The combined and interactive effects of orientation, strain amplitude, cycle number, stacking fault energy and hydrogen doping on microstructure evolution of polycrystalline high-manganese steels under low-cycle fatigue

Highlights

• Dislocation patterns of more than 1000 grains with defined loading conditions are observed by an integrated experimental approach combining DIC, EBSD and ECCI.

• Taylor factor is found to represent the crystallographic orientation information in polycrystalline materials more accurately than the commonly used loading direction parameter.

• The combined and interactive influences of crystallographic orientation, strain amplitude, cycle number, stacking fault energy and hydrogen doping on LCF behaviour of HMnSs are explored and discussed in detail.

Abstract

We studied the combined and interactive effects of crystallographic orientation, strain amplitude, cycle number, stacking fault energy (SFE) and hydrogen doping on the microstructure evolution of polycrystalline high-manganese steels (HMnSs) under low-cycle fatigue (LCF). An integrated experimental approach combining digital image correlation (DIC), electron backscatter diffraction (EBSD) and electron channelling contrast imaging (ECCI) at interrupted cycles was performed in the same region of interest on the bulk shear samples, which enables us to systematically compare the dislocation patterns of grains with defined loading conditions at a much larger field of view and less artefacts compared to transmission electron microscopy (TEM). We found that Taylor factor (M) works well with describing the effect of crystallographic orientation, which was further proved by the crystal plasticity finite element method (CPFEM). In detail, grains with a medium M value (3.3–3.9) tend to form more complex dislocation pattern, which is defined as sensitive value. Generally, increasing strain amplitude and cycle number both promote the evolution of dislocation pattern while their efficiencies depend strongly on grain orientation. The promotion effect of SFE on dislocation evolution becomes obvious at larger strain amplitudes (>0.4%) and non-sensitive M value. Hydrogen can strongly assist the formation of ε-martensite and reduce its critical resolved shear stress (CRSS), while it retards the evolution of dislocation pattern. The number of activated martensite variants in individual grain can be well predicted by its M value. With increasing strain amplitude, the fraction of ε-martensite increases in a manner of thinner but denser plates.

Introduction

High-manganese steels (HMnSs) show great potential in engineering applications, e.g. infrastructures and automotive industries, owing to their good combination of strength and ductility resulting from the twinning-induced plasticity (TWIP) and/or transformation-induced plasticity (TRIP) mechanisms (Bouaziz et al., 2011; Chung et al., 2011; De Cooman et al., 2018; Fischer et al., 2000). During these applications, they always undergo cyclic deformation, which makes fatigue an important failure mechanism within their service life. Therefore, comprehensive understanding the fatigue behaviours of HMnSs, especially the effects of different influencing factors on the microstructural evolution, is very important to accurately predict their fatigue life.

Series of experiments were performed to study the fatigue behaviour of different kinds of pure face-centred crystal (fcc) materials (Cu (Finney and Laird, 1975; Hancock and Grosskreutz, 1969), Ag (Sastry and Ramaswami, 1977), Al (Mitchell and Teer, 1969; Vorren and Ryum, 1987), Ni (Blochwitz and Velt, 1982), etc.). Based on these earlier work people realized that the fatigue microstructures depend on several factors, e.g. strain amplitude (Basinski et al., 1969; Hancock and Grosskreutz, 1969), orientation (Li et al., 2009, 2010) and stacking fault energy (SFE) (Li et al., 2011, 2017). Hancock et al. (Hancock and Grosskreutz, 1969) found that with increasing strain amplitude the dislocation patterns change from loop patches to cell walls in single crystalline copper. Li et al. (2010) further studied the effect of orientation and showed that the dislocation patterns vary from persistent slip band (PSB) ladders to cell-like structures when the loading orientation changes from {011} to {111}. However, most of these fatigue studies were performed on pure fcc materials with single crystalline microstructures and studied their fatigue behaviour as a function of single-parameter.

Fatigue behaviours of polycrystalline fcc alloys, e.g. HMnSs, are more complicated (Kubler et al., 2011). It was reported that the deformation mechanisms in HMnSs depend strongly on stacking fault energy (SFE) (Pierce et al., 2014). With SFE below 20 mJm⁻² the TRIP mechanism occurs, while the critical range for the occurrence of TWIP mechanism is between 20 and 40 mJm^-2. Moreover, the slip mode of materials is also greatly affected by SFE (Allain et al., 2004; Li et al., 2011), which strongly affect their fatigue behaviour (Marnier et al., 2016). Therefore, the fatigue behaviour of HMnSs strongly depends on the SFE, which was quantitatively studied by Nikulin et al. (2013). Shao et al., 2016a, 2016b, 2017 addressed the effect of SFE on fatigue microstructures in HMnS in detail. Furthermore, Niendorf et al. (Lambers et al., 2012; Niendorf et al., 2009, 2010) studied the effect of strain amplitude, pre-deformation and twinning on fatigue response of HMnSs. However, none of them took grain orientation into consideration, which was reported to be a determining factor on dislocation patterns in cyclically deformed polycrystalline nickel (Buque et al., 2001). Besides, the dislocation patterns were reported to develop significantly at the very beginning of cyclic loading (An and Zaefferer, 2019; Pham and Holdsworth, 2011; Pham et al., 2013), however, most studies focused on the microstructures after high cycle numbers or even after fracture. More attention should be paid on studying the low-cycle fatigue (LCF) behaviour of polycrystalline HMnSs. Moreover, to estimate the fatigue life of the studied materials, investigations on the cyclic deformation response and the deformation microstructure evolution with increasing cycle numbers are necessary (Anahid and Ghosh, 2013; Ghosh and Anahid, 2013).

Hydrogen is also an important factor on mechanical properties of HMnSs, as hydrogen was reported to induce severe and unpredicted damage, i.e. hydrogen embrittlement (Koyama et al., 2012; Ryu et al., 2012; So et al., 2009). However, most of these studies done so far only investigated the effect of hydrogen on their mechanical properties under monotonic deformation, our knowledge on the fatigue behaviour of HMnSs with hydrogen doping is still incomplete. In a previous study(An et al., 2020), we reported that hydrogen can assist the formation of stacking faults (SFs) and ε-martensite in a TWIP steel. In the present study, the effect of hydrogen on fatigue behaviour of HMnSs was investigated more systematically by means of considering the effects of grain orientation and strain amplitude.

The aim of this study is to comprehensively understand the influences of crystallographic orientation, strain amplitude, cycle number, SFE and hydrogen on the LCF microstructures of the polycrystalline HMnSs in conjunction with correlating their interdependence. To this end, the fatigue behaviours of two HMnSs (TRIP and TWIP) with different SFE as well as the TWIP steel with pre-charging hydrogen were investigated and compared. Digital image correlation (DIC), electron backscattering diffraction (EBSD) and electron channelling contrast imaging (ECCI) at interrupted cycles were performed to obtain the local strain amplitude, the grain orientation and the dislocation pattern evolution of the fatigued samples, respectively. The integrated experimental approach enables observing the dislocation patterns of grains with detailed loading information at a much larger field of view and less artefacts compared to transmission electron microscopy (TEM).