Elsevier

Materials Characterization

Volume 167, September 2020, 110470
Materials Characterization

TEM analysis of deformation bands created by tensile deformation of superelastic NiTi wires

https://doi.org/10.1016/j.matchar.2020.110470Get rights and content

Highlights

  • Deformation bands as permanent lattice defects left in the austenitic microstructure of deformed superelastic NiTi wires.

  • A TEM method was introduced for analysis of microstructures with deformation bands using SAED combined with DF imaging.

  • Deformation bands contain {114} austenite twins and/or B19’ martensite variants with high density of slip dislocations.

  • Orientation mapping revealed the mechanism of refinement of austenitic microstructure by deformation twinning in martensite.

Abstract

Deformation processes derived from martensitic transformation in shape memory alloys are theoretically fully recoverable in a complete thermomechanical loading cycle across transformation range and do not leave any lattice defects in the microstructure. In reality, this is rarely the case in NiTi, since plastic deformation tends to accompany the martensitic transformation, particularly if it proceeds under large stress. Lattice defects observed in the microstructure of deformed NiTi wires (slip dislocations and deformation bands) attract the attention of researchers, since they are linked to unrecovered strains and play significant role in functional fatigue, shape setting or two-way shape memory effect. In this work, we present an experimental approach allowing for analysis of deformation bands in deformed NiTi consisting in: i) preparation of superelastic NiTi wires with recrystallized, small grained microstructure, ii) subjecting these wires to desired tensile test (e.g. superelastic or shape memory cycle) and iii) characterizing the deformation bands in TEM using selected area electron diffraction combined with dark field imaging following simple rules described in this work. If the deformed microstructure becomes too complex due to the high density and small size of deformation bands, ASTAR orientation mapping can be beneficially applied to reveal the refinement of the microstructure through the introduction of deformation bands.

Introduction

Superelastic NiTi wires display large recoverable tensile strains (~4–8%), the magnitude of which depends on the wire texture, virgin microstructure, and test temperature [1,2]. Since these large strains are due to fully recoverable deformation processes of elastic deformation and stress induced martensitic transformation [2], the original parent austenite is, in the ideal case, completely restored in a closed loop thermomechanical load cycle through the transformation range without leaving any lattice defects in the microstructure. In reality, however, unrecoverable strains [3] were frequently recorded and permanent lattice defects, which do not disappear from the microstructure upon unloading, were frequently observed in the microstructure of superelastically cycled NiTi wires by many investigators in the literature. The observed permanent lattice defects are mainly slip dislocations [[3], [4], [5], [6]] and deformation bands [1,[7], [8], [9], [10], [11], [12], [13]].

When superelastic NiTi wires are deformed in the low temperatures martensite state up to the maximum tensile strain within the plateau range [1,3,12] and heated above the Af temperature, the applied strains are expected to be ideally completely recovered. This phenomenon is known as the shape memory effect. In reality, the situation is very close to the ideal case (the strain recovery is almost complete and practically no lattice defects are left in the microstructure of the wire subjected to complete shape memory thermomechanical cycle [3,11]). In fact, complete strain recovery was claimed to occur in the shape memory cycle since the martensitic transformation never proceeds under external stress in it [3]. This sharply contrasts with the superelastic deformation, which inherently involves martensitic transformation proceeding under external stress. When the martensitic transformation proceeded in superelastic tensile test at high temperature and high stress [1], deformation bands besides slip dislocations were frequently observed in the microstructure of the unloaded wire. Moreover, if NiTi wire is strained in tension beyond limits posed by the crystallography of the B2-B19’ transformation in NiTi (~12% strain) and heated above the Af temperature [11], the strains are not recovered completely and deformation bands containing austenite twins appear to dominate the microstructure of deformed and heated NiTi wires [11,12]. It is important to realize that one needs to perform a complete thermomechanical load cycle through the transformation range in order to determine unrecovered strains and reveal permanent lattice defects created by the mechanical load [1,3,11,12].

Since the lattice defects created by tensile deformation in NiTi (dislocations and deformation bands) are linked to the irreversible processes that accompany the stress induced martensitic transformation and cause unrecovered strain (functional fatigue and preliminary fatigue failure), they have been frequently investigated in the field in the past decades. Unfortunately, these lattice defects are very difficult to detect and analyze in TEM in case of commercial NiTi wires [3] exhibiting nanograined, partially recrystallized microstructures (several small grains overlap within the ~100 nm thick TEM foils which, moreover, contain lattice defects and internal stress from the previous cold work). On the other hand, the results of dedicated experiments on solution annealed NiTi [5,6,[8], [9], [10]], which frequently show a high density of slip dislocations created by the martensitic transformation, are not representative for high quality commercial nitinol. This is the case because the stress induced martensite in the solution annealed NiTi alloys tends to deform plastically at a level of stress similar to that at which the austenite transforms at room temperature [13]. A compromised grain size (100 < d < 500 nm), assuring that the alloy is superelastic but its microstructure does not contain lattice defects persisting from the cold work and that its grains are small enough but do not overlap in the TEM foil, is hence needed for meaningful TEM analysis of permanent lattice defects in deformed NiTi [1,7,11,12]. Besides the TEM analysis of lattice defects in deformed NiTi, it has recently become possible to analyze the martensitic transformation generating unrecoverable strains theoretically by molecular dynamics modeling [14] and phase field simulations [15] methods.

