Journal of Materials Science & Technology  2020 , 41 (0): 76-80 https://doi.org/10.1016/j.jmst.2019.09.022

Letter

Effects of temperature and alloying content on the phase transformation and {10$\bar{1}$1} twinning in Zr during rolling

Xinglong Ana, Hao Zhanga, Song Nia*, Xiaoqin Oua, Xiaozhou Liaob, Min Songa*

aState Key Laboratory of Powder Metallurgy, Central South University, Changsha 410083, China
bSchool of Aerospace, Mechanical and Mechatronic Engineering, The University of Sydney, Sydney, NSW 2006, Australia

Corresponding authors:   *Corresponding authors. E-mail addresses: song.ni@csu.edu.cn (S. Ni), msong@csu.edu.cn (M. Song).*Corresponding authors. E-mail addresses: song.ni@csu.edu.cn (S. Ni), msong@csu.edu.cn (M. Song).

Received: 2019-07-14

Revised:  2019-09-1

Accepted:  2019-09-4

Online:  2020-03-15

Copyright:  2020 Editorial board of Journal of Materials Science & Technology Copyright reserved, Editorial board of Journal of Materials Science & Technology

More

Abstract

The effects of temperature and Ti content on the deformation mechanisms of pure Zr and Zr-Ti alloys were investigated by transmission electron microscopy. The results indicate the existence of a relation between deformation-induced phase transformation from a hexagonal close-packed structure to a face-centered cubic structure and {10$\bar{1}$1} deformation twinning. That is, when one is suppressed, the other will be promoted. The phase transformation was suppressed while the {10$\bar{1}$1} compressive twinning was promoted with increasing the rolling temperature and/or Ti content. This can be attributed to the activation of basal <a> dislocations at high temperature and the increased stacking fault energy with Ti content.

Keywords: Zr ; Zr-Ti alloy ; Twinning ; Phase transformation ; Stacking fault energy

0

PDF (2539KB) Metadata Metrics Related articles

Cite this article Export EndNote Ris Bibtex

Xinglong An, Hao Zhang, Song Ni, Xiaoqin Ou, Xiaozhou Liao, Min Song. Effects of temperature and alloying content on the phase transformation and {10$\bar{1}$1} twinning in Zr during rolling[J]. Journal of Materials Science & Technology, 2020, 41(0): 76-80 https://doi.org/10.1016/j.jmst.2019.09.022

Deformation mechanisms of hexagonal-closed packed (HCP) materials have attracted much attention due to their high structural anisotropy and limited slip systems [[1], [2], [3], [4], [5], [6], [7]]. Slip and twinning are the most common deformation mechanisms in HCP materials, while martensitic phase transformation and shear band formation also play important roles in coordinating the strain [[3], [4], [5], [6], [7]]. Factors including strain rate, temperature, grain size, texture, axis ratio (a/c) and alloy composition affect significantly the deformation mechanisms of HCP materials [[8], [9], [10], [11], [12], [13]].

