Researchers at Nanyang Technological University developed a fast, low-cost imaging method that can analyze the structure of 3D-printed metal parts to assess the quality of the material. The system uses an optical camera, a light source, and a notebook computer running a proprietary machine-learning software developed by the team. Most 3D-printed metal alloys are composed of a variety of microscopic crystals, which differ in shape, size, and atomic lattice orientation. Mapping this information out, one can infer the alloy’s properties, such as strength and toughness. To analyze this microstructure, typically a scanning electron microscope is used, which is an expensive and time-consuming process. The microscopes themselves cost between $75,000 and $1.5 million. Matteo Seita, assistant professor at NTU, analyzing a piece of 3D-printed alloy for strength and hardness using a prototype imaging system. Courtesy of NTU Singapore. On the other hand, the hardware involved in the team’s method costs less than $20,000. The method first treats the metal surface with chemicals to reveal the microstructure, and then places the sample facing the camera, which takes multiple images as the light source illuminates the metal from varying directions. The software then analyzes patterns produced by the light reflecting off the surface of different metal crystals, and then deduces their orientation. The entire process takes about 15 minutes. “Using our inexpensive and fast-imaging method, we can easily tell good 3D-printed metal parts from the faulty ones. Currently, it is impossible to tell the difference unless we assess the material’s microstructure in detail,” said Matteo Seita, assistant professor in NTU’s School of Mechanical and Aerospace Engineering and School of Materials Science and Engineering. Analyzing the unique crystal patterns on the surface of a 3D-printed metal may pave the way for the certification and quality assessment of parts made through additive manufacturing. Courtesy of NTU Singapore. “No two 3D-printed metal parts are created equal, even though they may have been produced using the same technique and have the same geometry. Conceptually, this is akin to how two otherwise identical wooden artefacts may each possess a different grain structure.” Seita believes that the team's imaging method has the potential to simplify the certification and quality assessment of metal alloy parts produced by 3D printing, or additive manufacturing. One of the most commonly used techniques to 3D print metal parts uses a high-powered laser to melt metal powders and fuse them together, layer by layer. However, the microstructure, and therefore the quality of the printed metal, depends on several factors. These include the speed and intensity of the laser, the time allotted for cooling before the next layer is added, and the type and brand of metal powders used. Rather than use a complex program designed to measure the crystal orientation from the optical signals acquired, the team used a neural network, feeding it hundreds of optical images from which to learn. Eventually, it learned to predict the orientation of crystals in the metal from the images, based on differences in how light scatters off the metal surface. They then tested the method to be able to create a complete “crystal orientation map,” which provides comprehensive information about the crystal shape, its size, and atomic lattice orientation. The team is currently in discussions with NTUitive, NTU’s innovation and enterprise company, to explore the possibility of starting a spinoff company or to license their patent. The research was published in npj Computational Materials (www.doi.org/10.1038/s41524-021-00688-1).