In this work, we put aside slip dislocations (see [[3], [4], [5], [6]] for information on slip dislocations generated by the forward and reverse martensitic transformations in superelastic NiTi) and pay special attention to the deformation bands. The deformation bands are permanent lattice defects observed mainly in the microstructure of NiTi wires deformed at high temperatures [1,3] and/or in heavily deformed NiTi (see [11] and references therein). Besides of that, however, deformation bands tend to appear in the solution treated NiTi deformed within the plateau range at room temperature [1,5,6,[8], [9], [10]]. This sharply contrasts with commercial nanograined NiTi wires, where deformation bands appear only in tensile tests at very high temperatures [1,3]. This is due to the strong grain size dependence of the yield stress for the plastic deformation of martensite [13]. Deformation bands were also observed to appear in the microstructure of any NiTi wire after a given number of superelastic cycles characteristic for its virgin microstructure and test temperature [12,13]. It shall be mentioned that deformation bands were frequently confused in the literature with conventional residual martensite plates (disappear from the microstructure upon heating above Af without leaving lattice defects in it) [9,10]. In summary, the occurrence of deformation bands in the microstructure of NiTi wires deformed in tension depends on the virgin microstructure of the wire [1] and increases with increasing test temperature [1], maximum strain [11] and number of superelastic cycles [12].

Deformation bands may contain B2 austenite, R-phase and/or B19’ martensite phases depending on the deformation history [1,7,8,[11], [12], [13]]. There is extreme strain inhomogeneity in the microstructure of deformed NiTi associated with the formation of deformation bands (plastic strains as large as 30% were determined within the wide deformation bands of solution treated NiTi by Digital Image Correlation (DIC) method [8]). Significant macroscopic unrecoverable strains associated with the occurrence of deformation bands in the microstructure and refinement of the parent austenite microstructure via introduction of deformation bands were reported in [1,11,12]. We have recently proposed that deformation bands are created by further distortion of the B19’ martensite via dislocation slip accommodated deformation twinning taking place at high stresses [1,7,11]. Upon unloading and heating above Af, the B19’ martensite within the microstructure of the deformed wire retransforms back to the parent austenite everywhere except the deformation bands, where it retransforms to austenite twins yielding permanent unrecoverable strains (see also Gao [16,17] who investigated that mechanism theoretically). The permanent lattice defects, observed in the austenite state after complete thermomechanical cycle across the transformation range, are thus created by the irreversible deformation processes activated in the martensite state. To fully characterize the observed deformation bands, we need to specify the phases, lattice misorientations and interface planes, as presented and discussed in this work.

Section snippets

Materials and methods

In order to analyze the deformation bands created by superelastic deformation, it is necessary to perform tensile tests on fully recrystallized but still superelastic NiTi wires. To prepare superelastic NiTi wires with such microstructure, we heat treated NiTi wires (Fort Wayne Metals #1) in cold worked state by the electropulse method [13] using a 16 ms pulse rendering it fully recrystallized microstructure with defect free grains having diameter d ~500 nm (Fig. 1a).

This wire displays almost

Lattice defects created by superelastic deformation – analysis by selected area electron diffraction combined with dark field imaging in TEM

A grain containing few deformation bands (Figs. 1d, 2a) in the microstructure of the deformed wire was selected for demonstration of the TEM analysis. The lamella was tilted so that the austenite matrix in the grain was oriented in the [011]B2 zone. Bright field image of the whole grain including the deformation bands in this particular foil orientation (Fig. 2b) becomes completely dark since all crystal orientations diffract strongly. The same grain becomes bright in the DF image taken with [01

Conclusions

An experimental approach for analysis of deformation bands created within the grain microstructure of superelastic NiTi wires upon tensile testing was presented. Deformation bands are permanent lattice defects containing {114} austenite twins, R-phase, and/or B19’ martensite variants with a high density of slip dislocations observed within the microstructure of deformed NiTi which do not disappear upon unloading and heating above the Af temperature.

The approach consists in: i) preparation of

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

Support of this research from Czech Science Foundation projects 18-03834S (O. Molnárová), 20-14114S (P. Šittner) is acknowledged. MEYS of the Czech Republic is acknowledged for the support of infrastructure projects LNSM (LM2018110), SOLID 21 (CZ.02.1.01/0.0/0.0/16_019/0000760) and ESS-CZ–OP (CZ.02.1.01/0.0/0.0/16_013/0001794).

References (25)

Cited by (23)

View all citing articles on Scopus
View full text