PSI - Issue 68

Ritsuki Morohoshi et al. / Procedia Structural Integrity 68 (2025) 701–707 R. Morohoshi et al. / Structural Integrity Procedia 00 (2024) 000–000

704

4

(c) Phase map. Blue: , Red: ′

(a) SEMx50 Fig. 4: Basic results from uniaxial test. tensile direction: ⇔ Now, to briefly recap the objectives of this study. The goal of this experiment is to quantitatively evaluate the effect of different stress triaxialities ( ) on the ′ transformation by using specimens that represent these three values of . In previous studies, XRD (Lebedev and Kosarchuk (2000)) and ferrite scopes (Beese and Mohr (2011)) were employed to observe the transformation amount for similar evaluations. However, this study utilizes SEM-EBSD for this purpose. Generally, SEM-EBSD is not commonly used for such evaluations because techniques like XRD or ferrite scopes can provide an averaged result over the irradiated area by calculating the ′ transformation rate from hit rates or magnetic measurements, thus ensuring that the results reflect more global properties. On the other hand, SEM-EBSD generates diffraction pattern images that directly correspond to specific crystal phases and orientations. This method cannot provide coarse, averaged information because the spatial resolution at which the diffraction patterns are measured should be smaller than the grain size. Consequently, a quantitative evaluation of the ′ transformation rate can only be achieved through in-situ testing, specifically by tracking the transformation at the grain level. Additionally, this study leverages the high resolution of SEM-EBSD to explore the stress triaxiality dependence of the ′ transformation, an aspect that previous experimental methods have struggled to capture. To specifically evaluate the stress triaxiality dependence of the ′ transformation, it is necessary to stitch together the continuously captured EBSD images taken during testing. Therefore, an automated solution through programming was adopted to tackle this issue. For simplicity, this challenge will be referred to as the ”grain tracking problem” throughout the discussion. The grain tracking problem, in essence, is an attempt to quantify the displacement and strain between consecutive EBSD images. For instance, referring to fig. 5, the following can be qualitatively observed. The red lines appear to be stretched more horizontally, and the correspondence between the grains at 105N and the crystal grains at 115N can be visually identified. However, it is practically impossible to manually label all the corresponding crystal grains across the images. The following explains the solution to the grain tracking problem. Broadly speaking, the triple junctions of grain boundaries are used as feature points for matching within the image, serving as a starting point for matching the entire image and tracking individual grains. A schematic of this process is shown in fig. 6. At the first step, temporary crystal tracking graph, search , is constructed to detect and track the grains of interest. The basic properties of this graph are outlined as follows: (1) It is a directed graph. (2) The nodes represent individual grains in the EBSD image captured under a specific load. (3) The direction of the edges corresponds to increasing load during the measurement. (4) The weight of the edges is determined by the crystallographic orientation deviation. The orientation deviation between and ′ phases is calculated considering the Kurdjumov-Sachs (KS) relationship. (5) The steps for generating the edges are as follows: (a) First, a broad search for matching candidates is performed. Matching candidates for a given grain at a certain load, (hereafter referred to as the reference grain), are searched for in the EBSD image taken at load . The candidate grains are those whose coordinates fall within the set of coordinates occupied by all grains adjacent to the reference grain in . Additionally, the candidate set includes all grains adjacent to the candidate grains (hereafter referred to as the candidate grain set). Conceptually, the search range extends up to a distance approximately equal to twice the average grain diameter. (b) To narrow down the most probable candidates from the candidate grain set, only those grains with a crystallographic orientation (b) SEM x500 (d) Euler map