Large-Scale Serial-Sectioning Observation of 3D Steel Microstructures Based on Efficient Exploring of Etching Conditions Using 3D Internal Structure Microscope

Abstract

This paper describes large-scale three-dimensional (3D) observation of microstructures in ferrous materials, using a three-dimensional internal structure microscope that enables automated serial sectioning of ferrous materials, assisted by efficient exploring of etching conditions. Our system uses precision cutting for surface fabrication, which can create a mirror-like cross-section within a minute at depth intervals of 1 μm. Our approach consists of two steps: Exploring the etching conditions and 3D observation. For the first stage, etching times were automatically changed for each cross-section from five to 40 s in 2.5 s steps, and an optical microscope with a digital camera captured the cross-sections. In this process, the specimen did not need to be detached from the device, and it took less than an hour to obtain the images for 15 conditions. A suitable image could be quickly selected from these. The following 3D observation step demonstrated automated large-scale serial sectioning of 0.15C-1.5Mn steel using the etching condition. The 3D model offered a range of 867 × 645 × 1500 μm3, which had 1000 cross-sections at 1.5-μm intervals and a resolution of 0.066 × 0.066 × 1.5 μm3. The total voxel size became 13196 × 9824 × 1000 voxels. The observation time was 3.5 min per section and took about 60 h in total. The quality of the 3D image was sufficient for recognizing clear microstructures even in the reconstructed side surfaces.