Oak Ridge National Laboratory (ORNL) has released comprehensive additive manufacturing datasets for public use. These datasets aim to enhance the evaluation and quality control of 3D printed parts through in-process measurements, reducing reliance on post-production testing.
The data, gathered over a decade at ORNL’s Manufacturing Demonstration Facility, encompasses various 3D printing processes, materials, and controls. The latest 230-gigabyte dataset includes design, printing, and testing details of parts created using a laser powder bed fusion system. This dataset features machine health sensor data, laser scan paths, 30,000 powder bed images, and 6,300 tensile strength tests.
Traditionally, quality control in additive manufacturing has involved expensive techniques like destructive testing or X-ray computed tomography, which are often impractical for large parts. ORNL’s datasets offer an alternative by enabling machine learning models to predict part performance from in-process measurements. This approach can reduce errors in predicting tensile strength by 61%.
The datasets, now freely accessible online, support industry-scale additive manufacturing by linking manufacturing intent with outcomes, helping to determine when additional testing is necessary. This release is part of DOE’s Advanced Materials and Manufacturing Technology Program, aimed at advancing reliable and economical nuclear energy through smart manufacturing approaches.
Source: ornl.gov
