Intensified texture of Zr702 sheet after slow cooling from near β-transus temperature

Abstract

Microstructures and textures of a Zr702 sheet subjected to slow cooling (air cooling (AC) and furnace cooling (FC)) from a near β-transus temperature (980°C) were characterised by electron channelling contrast imaging and electron backscatter diffraction techniques. Results show that textural intensities of both the AC and the FC are markedly higher than that of the initial specimen and the FC specimen owns the strongest texture. After both the heat treatments, the initial bimodal basal textural features are retained with the recrystallisation textural component (0°, 30°, 30°) becoming dominant but the deformation textural component (0°, 30°, 0/60°) largely weakened. The textural intensification is attributed to strong variant selection during the β → α phase transformation and slow cooling-induced sufficient growth of residual prior α grains.

Keywords

Zr sheet microstructure texture phase transformation variant selection misorientation electron channelling contrast imaging electron backscatter diffraction

Introduction

Thanks to low neutron absorption cross section, satisfactory strength and ductility, excellent biocompatibility and corrosion resistance, Zr and its alloys have important structural applications in nuclear, biomedical and chemical industries [1 –4]. Most Zr alloys have a close-packed hexagonal (hcp) structure (α-Zr) at room temperature. Since the hcp structure can only provide a limited number of slip systems during plastic deformation, the crystallographic texture is easy to be developed in α-Zr after processing [5 –7]. At elevated temperatures, α-Zr can be transformed into β-Zr with a body centred cubic (bcc) structure through allotropic phase transformation. During fabrications of Zr alloy products (like tubes and sheets), heat treatments at near the β-transus temperatures are often employed, which could not only markedly change their microstructural characteristics [8 –11], but also effectively alter the textures developed earlier [12,13].

It is known that the transformation between α-Zr and β-Zr usually obeys the Burgers orientation relationship (OR) [14], namely {0001}_α//{110}_β and ⟨11–20⟩_α//⟨111⟩_β. During the β → α phase transformation, the crystal symmetry allows 12 equivalent α variants to be produced from one prior β orientation according to the Burgers OR. In practice, however, variant selection may occur during β → α cooling, leading to only a few out of all the α variants to appear. When the variant selection is strong, the transformation textures can be very similar to or even stronger than the original one, which is called as texture inheritance/memory phenomena. To date, some efforts have been made to explore factors (like applied stresses and β grain sizes) affecting such variant selection in Zr alloys [15 –19]. Nevertheless, their textural characteristics after cooling from near β-transus temperatures are still not well documented with formation mechanisms unclear as well.

In the present study, therefore, a Zr702 sheet with a typical bimodal basal texture was subjected to slow cooling from a near β-transus temperature (980°C). Their microstructural and textural features after the β → α phase transformation were well characterised and analysed by use of electron channelling contrast (ECC) imaging and electron backscatter diffraction (EBSD) techniques.

Experimental

The as-received material is a rolled and annealed Zr702 sheet (containing 1.15 wt-% Hf, 0.07 wt-% Fe, 0.05wt-% Cr and 0.15 wt-% O) with an equilibrium β-transus temperature of ∼990°C (the determination method described in Ref. [20]). Rectangular specimens with dimensions of 11, 9 and 2 mm along rolling, transverse and normal directions (RD, TD and ND), respectively were cut from the as-received sheet. After cleaning surfaces, they were sealed into quartz tubes in vacuum and then heated at 980°C for 10 min in a box furnace (SX2-8-16). Subsequently, the specimens were cooled to room temperature in air and furnace, and denoted as AC and FC specimens, respectively. Cooling rates corresponding to AC and FC were estimated to be ∼5 and ∼0.02°C/s in a temperature range of 980°C and ∼800°C (roughly the α + β region), respectively [20,21].

A field emission gun scanning electron microscopy (Zeiss Sigma HD) equipped with a back-scattered electron detector and an EBSD system was employed for microstructural and textural characterisations. The EBSD system consisted of the Oxford Instruments NordlysMax² detector, with AZtec 2.4 and Channel 5 software packages used for data acquisition and analyses, respectively. Mechanical grinding was performed using silicon carbide paper (800-3000#) for the to-be-analysed surfaces (RD-ND) of the specimens before ECC and EBSD examinations, followed by electro-polishing in a mixture of 10% perchloric acid, 20% butyl cellosolve and 70% methanol at 20 V and −30°C for 40 s.

