Automatic Feature Based Inspection and Qualification for Additively Manufactured Parts with Critical Tolerances

Abstract

This work focuses on expanding the capabilities of the Digital Additive and Subtractive Hybrid (DASH) system by extending the functionality to include some "geometric qualification" of mechanical products. Specifically, the developments for this research incorporate the extended Additive Manufacturing Format files (AMF-TOL) which include American Society of Mechanical Engineers (ASME) Y14.5 specifications for planes, cylinders and other features so that "in-process" inspection can be completed automatically as part of the hybrid process. A Renishaw touch probe is used to perform On Machine Measurements (OMM), collecting sample radii to create confidence bounds on the accuracy of size and position of finished cylindrical features. Measurements are imported to a spreadsheet where a nonlinear least squares regression algorithm creates statistical bounds on the true size and position of the cylindrical feature. These statistical bounds can then be compared to the specified tolerance callouts to determine whether a part is within specification, with a user-defined level of confidence. Seven different sampling strategies were evaluated on a DASH part to determine that four different sampling techniques, including the bird cage sampling strategy defined in ISO 12180, can be utilized with no statistically significant difference in accuracy. Special considerations for each method are discussed. Finally, analysis shows that using the DASH process with automatic inspection is economically advantageous for low volume production runs.