Multiscale TRIP-based investigation of low-cycle fatigue of polycrystalline NiTi shape memory alloys

Highlights

• A mechanism of fatigue crack initiation during cyclic loading in SMAs is modeled and experimentally validated.

• TRIP appearing in A-M interfaces is the key factor that drives fatigue crack initiation during cyclic loading in SMAs.

• Maximum temperature during phase transformation can be used as an indicator to predict low-cycle fatigue of SMAs.

• In the pseudo-elastic domain, if the maximum temperature is kept constant the fatigue lifetime remains unchanged.

• Fatigue failure almost always occurs at one of the martensite front at the end of the forward phase transformation.

Abstract

In this paper, a multiscale investigation of fatigue crack initiation in shape memory alloys (SMAs) based on Transformation Induced Plasticity (TRIP) is presented. A mechanism for fatigue crack initiation during cyclic stress-induced phase transformation along with theoretical model is proposed. To validate the TRIP-based model, quasi-static tests at different ambient temperatures, ${40}^{\circ }\text{C}$, ${52}^{\circ }\text{C}$ and ${65}^{\circ }C$, and strain and stress controlled low-cycle fatigue tests at different frequencies ranging from 0.16 Hz to 5 Hz on pseudoelastic NiTi wires are carried out. The results show that, (i) TRIP appearing on phase transformation interfaces is the key factor that drives the fatigue crack initiation during cyclic stress-induced phase transformation in SMAs; (ii) maximum temperature during phase transformation is a relevant indicator to predict low-cycle fatigue of SMAs and, (iii) within the range of pseudoelasticity and below the plastic yield, low-cycle fatigue of SMAs is not directly correlated with the mechanical loads applied at macro-scale, in the sense that, if the maximum temperature reached during loading cycles is kept constant, the fatigue lifetime remains unchanged whatever the amplitude of the mechanical loading is. Based on the findings, a new criterion for pseudoelastic low-cycle fatigue of SMAs as well as fatigue-isolines diagram are proposed and validated experimentally.

Introduction

The pseudoelastic behavior of shape memory alloys (SMAs), i.e. the accommodation of large recoverable strains and the dissipation of a significant hysteresis energy, can be explained by the stress induced solid-solid martensitic phase transformation. This unique property of SMAs has been widely used in various applications ranging from biomedical to space industries, where fatigue could be a key issue.

It is well known that phase transformation makes the mechanism of structural fatigue in SMAs much more complex than the one in “classical” materials because the material structure and the related physical fields (such as stress, strain, temperature, etc.) are highly influenced by phase transformation. The fatigue lifetime of SMAs strongly depends on factors related to phase transformation, such as thermomechanical coupling (Eggeler et al., 2004; Matsui et al., 2004; Wagner et al., 2004; Zhang et al., 2016, 2017), grain size (Yin et al., 2016), pre-training (Gupta et al., 2015; Zhang et al., 2018) and loading types [proportional loading (Song et al., 2015b) or non proportional loading (Song et al., 2015c, 2016)]. To predict the fatigue lifetime, some macroscopic fatigue criteria have been proposed in both 1-D (Maletta et al., 2012, 2014; Song et al., 2015a) and 3-D (Moumni et al., 2005; Runciman et al., 2011; Zhang et al., 2016; Song et al., 2017). In order to take into account thermomechanical coupling, a total strain energy-based criterion was proposed and experimentally validated in (Zhang et al., 2017).

In all these works, the mechanical response at macro-scale has been used for the prediction of fatigue lifetime of SMAs without a proper explanation of related physical mechanisms, and this makes an investigation at lower scales worthwhile. In fact, it is well established that fatigue of SMAs is related to local phenomena occurring at meso-scopic (grain) scale (Shaw and Kyriakides, 1997; Hallai and Kyriakides, 2013; Reedlunn et al., 2014; Zhang and He, 2018). A series of experimental works on fatigue behavior of pseudoelastic NiTi plates have been proposed by Zheng et al. (2016a, b, 2017) and the results show that fatigue cracks initiate in the so-called active zone where SMAs undergo locally cyclic phase transformation. Many other experimental works at micro-scale or meso-scale using X-ray diffraction (XRD) (Koster et al., 2015; Sedmák et al., 2015), Scanning Electron Microscope (SEM) (Eggeler et al., 2004; Rahim et al., 2013; Koster et al., 2015) and Transmission Electron Microscope (TEM) (Frotscher et al., 2009; Pelton, 2011) have been carried out. It is believed that fatigue damage in pseudoelastic SMAs is resulting from permanent dislocation slips that occur on phase transformation interfaces (Delville et al., 2011; Kundin et al., 2015; Sedmák et al., 2015). Furthermore, dislocation slips (Delville et al., 2011; Polatidis et al., 2015), as well as residual oriented martensite locked-in by internal stress field induced by dislocations, (Hamilton et al., 2005; Yu et al., 2013, 2014; Sedmák et al., 2015; Chowdhury and Sehitoglu, 2017), induce residual deformations. This residual strain has been chosen as an indicator for fatigue damage of SMAs (Lagoudas et al., 2009; Zheng et al., 2017). During stress induced phase transformation, the macro-stress applied to trigger the phase transformation is usually lower than the plastic yield of the material. Nevertheless, even when the macroscopic stress field is below the plastic yield, dislocation slips are triggered due to the so-called Transformation Induced Plasticity (TRIP) (Kang et al., 2009; Norfleet et al., 2009; Kato and Sasaki, 2013; Wang et al., 2014; Yu et al., 2015a, b; Cisse et al., 2016; Paranjape et al., 2017; Heller et al., 2018): when phase transformation takes place, a high-level local stress field is created as a result of the unmatched deformation in phase transformation interfaces (i.e., austenite-martensite interface, hereinafter referred to as A-M interface) which assist dislocations slip in the A-M interface. Upon cycling, the accumulation of dislocations get saturated and the mechanical response of SMAs reaches a stabilized state after dozens of cycles (Moumni et al., 2005; Morin et al., 2011; Gu et al., 2017; Wang et al., 2017; Zhang et al., 2017; Xiao et al., 2018). However, although TRIP is believed to be the key factor in fatigue crack initiation during cyclic stress induced phase transformation in SMAs (Wagner et al., 2008; Lagoudas et al., 2009; Zheng et al., 2016a, 2017), the physical mechanisms related to this process is still unclear and there is no theoretical model available to present a quantitative assessment of the fatigue crack initiation in SMAs.

This paper aims at filling the aforementioned gap by giving an experimental and theoretical assessment of the link between TRIP and fatigue crack initiation in SMAs. In section 2, first the physical mechanism of fatigue crack initiation in shape memory alloys is presented, and then a theoretical TRIP-based model for crack initiation is established. Experimental validations of the model are presented in section 3. In section 4, some issues related to low-cycle fatigue of SMAs are discussed including the dependence of the fatigue lifetime on the macroscopic mechanical loads as well as the spatial location of the fatigue failure. Finally, summary and conclusions are given in section 5.