Results

Microstructural characteristics

As revealed in Figure 1(a) (ECC image), microstructures of the as-received material are mainly comprised of well-equiaxed grains with relatively uniform sizes. There are also some intra- or inter-granular black dots randomly distributed and composition measurements suggest that they are second phase particles (SPPs) containing Fe and Cr. Grain sizes in the initial microstructure are measured by the linear intercept method to be about 8.3 µm on average. Figure 1(b) is an EBSD inverse pole figure (IPF) map with high angle boundaries (HABs, θ > 15°) and low angle boundaries (LABs, 2° < θ < 15°) indicated by black and grey lines, respectively. It can be seen that most grains are clearly separated by HABs while there exist few LABs. According to the colour code shown in Figure 1(b), the majority of grains have their c-axes aligned close to the ND, suggesting the existence of grain orientation preference (crystallographic texture) in the as-received material. Figure 1(c) is a misorientation angle distribution histogram which reveals that most misorientation angles are higher than 15° while those lower than 15° are rare. In general, the above microstructural characteristics are typical for Zr alloy sheets after recrystallisation annealing [22,23].

Figure 2 presents ECC images of the heat-treated specimens. From Figure 2(a,b), one can see that most prior equiaxed grains were replaced by plate structures in the AC specimen with a small number of equiaxed grains retained. Also, many tiny SPPs are observed at plate/grain boundaries (as arrowed in Figure 2(b)), which should be resulted from diffusion and segregation of alloying elements during the air cooling [24]. With respect to the FC specimen (Figure 2(c,d)), in spite of the existence of a few plate structures, the overwhelming majority of grains keep the equiaxed morphology like in the initial microstructure. A few SPPs preferably distributed along grain boundaries appear in the FC specimen (Figure 2(d)), different from their random distribution in the initial microstructure (Figure 1(a)). A comparison between Figures 1(a) and 2(d) suggests that sizes of the equiaxed grains and the SPPs in the FC specimen are evidently increased, compared to those in the as-received material. A part of these equiaxed grains may directly grow from residual prior α grains during heating while the other equiaxed grains (along with very few plate structures) should be transformed from high-temperature β phase during cooling. It has to be admitted that reliable distinguishing between them is difficult due to very similar appearance.

Microstructural characteristics revealed by EBSD for the AC and the FC specimens are shown Figure 3. It can be more clearly seen from Figure 3(a,c) that there are both plate structures and equiaxed grains in the AC specimen while the microstructure of the FC specimen essentially consists of equiaxed grains, similar to ECC observations (Figure 2). By use of the linear intercept method, average grain sizes of the AC and the FC specimen are measured to be 11.9 and 22.6 µm, respectively. Figure 3(b,d) (IPF images) show that most grains in both the specimens are coloured (near) red, consistent with grain orientation features of the as-received specimen (Figure 1(b)). Figure 4(a,b) further show the misorientation angle distributions (MAD) of the AC and the FC specimens, from which both of them are found to have greatly different MAD characteristics from that of the as-received material (Figure 1(c)). Two distinct peaks around 60° and 90° are observed for the AC specimen, while the FC specimen has three peaks around 10°, 60° and 90°. According to the Burgers OR [14], there exist five specific misorientation angles between different α variants transformed from one β grain, namely 10.5°, 60°, 60.8°, 63.3°and 90° [25]. The presence of these misorientation angle peaks in the AC and the FC specimen is well consistent with the theoretical predictions and suggests that the Burgers OR is well followed during the β → α cooling [26]. The other misorientations not corresponding to the Burgers misorientations (Figure 4(a,b)) should have reflected grain boundary characters of the prior α grains untransformed during the heat treatments.

