The growth in digital manufacturing has intensified the need for fast, reliable quality control methods to assess the microstructures of critical metal parts, like those used in aerospace. While 3D scanners and other technologies are available to analyze the build quality and integrity of metal parts, there is no equivalent solution for imaging and analyzing the microstructure of a part. A new approach to analyzing microstructures, based on directional reflectance microscopy (DRM), enables direct microstructure imaging and measurement of entire parts, and provides microstructure information for curved, complex surfaces. The method, developed by researchers at the University of Cambridge and Nanyang Technological University, is fast, accurate, and can be used to characterize microstructures at a fraction of the cost of scanning electron microscopy (SEM) based on electron diffraction — the current gold standard for microstructure analysis. The faulty microstructure of a jet engine turbine blade is revealed by using directional reflectance microscopy (DRM). Rather than a single crystal, the blade comprises two crystals with different orientations (i.e., different colors). Such a manufacturing defect compromises the performance of the blade at high temperature and thus prevents its use. Courtesy of Chenyang Zhu. In addition to costing a lot, SEM prohibits the direct analysis of entire parts because of the small size of the vacuum chamber. This limits microstructure analysis to small-sized, flat samples that must be extracted from the metal parts. Due to the high cost and low scalability of these measurements, industry traditionally has leaned toward conservative microstructure estimates to minimize safety concerns. “Indeed, being able to directly measure the microstructure of metal parts rather than relying on estimates may provide an opportunity to relax safety factors around part performance and life predictions,” professor Matteo Seita, who led the research, said. “These benefits translate into a longer use of the metal parts produced and thus into a more efficient and sustainable use of resources.” The DRM-based technique uses an optical camera and a rotating source of white light to assess the microstructure of metal parts. The revolving white light illuminates the surface of the metal part from different directions. After the metal surface is etched with chemical reagents, the reflected light intensity is measured by the optical camera. The measurements from the camera are fed to image analysis algorithms. The microstructure of a metal part is composed of crystal grains that differ in size, shape, and crystal lattice orientation. The algorithms used with DRM extrapolate the orientation of the crystal grains in the microstructure. This information is used to reconstruct the microstructure of the metal part. Depending on the manufacturing processes used to make a part, its microstructure will vary. The microstructure of a metal part determines properties like strength, fracture toughness, and resistance to degradation. Because the properties of a metal part are linked to its microstructure, it is crucial to assess microstructure when a part is produced and while it is still in use. This allows quality control to certify the safety of a metal component or make an informed decision about whether the part has reached end of life. A curved sample of polycrystalline aluminum seen through DRM. In contrast to conventional characterization techniques, DRM enables microstructure analysis on non-flat surfaces. Courtesy of Chenyang Zhu. The researchers capitalized on the large field of view and depth of focus of DRM to extend the crystallographic measurements to curved, 3D surfaces. To enable DRM to provide microstructural information directly from the complex, non-flat surfaces of life-size metal components, the team decoupled the optical signal generated by the microstructure from the signal produced by the variable surface of the metal part. “This is a game-changer in the field of nondestructive analysis,” Seita said. “There is no need to dissect metal components into small, flat specimens so that they can fit into the electron microscope. The material’s microstructure can be imaged directly onto the curved surface of the metal part.” The researchers provided a rigorous quantification of the accuracy of their measurements with experiments and simulations. They demonstrated their method on research samples and on a turbine blade with a bicrystalline microstructure. Because DRM captures the nuanced reflectance generated by multiple surface features that characterize the etch-induced surface of a metal, it allows for greater accuracy. The team’s analysis showed that DRM could yield reliable measurements up to a surface tilt of 20°, both when analyzing pure metals like aluminum and complex metal alloys. “We believe that DRM could open a completely new quality control process flow, whereby metal parts can be analyzed in real time during manufacturing,” Seita said. “This approach is perfectly aligned with the idea of digital manufacturing.” The DRM workflow only requires knowledge of the topographies induced by the chemical etching. It does not require prior knowledge about the geometry of the part to be inspected. The electrochemical etching technique used with DRM is a routine step taken to identify manufacturing defects or assess damage. It can be performed after part production and periodically to monitor the integrity of metal components throughout their life. As such, the DRM-based approach to assessing quality is considered nondestructive. Moreover, given that etching is such a common practice in the industry, it does not add any entry barrier to using DRM in quality control. The DRM technique provides fast data acquisition and an exceptional field of view, making it a promising option for inspecting entire components as they exit the manufacturing line in real time and throughout their ongoing service. “We believe that incorporating DRM into a digital manufacturing paradigm could create more responsive and resilient production systems,” Seita said. “As a tool that aligns with the vision of interconnected, intelligent factories, DRM represents a significant step forward towards achieving Industry 4.0’s promise of efficient, high-quality, and sustainable manufacturing.” The research was published in NPJ Computational Materials (www.doi.org/10.1038/s41524-024-01458-5).