Pulling at the Digital Thread: Exploring the Tolerance Stack Up Between Automatic Procedures and Expert Strategies in Scan to Print Processes

Abstract

While the combination of 3D scanning and printing processes holds much promise for the field of new product development, problems with repeatability and accuracy have limited the wider spread adoption of some digital prototyping tools, such as 3D scanners. Studies have explored the errors inherent in higher fidelity scan to print (S2P) processes, yet few have explored the errors in S2P processes that leverage affordable rapid noncontact scanners. Studies have yet to explore the strategies that designers, who are experienced with additive manufacturing, employ to mitigate errors. To address these gaps, a controlled study was conducted using data from 27 scans collected with a prototypical off-the-shelf noncontact optical scanner. The geometric and dimensional integrity of the digital models was found to be significantly out of tolerance at various phases of the S2P process, as compared to the original physical model. Larger errors were found more consistently in the data acquisition phase of the S2P process, but results indicate these errors were not sufficiently filtered out during the remainder of the process. A behavioral study was conducted with 13 experienced designers in digital fabrication to determine strategies for manually cleaning Point Clouds. Actions such as increase or decrease in brush size and select or de-select points were recorded. These actions were analyzed using hidden Markov modeling, which revealed distinct patterns of behavior. Designer strategies were not beneficial and digital models produced by designers were found to be significantly smaller than original physical models.