Textural characteristics

Pole figures of specimens before and after the heat treatments are presented in Figure 5. To reliably compare their textural characteristics, the pole figure of each specimen is derived from the same EBSD scanning area (900 × 600 µm²) that is believed to be large enough (containing thousands of grains). Figure 5(a) reveals that the as-received material owns a typical bimodal basal texture with basal maxima (5.9 times of random) tilted about ±20–40° from the ND towards the TD and the ⟨11-20⟩ or ⟨10-10⟩ of most grains aligned close to RD. For the AC specimen (Figure 5(a)), although a few new textural components emerge, their intensities are generally weak. In contrast, the initial bimodal basal texture characteristics are hardly changed with the intensity value increased to 9.7. Referring to its {11-20} and {10-10} pole figures, it seems that the prior⟨10-10⟩//RD component almost disappears while the ⟨11-20⟩//RD component is significantly enhanced. This means that after air cooling from 980°C, the specimen texture is intensified, accompanied by significant textural component preference. The major textures of the FC specimen are also featured by the retained bimodal basal texture with the enhanced ⟨11-20⟩//RD component (Figure 5(c)), similar to those of the AC specimen. After the slower cooling in the furnace, the specimen textural intensity is remarkably increased to 16.0.

Figure 6 presents distributions of deviation angles between grain c-axes and the ND in various specimens. One can see from Figure 6(a), such deviation angles in the as-received material concentrate on ∼20–40°, which is consistent with the bimodal basal texture distribution revealed in Figure 5(a). Figure 6(b) shows that the deviation angles in the AC specimen also have evident peaks around 20–40°. Meanwhile, concentrations at about 65° and 80–90° are noticeable as well, which may correspond to the new textural components in the basal pole figure in Figure 5(b). For the FC specimen, similar concentrations around 20–40° can be seen for the deviation angles (Figure 6(c)) and their frequencies appear to be much higher than those of both the as-received and the AC specimens. This agrees with the results revealed by pole figures in Figure 5 that the FC specimen has the highest textural intensity.

Orientation distribution function (ODF) maps are frequently used to provide quantitative descriptions for textures and Figure 7 presents constant ODF sections at ϕ1 = 0° (containing the most important information) of all the specimens. From Figure 7(a), the main textural features of the as-received material may be described as (0°, 30°, 0–60°). In fact, Zr alloy sheets are usually known to develop a textural component of (0°, 30°, 0/60°) after deformation while a recrystallisation component of (0°, 30°, 30°) can be produced after sufficient grain growth [23]. Thus, the texture of the as-received material can be approximately regarded as the combination of both the deformation and the recrystallisation components with comparable textural intensities. Note that the deformation and the recrystallisation components represented by Euler angles essentially correspond to ⟨10-10⟩//RD and ⟨11-20⟩//RD in Figure 5, respectively [27]. After the heat treatment, the texture of the AC specimen seems to be largely dominated by the recrystallisation component with the deformation component comparatively weakened (Figure 7(b)). Moreover, the deformation component almost completely disappears in the FC specimen, with only the recrystallisation component left (Figure 7(c)). In general, changes in textural intensity in each specimen agree well with those revealed by pole figures in Figure 5.

Although Figures 2(a) and 3(a) reveal that both plate structures and equiaxed grains exist in the AC specimen, only the plate structures should be produced by β → α cooling. By use of their shape difference, they can be feasibly separated and further analysed individually by the EBSD post-processing software (Channel 5), which would facilitate understanding reasons for the textural variation induced by the phase transformation. Figure 8 presents the EBSD results with the plate structures and the equiaxed grains in the AC specimen separately presented. From Figure 8(a,b), the plates seem to have diverse colours and their c-axes are often deviated from the ND, different from the case of the as-received material. Figure 8(c) reveals that the peaks corresponding to the Burgers misorientations (around 10°, 60° and 90°) are more intensified than those shown in Figure 4(a), confirming such plates to be products of the β → α transformation. It can be seen from the basal pole figure in Figure 8(d) that the c-axis distribution is more scattered compared to that in Figure 5(a). Nevertheless, the maximum textural intensity of the plate structures (8.3 times of random) is still higher than that of the as-received material, with the basal bimodal texture and the ⟨11-20⟩//RD component retained. This indicates that although some new α orientations can be produced during the phase transformation, the features of the transformation texture are still mainly determined by those of the as-received material, possibly related to the occurrence of strong variation selection.

