidelity scan to print (S2P) processes, no studies have explored the errors in S2P processes that leverage affordable rapid non-contact digitizers. The purpose of this study was to explore at which phases of the S2P process errors are introduced into the digital model. A controlled study was conducted using data from 27 scans using a common off-the-shelf non-contact optical digitizer and a relatively simple workpiece. Data from the digital thread was collected between each phase of the S2P process and compared against a truth model; the geometric and dimensional integrity of the data was calculated through a comparison between the digital model and the original truth model. Results indicate significant differences between digital models at the various steps of the S2P process.