Zr and Zr alloys with HCP structures at ambient temperature show great applications in nuclear industry, chemical industry, and other fields due to their appropriate densities, high melting point, low thermal neutron absorption cross-section and excellent corrosion resistance [14,15]. There are three types of slip systems: basal <a>, prismatic <a> and pyramidal <c+a>, and four types of twinning modes: {10$\bar{1}$2}<10$\bar{1}$1>, {11$\bar{2}$1}<11$\bar{2}$$\bar{6}$>, {11$\bar{2}$2}<11$\bar{2}$$\bar{3}$> and {10$\bar{1}$1}<10$\bar{1}$2> to accommodate the plastic deformation in Zr [[11], [12], [13],16,17]. Prismatic <a> is the easiest slip mode in many conditions, while basal <a> and pyramidal <c+a> are usually activated at elevated temperatures [13]. Under tensile stress along the c-axis, {10$\bar{1}$2}<10$\bar{1}$1> is the most common twinning mode in Zr, while some {11$\bar{2}$1}<11$\bar{2}$$\bar{6}$> twins can also form [13]. Under compressive stress along the c-axis, the most favorable twinning mode is {11$\bar{2}$2}<11$\bar{2}$$\bar{3}$> at low temperatures and {10$\bar{1}$1}<10$\bar{1}$2> at elevated temperatures [13,16,[18], [19], [20]]. Yang et al. [21] reported that the main deformation mechanism in cold-rolled Zr at room temperature is prismatic <a> slip, and no deformation twin was observed. Chai et al. [22] and Jian et al. [23] observed {10$\bar{1}$2}<10$\bar{1}$1>, {11$\bar{12}$1}<11$\bar{2}$$\bar{6}$> twinning and {11$\bar{2}$2}<11$\bar{2}$$\bar{3}$> twinning in cold-rolled Zr at liquid nitrogen temperature, suggesting that twinning was activated prior to dislocation slip at low temperature and played a major role in accommodating the strain at the initial deformation stages. Knezevic et al. [8] investigated the effect of temperature on the deformation mechanisms of Zr, proposing that with increasing the temperature, basal <a> and {10$\bar{1}$1}<10$\bar{1}$1> twinning gradually replaced the {11$\bar{2}$2}<11$\bar{2}$$\bar{3}$> twinning to coordinate the strain. Recently, twinning induced by intergranular thermal residual stresses has also been reported in Zr alloys during cooling process [24,25].

In addition to dislocation slip and twinning, phase transformation can also coordinate deformation in Zr [26,27]. For example, Zhao et al. [26] and Hu et al. [27] reported two types of HCP to face-centered cubic (FCC) phase transformation in cold-rolled Zr, leading to two types of orientation relationships between the two phases: (1) <1$\bar{2}$10>HCP/<$\bar{1}$10>FCC and (0001)HCP//{111}FCC, and (2) [0001]HCP//<001>FCC and {10$\bar{1}$0}HCP//{220}FCC. The first type formed via Shockley partial dislocations gliding on every other (0001) plane and the second type was accomplished by pure-shuffle and shear-shuffle mechanisms [28]. So far, there have been only a few studies focusing on the HCP to FCC transformation in Zr and Zr alloys and none of these studies discussed factors affecting the phase transformation. In this study, the effects of temperature and Ti content on the deformation mechanisms of Zr and Zr-Ti alloys, especially the HCP to FCC phase transformation, were studied. The results indicate that there was a competition between HCP to FCC phase transformation and {10$\bar{1}$1} twinning during plastic deformation. Both temperature and Ti content determine the deformation mechanisms.

The materials used in this study were high purity Zr (>99.9%) and Zr100-xTix (x = 10, 30, 50 at.%) alloys. The raw materials with equiaxed grains had average grain sizes of ∼ 50 μm. Small bars with a dimension of 50 mm × 20 mm × 5 mm were cut from the raw materials and annealed at 650 °C for 2 h to eliminate the deformation history. Then all the bars were cold rolled multiple times (0.2 mm per pass) until a total thickness reduction of 60% at room temperature except for the Zr50Ti50 bar, which was already fractured when the thickness reduction reached 30%. The pure Zr was also rolled at 650 °C with a total thickness reduction of 60%. TEM specimens were cut from the normal plane of the rolled samples and mechanically grounded to 40 μm in thickness, and then electro-polished. TEM and high-resolution TEM (HRTEM) observations were performed using a Titan G2 60-300 Cs-corrected TEM. The generalized stacking fault energies (GSFE) of Zr-Ti alloys were calculated by the density function theory (DFT) method using the Vienna Ab initio Simulation Package (VASP) code [29]. Molecular dynamics (MD) simulations of uniaxial tensile deformation on pure Zr was performed using the LAMMPS code [30]. The angular dependent potential (ADP) developed by Smirnova and Starikov was adopted [31]. Details for DFT calculations and MD simulations are shown in the Supplementary Materials.