Figure 8(e),(f) are EBSD maps showing only the equiaxed grains, from which less scattered orientations (colours) than those plate structures (Figure 8(b)) can be noticed. For these extracted equiaxed grains, Figure 8(g) shows their misorientation angle distribution is completely different with that predicted by the Burgers OR but closer to that in Figure 1(c), confirming that they are evolved from the prior equiaxed grains. Figure 8(h) reveals that their textural intensity (15.2 times of random) is significantly higher than that of the as-received material, suggesting that considerable texture strengthening has occurred along with the growth of prior equiaxed α grains during heating at 980°C. Also as a result of grain growth, the untransformed equiaxed grains seem to preferably develop the ⟨11-20⟩//RD component, the same major textural component as that of the plate structures (Figure 8(d)). In addition, by use of Figure 8(a,e), the area fractions of the plate structures and the equiaxed grains are measured to be ∼72% and ∼28%. Note that the fraction of the untransformed equiaxed grains is higher than their equilibrium fraction during annealing, which is estimated to be <10% by examining a water-quenched specimen at 980°C. This should be related to further growth of these equiaxed grains during slow cooling in air.

Discussion

According to the above results, in spite of drastic microstructural changes, the main features of the initial texture are retained (texture inheritance) after air and furnace cooling from the near β phase region (980°C) and the textural intensities are markedly increased (Figures 5 and 8). When Zr alloys are subjected to slow β cooling (like AC and FC in this work), the α phase tends to preferably nucleate at β boundaries and then grow towards their interiors [11,24,28]. Although one β orientation can give birth to 12 different α variants according to the Burgers OR, these variants usually have different chances to appear, i.e. the variant selection occurs. A previous study on the β → α transformation of Ti alloys [29] suggested that the α nuclei formed near β boundaries would prefer orientations maintaining simultaneously the Burgers OR with the β grains on both sides. Such variant selection behaviour would be greatly enhanced when ⟨110⟩ axes of adjacent β grains were parallel or close to each other, leading to intensified transformation textures. Recently, this boundary-related mechanism has also been confirmed in Zr alloys [30]. Besides, Romero et al. [15] pointed out that in Zr-2 alloys heated at temperatures lower than the β-transus, residual prior α grains could act as α-nucleation sites during subsequent cooling. α variants with orientations similar to these prior α grains would then be preferably selected, resulting in texture inheritance. Since the annealing temperature (980°C) in this work is in the α + β region, a certain amount of residual prior α grains exist and they should be able to affect the α-variant selection during cooling. In an earlier study [21], the Zr702 air-cooled after annealing at 1000°C (above the β transus) was found to own a much weaker and more scattered texture than the present AC specimen, verifying the important role played by the residual α grains in inheriting the original texture. Nevertheless, the fraction of the untransformed α grains is relative low (<10%) during annealing at 980°C because it is close to the β-transus of the Zr702 sheet. Thus, the mechanism based on β-boundary characteristics should have also affected the α-variant selection.

With respect to reasons why the FC specimen has a stronger texture than the AC specimen, on the one hand, the lower cooling rate could increase the nucleation advantage of α phases at β boundaries and then enhance the above-mentioned variant selection behaviour [28]. On the other hand, slow cooling allows the residual α grains (untransformed during annealing) to easily grow, which will promote the deformation component (0°,30°,0/60°) to be replaced by the recrystallisation component (0°,30°,30°) with higher intensities [31]. This can also be verified by comparing textural features of the equiaxed grains in the as-received material and those retained in the AC specimen (Figures 5(a) and 8(h)). The grain growth is more sufficient in the FC specimen with slower cooling than the AC specimen (Figures 2 and 3), which could contribute to the intensified texture.

Conclusions

After cooling from the near β-transus temperature (980°C), both plate structures and equiaxed grains exist in the AC specimen while the microstructure of the FC specimen mainly consists of coarse equiaxed grains. The presence of typical Burgers misorientation suggests the Burgers OR is still obeyed during the slow β → α transformation.

