Effect of post-heat treatment conditions on shape memory property in 4D printed Fe–17Mn–5Si–10Cr–4Ni shape memory alloy

Abstract

The influence of post-heat treatment on the microstructure and properties of 4D printed Fe–17Mn–5Si–10Cr–4Ni (wt. %) shape memory alloy (SMA) produced via a laser powder bed fusion process is investigated in this study. It is shown that heat-treatment temperatures lower than 800 °C are not sufficient to completely undergo phase transformation from bcc-δ to fcc-γ, whereas temperatures higher than 800 °C lead to grain growth and thickening of hcp-ε phases. The latter has a severely negative effect on shape memory. Furthermore, heat treatment at 800 °C for more than 3 h leads to σ-phase formation, which also negatively affects the shape memory effect (SME) by changing the chemical composition of the fcc-γ matrix. The SME along the building direction (BD), is higher than that along the laser scanning direction (SD) and in 45° to the SD (45SD). The higher SME along the BD is attributed to the {110} texture in the plane normal to the BD, whereas almost random but weak {111} and {100} textures are observed in SD and 45SD. Furthermore, transmission electron microscopy revealed that a high stacking fault density and very thin hcp-ε phase are formed in the heat-treated sample, which explains the higher SME regardless of direction, compared with conventionally fabricated Fe-based SMAs with a similar composition.

Introduction

Fe-Based shape memory alloys (Fe-SMAs), which have a representative composition of Fe–Mn–Si, have received significant attention because of their lower production cost than Ni–Ti alloys, machinability, and moderate shape memory effect (SME), that is, deformed specimens recover their original shapes when heated. Following the first observation of SME in single-crystalline Fe–30Mn–6Si by Sato et al., in 1982 [1], polycrystalline Fe-SMAs have been widely studied to enhance the SME because it is less in polycrystalline alloys than in single crystals. For the first generation, cyclic thermo-mechanical treatment, that is, repetitive deformation with heat treatment, can effectively enhance the SME by creating a high density of stacking faults in face-centered cubic (fcc)-γ grains [[2], [3], [4]]. However, owing to the additional procedure, the cost of production is higher. In this regard, Kajiwara et al. reported a new training-free Fe-SMA containing Fe–28Mn–5Si-xCr-yNi-(Nb, C) through precipitation of NbC [5]. The precipitates generated more stacking faults in their vicinity to relieve the misfit strain between the matrix and the precipitate. The resulting stacking faults act as easy nucleation sites for the fcc-γ to hexagonal close-packed (hcp)-ε transformation. Dong et al. discovered that VC precipitation improved the SME in the Fe–17Mn–5Si–10Cr–4Ni-(V, C) alloy [6]. This alloy exhibits a high recovery stress and SME. The basic approach for improving the SME is to increase the stacking fault density. However, there are other important factors for enhancing SME, such as grain size, morphology, and texture [7,8]. Arabi et al. reported the effects of the grain size and texture [9]. However, it is challenging to control the texture in near-net-shaped products.

Beside the conventional fabrication method, such as casting and rolling, powder metallurgy processes have been subjected to fabricate SMAs [10,11]. Among them, SMA fabrication by additive manufacturing, known as three-dimensional (3D) printing, has recently been attempted. SMA printing is typically referred to as 4D printing, implying that a time-dependent shape-changing functionality is added to the 3D printed components. In this regard, Gustmann et al. have tried to find optimal processing parameters for 4D printing of Cu–Al–Ni–Mn shape-memory alloy by selective laser melting [12]. Yang et al. investigated the 4D printing of Ni–Ti-based SMAs, which is the widest used in the field of SMAs [13]. However, the exceptionally high compositional sensitivity of Ni–Ti-based SMAs makes their application challenging. Fe-SMAs, however, are considered more suitable for 4D printing because of their lower compositional sensitivity than Ni–Ti-based SMAs. Few researchers reported the shape memory propert Ferretto et al., who are a part of authors of this study, successfully fabricated a Fe–17Mn–5Si–10Cr–4Ni (wt. %) SMA via a laser powder bed fusion (L-PBF) process, which showed higher SME than conventionally fabricated alloys with a similar chemical composition, that is, hot rolled Fe–17Mn–5Si–10Cr–4Ni-1(V, C) (wt. %) [14]. They further reported the basic properties of 4D printed SMAs in terms of processing parameters and textural components. It was found that by increasing the volumetric energy density (VED), the dominant phase was changed from body-centered cubic (bcc-δ) to fcc-γ phase. Heat treatment of this alloy at 800 °C transformed the residual bcc-δ to fcc-γ, which considerably improved the SME. Furthermore, different textures were developed in different sample directions, such as the building direction (BD) and scanning direction (SD), resulting in anisotropy in the SME and tensile properties. When the sample was deformed parallel to the BD, where the {110} texture was developed, the SME higher than in the SD was observed. Later, the mechanism of solid stated phase transformation from bcc-δ to fcc-γ and their crystallographic orientation relationship was investigated [15]. It was shown that intrinsic heat treatment (IHT) effect during the laser processing could ascribe to increase the fraction of fcc-γ as increase laser power. A coherent crystallographic orientation relationship between bcc-δ and fcc-γ was confirmed by transmission electron microscope analysis.

The crystallographic texture can significantly affect the SME of the Fe-SMA owing to the differences in the Schmid factors for the movement of partial dislocations. According to literature, a higher Schmid factor against the deformation direction suggests easier movement of Shockley partial dislocations, resulting in easier phase transformation from fcc-γ to hcp-ε during deformation and higher SME [16,17]. In this regard, this study aimed to determine how post-heat treatment affects the shape memory property to determine the optimal combination of heat-treatment temperature and time. Microstructural analysis using scanning electron microscopy and electron backscatter diffraction revealed how SME is associated with the microstructure and texture resulting from each heat-treatment condition. In addition, the primary reason for the differences in the texture along each sample direction was determined. Finally, the origin of the higher SME in the printed Fe-SMAs was investigated using transmission electron microscopy.