Fig. 1(a) shows a typical bright-field TEM image and a corresponding selected area electron diffraction (SAED) pattern of the pure Zr rolled at room temperature. Lots of lamellas (marked by red arrows) with tens of nanometers in width and hundreds of nanometers in length were observed in the HCP matrix. Fig. 1(b) shows an HRTEM image of an area containing both the matrix and a lamella. Both the SAED pattern and the HRTEM image present the orientation relationship between the lamella and the matrix. From Fig. 1(b), the lamella is an FCC phase viewed from a <$\bar{1}$10> zone axis and the orientation relationship between the HCP matrix and FCC lamella is <1$\bar{2}$10>HCP//<$\bar{1}$10>FCC and (0001)HCP//{111}FCC. The longitudinal boundary between the two phases is parallel to the {0001} basal plane of the HCP matrix. The enlarged Fourier-filtered TEM image of the HCP-FCC interface as inserted in Fig. 1(b) shows two 90° and one 30° partial dislocations gliding on three (0001) planes. These partials were identified by drawing three Burgers circuits in the HCP-FCC interface based on the HCP lattice (detailed explanation on how to identify the type of partial dislocations was provided in Ref. [32]). Gliding of these partials changed the atomic stacking sequences from …ABABAB… (HCP) to …ABCABC… (FCC). Our previous investigations [26,32] indicated that this type of phase transformation was triggered by Shockley partial dislocations gliding on every other (0001) planes in the HCP matrix. These partial dislocations stemmed from the dissociation of <a> dislocations [26]. Fig. 1(c) and (d) shows two bright-field TEM images of the hot-rolled pure Zr. A large lamella with ∼ 100 nm in width and ∼ 2 μm in length was found and marked with the red arrow in Fig. 1(c). The inset corresponding SAED pattern indicates that this lamella is also the FCC phase. Compared to the cold-rolled sample, the amount of the FCC phase in the hot-rolled sample decreased drastically but the size became larger. Fig. 1(d) shows a group of twins in the HCP matrix of the hot-rolled pure Zr, which were never found in the cold-rolled sample. The SAED pattern in Fig. 1(d) indicates that these twins are {10$\bar{1}$1}<10$\bar{1}$2> compressive twins.

Fig. 1.   (a) A bright-field TEM image and a corresponding SAED pattern of the cold-rolled Zr with a thickness reduction of 60%, (b) an HRTEM image of an FCC lamella with an enlarged Fourier-filter image of the phase interface inset in (b), and (c, d) TEM images and corresponding SAED patterns of the hot-rolled Zr with a thickness reduction of 60%.

Fig. 2 presents typical TEM and HRTEM images of the cold-rolled Zr-Ti alloys with different Ti contents. Fig. 2(a) and (b) shows the bright-field TEM and HRTEM images, respectively, of the cold-rolled Zr90Ti10 alloy. Fig. 2(a) presents two FCC lamellas with similar sizes to and the same orientation relationship as those observed in the cold-rolled pure Zr (as shown in Fig. 1(a)). In addition, some FCC lamellas with much smaller size (a few nanometers in width and several to tens of nanometers in length) were also observed in the Zr90Ti10 alloy, as shown in Fig. 2(b). Two enlarged Fourier-filtered TEM images of the interfaces at two sides of the lamella on the same atomic planes were inset in Fig. 2(b). The yellow dashed lines indicate the same atomic layer. A 30° and a 90 °Shockley partials were observed at the left side of the lamella, while two 30 °Shockley partial dislocations were observed at the right side of the lamella. It has been reported that a 60° mixed-type <a> dislocation can dissociate into a 30° mixed-type partial and a 90° pure edge partial, while a screw <a> dislocation can dissociate into two 30° mixed-type partials [33,34]. Therefore the Shockley partials located at the interface regions were derived from the dissociation of one screw <a> dislocation and one 60° mixed-type <a> dislocation. Extensive TEM observations indicated that both the amount and size of the FCC lamellas observed in the cold-rolled Zr90Ti10 alloy decreased obviously compared to that observed in the cold-rolled pure Zr. No obvious twin was found in the cold-rolled Zr90Ti10 alloy. Fig. 2(c) and (d) shows the TEM images of the cold-rolled Zr70Ti30 alloy. A high density of nano-twins was observed in Fig. 2(c). The SAED pattern in Fig. 2(c) indicates that these twins were {10$\bar{1}$1} compressive twins. Fig. 2(d) shows an HRTEM image of a twin in this sample. The coherent twin boundaries (CTBs) were marked by yellow dashed lines which are {10$\bar{1}$1} planes for both the matrix and the twin, while the incoherent twin boundaries (ITBs) were marked by red dashed lines. These ITBs can also be described as basal-pyramidal (B-Py) or pyramidal-basal (Py-B) planes [35] since the orientation relationship between the twin and the matrix is (0001)M // {10$\bar{1}$$\bar{1}$} T or {10$\bar{1}$$\bar{1}$} M//(0001)T. No obvious FCC lamella was found in this sample. Fig. 2(e) and (f) shows two bright-field TEM images of the cold-rolled Zr50Ti50 with a thickness reduction of 30%. As shown in Fig. 2(e), there were lots of band structures in most areas. No obvious cell structure was found, suggesting poor plasticity of the sample. In Fig. 2(f), a group of {10$\bar{1}$1} compressive twins was observed. No obvious FCC lamella was found in this sample.