Textural intensities of both the AC and the FC are markedly higher than that of the initial specimen and the FC specimen owns the strongest texture. The initial bimodal basal textural features are retained after the heat treatments, with the recrystallisation textural component (0°, 30°, 30°) becoming dominant and the deformation textural component (0°, 30°, 0/60°) weakened (even disappeared).

The textural intensification can be attributed to strong variant selection during the β → α phase transformation and slow cooling-induced sufficient growth of residual prior α grains. The variant selection is closely related to the existence of such residual α grains and specific β-boundary characteristics developed at 980°C.

Footnotes

Acknowledgements

Ms. Tingting Wang is acknowledged for providing assistance in post-processing EBSD data. The reviewers of this paper are also gratefully appreciated for their critical comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Linjiang Chai

Ning Guo

References

Zinkle

, Was

GS.

Materials challenges in nuclear energy. Acta Mater. 2013;61:735–758. doi: 10.1016/j.actamat.2012.11.004

Yau

TL.

Corrosion of Zirconium and its alloys. Shreir's Corros. 2010;3:2094–2134. doi: 10.1016/B978-044452787-5.00101-3

Chen

L-Y

, Shen

, Zhang

, Corrosion behavior of non-equilibrium Zr-Sn-Nb-Fe-Cu-O alloys in high-temperature 0.01 M LiOH aqueous solution and degradation of the surface oxide films. Corros Sci. 2018;136:221–230. doi: 10.1016/j.corsci.2018.03.012

Mehjabeen

, Song

, Xu

, Zirconium alloys for Orthopaedic and Dental applications. Adv Eng Mater. 2018;20:1800207. doi: 10.1002/adem.201800207

Murty

, Charit

Texture development and anisotropic deformation of zircaloys. Prog Nucl Energy. 2006;48:325–359. doi: 10.1016/j.pnucene.2005.09.011

Luan

, Gao

, Chai

, Compression deformation behavior of Zr–1Sn–0.3Nb alloy with different initial orientations at 650°C. Mater Des. 2013;52:1065–1070. doi: 10.1016/j.matdes.2013.06.069

Tenckhoff

Review of deformation mechanisms texture and mechanical anisotropy in zirconium and zirconium base alloys. J ASTM Int. 2005;2:25–50. doi: 10.1520/JAI12945

Holt

RA.

The beta to alpha phase transformation in zircaloy-4. J Nucl Mater. 1970;35:322–334. doi: 10.1016/0022-3115(70)90216-3

Zhang

, Zhang

, Yang

, Effect of cooling process on the formation of duplex microstructure in Zr–2.3Nb alloy. J Alloys Comp. 2015;651:316–321. doi: 10.1016/j.jallcom.2015.08.105

10.

Jeong

, Rheem

, Choi

, Effect of beta heat treatment on microstructure and nodular corrosion of zircaloy-4. J Nucl Sci Technol. 1993;30:154–163. doi: 10.1080/18811248.1993.9734463

11.

Chai

, Luan

, Chen

, Effect of cooling rate on β→α transformation during quenching of a Zr-0.85Sn-0.4Nb-0.4Fe-0.1Cr-0.05Cu alloy. Sci China Technol Sci. 2012;55:2960–2964. doi: 10.1007/s11431-012-4956-9

12.

Dahlback

MLM

, Hallstadius

, Barberis

, The effect of beta-quenching in final dimension on the irradiation growth of tubes and channels. J ASTM Int. 2005;2:276–304.

13.

Grytsyna

, Stukalov

, Chernyayeva

, Destruction of crystallographic texture in zirconium alloy tubes. J ASTM Int. 2005;2:305–329. doi: 10.1520/JAI12339

14.

Burgers

WG.

On the process of transition of the cubic-body-centered modification into the hexagonal-close-packed modification of zirconium. Physica. 1934;1:561–586. doi: 10.1016/S0031-8914(34)80244-3

15.

