Investigation of the deformation mechanisms and defects in high‐Mn austenitic steel

Abstract

A new class of austenitic steels stabilized with high Mn contents (instead of Ni) exhibits exceptional mechanical properties, such as large energy absorption and high work‐hardening rate, owing to secondary deformation mechanisms such as mechanical twinning‐induced plasticity (TWIP) and martensitic transformation‐induced plasticity (TRIP) favored for low stacking‐fault energy (SFE) [1]. The interaction of dislocations with twin boundaries and martensite interfaces during mechanical deformation enhances the work hardening, i.e., a dynamic Hall‐Petch effect, with total elongations exceeding 70% and ultimate tensile strengths in the GPa regime. The influence of the strain rate, temperature, and changes in SFE on the deformation mechanisms in high‐Mn austenitic steels has been investigated using electron backscattered diffraction (EBSD), electron‐channeling contrast imaging (ECCI), conventional bright‐field/dark‐field imaging (BF/DF), and aberration‐corrected high‐resolution scanning/transmission electron microscopy (HRTEM/HRSTEM). The TWIP/TRIP secondary deformation mechanisms are related to the low SFE exhibited in these materials. Experimentally measured SFE from weak‐beam‐dark‐field (WBDF) imaging provides the basis to understand how changes in SFE influence mechanical twinning versus transformation induced martensite [2‐3]. However, adiabatic heating during deformation at high strain rates (20 ‐10,000 s ‐1 ) increases the SFE. Quantifying the twin or martensite density by EBSD/ECCI and BF‐DF images allows for comparison of the secondary deformation at different SFE, strain rates, and total elongation, but to study the details of the deformation mechanisms requires imaging at atomic resolution using aberration‐corrected electron microscopy. Figure 1(a) shows the EBSD/ECCI experimental procedure. EBSD identifies grains with a [110] orientation that are subsequently imaged using backscattered electrons where the incident beam strongly channels except for areas with twinning and martensite plates. Although the EBSD/ECCI method provides for a statistical number of measurements, the DF imaging method shown in figure 1 (b) has the advantages of improved resolution and the ability to differentiate between the hexagonal ε‐martensite and twins. The histograms in figures 1 (c‐d) summarize the spacing between the planar defects and their thickness from a Fe‐25Mn3Al3Si alloy deformed at a strain rate of 20 s ‐1 to a total strain of 20% using these two experimental methods. Figure 3 is a HRTEM image produced with an image corrected FEI‐Titan from a Fe‐25Mn3Al3Si alloy deformed at a strain rate of 2x10 3 s ‐1 to a total strain of 18%. The image shows an example of an individual dislocation trapped at the planar defect interface as well as a local region of martenite (bottom left) together with the mechanical twins. The high‐angle dark field (HAADF) HRSTEM image in figure 3 from a Fe‐16Mn14Cr0.3N0.3C alloy deformed to a total strain of 21% at a strain rate of 10 ‐4 s ‐1 was acquired using the ER‐C PICO operating at 300 kV. Similar to the high strain rate material, the quasi‐static deformed microstructure exhibits multiple mechanical twins with evidence of local hexagonal stacking (upper right). Advantages of the HRSTEM method compared to HRTEM images are more straight forward image interpretation from the HAADF amplitude contrast, the ability to image thicker samples, and the reduced sensitivity to local variation in crystallographic orientation.