Fig. 2.   (a) A bright-field TEM image and a corresponding SAED pattern of the cold-rolled Zr90Ti10 with a thickness reduction of 60%, (b) an HRTEM image of a nanosized FCC lamella and two enlarged Fourier-filter images on both sides of the phase interface in Zr90Ti10, (c) a bright-field TEM image and a corresponding SAED pattern of the cold-rolled Zr70Ti30 with a thickness reduction of 60%, (d) an HRTEM image of a twin in cold-rolled Zr70Ti30 alloy, and (e, f) bright-field TEM images and corresponding SAED patterns of the cold-rolled Zr50Ti50 alloy with a thickness reduction of 30%.

Comparison of the microstructures of pure Zr rolled at room and high temperatures indicates that the HCP to FCC phase transformation was suppressed while the {10$\bar{1}$1} compressive twinning was promoted with increasing the temperature. This phenomenon can be explained by the activation of basal <a> full dislocations at high temperature. At room temperature, basal <a> dislocations can hardly be activated and prismatic <a> dislocation slip is the dominant deformation mode in pure Zr [8,13]. To coordinate deformation, <a> dislocations can dissociate into Shockley partial dislocations on basal planes, and the gliding of these partials change the atomic stacking sequence, resulting in HCP to FCC transformation. Thus FCC lamellas were easily observed in Zr rolled at room temperature. However, at high temperature, basal <a> dislocations can be activated much easier. Knezevic et al. [8] calculated the initial slip resistance in pure Zr, pointing out that the temperature has a significant influence on decreasing the critical resolved shear stress (CRSS) of basal <a>. Therefore, by increasing the temperature to a certain level, basal <a> dislocations can be activated. The activation and piling up of basal <a> dislocations result in the non-planar dissociation of the leading basal <a> dislocation that produces {10$\bar{1}$1}<10$\bar{1}$2> twinning dislocations [36]. Thus {10$\bar{1}$1} twins were easily observed in hot-rolled pure Zr. Knezevic et al. [8] regarded basal <a> dislocation slip and {10$\bar{1}$1} twinning as a slip-twin combination since both of them are always activated under the same condition, which is at low temperature in Mg and at high temperature in Zr. Therefore, basal <a> dislocations would not dissociate into phase transformation-related partials at high temperature, and thereby no FCC lamella can be observed. In order to further prove this, MD simulations were carried out on Zr deformed at 300 K and 923 K. Fig. 3(a) shows that a basal <a> dislocation dissociates into two Shockley partial dislocations under strain. The reaction can be written as follows:

$\frac{1}{3}[11\bar{2}0]→\frac{1}{3}[10\bar{1}0]+\frac{1}{3}[01\bar{1}0]$ (1)

Fig. 3.   (a) A basal <a> dislocation dissociated into two Shockley partial dislocations under strain and (b) the density of Shockley partial dislocations as a function of the engineering strain at 300 K and 923 K.

Such dissociations were frequently seen in cold-rolled Zr, but were rarely observed in hot-rolled Zr. Fig. 3(b) shows the density of Shockley partial dislocations with increasing the engineering strain at 300 K and 923 K, summarized from MD simulations. The results indicate that although partial dislocations were activated at lower engineering strain at 923 K compared to that at 300 K, the density of partial dislocations at 300 K is higher than that at 923 K after the engineering strain reached 12%, which agrees well with our TEM observations that the FCC phase was less with increasing the temperature.