Romero

, Preuss

, Quinta da Fonseca

Texture memory and variant selection during phase transformation of a zirconium alloy. Acta Mater. 2009;57:5501–5511. doi: 10.1016/j.actamat.2009.07.046

16.

Daymond

, Holt

, Cai

, Texture inheritance and variant selection through an hcp–bcc–hcp phase transformation. Acta Mater. 2010;58:4053–4066. doi: 10.1016/j.actamat.2010.03.012

17.

Sattari

, Holt

, Daymond

MR.

Variant selection and transformation texture in zirconium alloy Excel. J Nucl Mater. 2014;453:120–123. doi: 10.1016/j.jnucmat.2014.06.045

18.

Wenk

, Lonardelli

, Williams

Texture changes in the hcp→bcc→hcp transformation of zirconium studied in situ by neutron diffraction. Acta Mater. 2004;52:1899–1907. doi: 10.1016/j.actamat.2003.12.029

19.

Gey

, Humbert

, Gautier

, Study of the β→α variant selection for a zircaloy-4 rod heated to the β-transus in presence or not of an axial tensile stress. J Nucl Mater. 2004;328:137–145. doi: 10.1016/j.jnucmat.2004.03.003

20.

Qiu

, Luan

, Chai

, Precise determination of the α → α + β phase transformation temperature of Zr-1.0Sn-0.3Nb-0.3Fe alloy. Sci China Technol Sci. 2013;56:60–65. doi: 10.1007/s11431-012-5063-7

21.

Chai

, Luan

, Chen

, Concurrent inheritance of microstructure and texture after slow β→α cooling of commercially pure Zr. Sci China Technol Sci. 2016;59:1771–1776. doi: 10.1007/s11431-016-0306-7

22.

Chai

, Luan

, Xiao

, Microstructural and textural evolution of commercially pure Zr sheet rolled at room and liquid nitrogen temperatures. Mater Des. 2015;85:296–308. doi: 10.1016/j.matdes.2015.06.088

23.

Zhu

, Bacroix

, Chauveau

, Texture evolution and associated nucleation and growth mechanisms during annealing of a Zr alloy. Metall Mater Trans A. 2009;40:2423–2434. doi: 10.1007/s11661-009-9909-y

24.

Crepin

, Bretheau

, Caldemaison

, Microstructural study of beta treated grade 702 zirconium: scanning electron microscopy and transmission electron microscopy, complementarity of two observational scales. J Mater Sci. 1997;32:4841–4846. doi: 10.1023/A:1018603619366

25.

Chauvy

, Barberis

, Montheillet

Microstructure transformation during warm working of β-treated lamellar Zircaloy-4 within the upper α-range. Mater Sci Eng A. 2006;431:59–67. doi: 10.1016/j.msea.2006.05.031

26.

Chai

, Chen

, Zhou

A special twin relationship or a common Burgers misorientation between α plates after β quenching in Zr alloy?

Mater Charact. 2015;104:61–65. doi: 10.1016/j.matchar.2015.04.008

27.

Wang

, Huang

JC.

Texture analysis in hexagonal materials. Mater Chem Phys. 2003;81:11–26. doi: 10.1016/S0254-0584(03)00168-8

28.

Massih

, Andersson

, Witt

, Effect of quenching rate on the β-to-α phase transformation structure in zirconium alloy. J Nucl Mater. 2003;322:138–151. doi: 10.1016/S0022-3115(03)00323-4

29.

Stanford

, Bate

PS.

Crystallographic variant selection in Ti–6Al–4 V. Acta Mater. 2004;52:5215–5224. doi: 10.1016/j.actamat.2004.07.034

30.

Ahmmed

, Daymond

, Gharghouri

MA.

Microstructural evaluation and crystallographic texture modification of heat-treated zirconium Excel pressure tube material. J Alloys Comp. 2016;687:1021–1033. doi: 10.1016/j.jallcom.2016.05.072

31.

Dewobroto

, Bozzolo

, Barberis

, On the mechanisms governing the texture and microstructure evolution during static recrystallization and grain growth of low alloyed zirconium sheets (Zr702). Int J Mater Res. 2006;97:826–833.