Pearlite Size Effects on Ductility at Cryogenic Temperature via In-Situ Cantilever Loading

BCC metals generally exhibit a ductile-to-brittle transition temperature (DBTT) because their structure lacks fully dense planes, impeding dislocation cross-slip. In this study, we investigate the temperature dependance of this effect in BCC Fe that is confined between cementite (Fe3C) laths at two different average spacings using micro-mechanical testing. Conventionally, the DBTT behavior in BCC metals is measured using either a Charpy impact testing or a tensile testing. However, the size of these test specimens makes it difficult, if not impossible, to observe how the deformation behavior initiates and propagates in the microstructure. With such a temperature dependent deformation response within the metallic phase confined between hard, brittle ceramic phases, it is of interest to understand the cooperative deformation evolution as a function of temperature. This lends insights into how catastrophic failure mechanisms propagate when the metal has such a duality of responses. Recent advances in in-situ testing have enabled the indentation of materials at sub-ambient temperatures as low as -160°C, which we report here. To yield the desired ‘composite’ metallic/ceramic microstructure, we have used a 1095 steel and a gray cast iron (2.6-3.75 wt.%C), each with a pearlite microstructure but with varying changes in the lath spacing/thickness. Microstructurally, these alloys are relatively similar in that they are predominantly pearlite, but the 1095 has some cementite grains whereas the higher carbon content in the gray cast iron yields large graphitic regions. Through micro-mechanical testing, we can confine our test to only reside within regions that contain the pearlite laths. Each sample was metallographically polished to reveal the grain structure with the lath morphology. After which a specific pearlite grain was identified, a Xe Plasma Focus Ion Beam (PFIB) was used to mill cantilevers that were approximately 3x3x10 µm in dimension. A notch was then made using a Ga FIB at 30 kV, 1 pA to an approximate depth of 1 µm. From scanning electron microscopy (SEM) images, the lath spaces and number of laths in each of the cantilevers for each sample type was measured and tabulated in Table 1. With the cantilevers milled, a 1 µm flat punch loaded the end of the cantilever to a depth of 2.5 µm over 45s with a hold time of 5s in a Bruker PI-89 in-situ indenter that was inserted into a FEI Versa 3D SEM. The mechanical tests were conducted at room temperature and -160 °C. All cantilevers were subjected to two repeat loadings using the same load function. For the 1095 steel at room temperature, the in-situ images in Fig. 1 a) reveals a clear example of ductile notch widening. This ductile response resulted in the flat punch, initially centered on the cantilever end for both tests, slipping off the side of the cantilever after substantial plastic bending. This slipping corresponds to the drops and modulations observed in the load-displacement curve. When the sample was cooled to -160 °C, it surprisingly behaved relatively ductile and did not reveal a clear DBTT response. While notch widening was observed, eventually small cracks nucleated and propagated horizontally around the notch tip. This is highlighted by a white arrow in Fig. 1 b). A similar response was also observed again in the second cryogenic temperature test, where the cracking was slightly lengthened. A similar room temperature and cryogenic load was applied to the cast iron pearlite microstructure. Here, the room temperature showed the expected ductile response, Fig. 2 a). As the test proceeds, the notch widens. When the sample was cryogenically cooled, it revealed clear cracking, highlighted within the dashed circle, Fig. 2 b). However, even with this cracking, the load-displacement did not manifest itself as a clear brittle response as the load curves are absent of an abrupt load drop. Rather the load curves exhibit a slightly downward slope consistent with a quasi-brittle response. Collectively, both samples resulted in an absence of a clear DBTT response at the cryogenic condition. It is suspected that the absence of the DBTT may be a result of a size effect, which occurs when samples are reduced in volume for micro-mechanical tests. This reduced volume limits obstacles for dislocations [1], even though we have confined the Fe between the cementite phase. In the reduced spacing sample (gray cast iron, Table 1), the confinement was indeed smaller, and a quasi-brittle response was noted [2]. Microstructure quantification SEM images of cantilevers in the 1095 sample as well as the corresponding load-displacement curves. a) is a cantilever deformed at RT and b) was deformed at -160°C with the white arrow pointing to a crack formed. Two tests were performed with each final test state displayed. SEM images of cantilevers in the 1095 sample as well as the corresponding load-displacement curves. a) is a cantilever deformed at room temperature and b) is deformed at -160°C with the white circle indicating the fracture surface. Two tests were performed with each final test state displayed.