According to Fig. 2, increasing the Ti content in Zr-Ti alloys decreased obviously the amount of the FCC phase but increased the amount of {10$\bar{1}$1} twins. In order to figure out the reasons, the temperature increment during deformation and the variation trend of SFE with the addition of Ti in the alloys were calculated. The radius of Zr (160 pm) is ∼10% larger than that of Ti (145 pm), leading to lattice distortion in Zr-Ti alloys. The solute Ti atoms would show a pinning effect on dislocation slip, resulting in a reduction in the plasticity and an increase in the strength. A simple approximate formula to calculate the temperature increase ΔT during cold rolling is as follows [11]:

fε=ρCpΔT (2)

Where, β = 0.9 is the thermal conversion factor, σf is the plane flow stress, σf=$\frac{2}{\sqrt{3}}σ$ [37], σ is the tensile strength, ε = 0.9 is the true strain in matrix when the thickness reduction reaches 60%, ρ = 5.97 g/cm3 is the density and Cp = 279 J/(kg K) is the specific heat capacity [20]. It can be seen that the temperature increment during deformation is related to the tensile stress. With increasing the Ti alloying content, the tensile stress increased from 350 MPa [21] for pure Zr to 1135 MPa [18] for the Zr50Ti50 alloy, which means when both materials reach the same deformation strain (60% in this case), the instantaneous temperature in Zr50Ti50 alloy (937.4 K) is much higher than that in pure Zr (496.5 K). Furthermore, notice that twins were mostly observed in microbands where the true strain was much higher than that in the matrix, and thus the temperature would be even higher in the microbands [37]. Therefore, with the temperature increase during cold rolling, the basal <a> dislocations can be activated in some areas especially inside the microbands, resulting in the formation of {10$\bar{1}$1} twinning in Zr-Ti alloys. In addition, the SFE of an alloy might also affect the deformation mechanism. The SFEs on basal planes of Zr-Ti alloys with different Ti contents (the content of Ti varied from 0, 2.08 at.%, 4.16 at.% to 6.25 at.%) were calculated by DFT method, as shown in Fig. 4. It can be seen that with increasing the Ti content, the SFE increases from 234 mJ/m2 to 272 mJ/m2, which means partial dislocations are more closely packed and have the tendency to gather in full dislocations for the alloys with higher Ti content. Therefore, phase transformation related partials were much harder to form for an alloy with higher SFE. Thus, with increasing the Ti content, the HCP to FCC phase transformation was suppressed while the {10$\bar{1}$1} twinning was activated due to the increases in both the SFE and temperature. It has also been reported that for Zr-Ti alloys rolled at high temperatures, {10$\bar{1}$1} twinning took place to coordinate the strain [18]. Both temperature increase and alloying contribute to activate the {10$\bar{1}$1} twinning.

Fig. 4.   Generalized stacking fault energy of Zr-Ti alloys with different Ti contents.

In conclusion, the effects of temperature and Ti content on the deformation mechanisms of pure Zr and Zr-Ti alloys were investigated by TEM. Comparison of the microstructures of pure Zr under room temperature and high temperature rolling showed that with increasing the temperature, the HCP to FCC phase transformation was suppressed while the {10$\bar{1}$1} compressive twinning was promoted. The activation of basal <a> dislocations at high temperature should be responsible for this phenomenon. With increasing the Ti content for the Zr-Ti alloys, the HCP to FCC phase transformation was suppressed and {10$\bar{1}$1} twinning was activated due to the increase in both the SFE and temperature.

Acknowledgements

This work was financially supported by the Natural Science Foundation of China (No. 51828102), the Natural Science Foundation of Hunan Province (No. 2018JJ3649) and The Project of Innovation-driven Plan in Central South University (No. 2019CX026). X.Z. Liao appreciates financial support from the Australian Research Council (No. DP190102243). The Advanced Research Center of Central South University is sincerely appreciated for TEM technical support.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jmst.2019.09.022.


/