Elsevier

Acta Materialia

Volume 87, 1 April 2015, Pages 283-292
Acta Materialia

Dislocation mechanisms in a zirconium alloy in the high-temperature regime: An in situ TEM investigation

https://doi.org/10.1016/j.actamat.2015.01.016Get rights and content

Abstract

Dislocation mechanisms responsible for the high-temperature mechanical properties of a Zr alloy have been investigated using in situ straining experiments between 250 °C and 450 °C. At 250 °C and 300 °C, the results show a steady and homogeneous dislocation motion in prismatic planes, with little cross-slip in the pyramidal and/or basal planes. At 350 °C, the kinetics of mobile dislocations becomes very jerky and inhomogeneous, in agreement with a dynamic strain aging mechanism. Above this temperature, the motion is again steady and homogeneous. Extensive cross-slip forms super-jogs which are efficient pinning points against the glide motion. These super-jogs move by glide along the Burgers vector direction, never by climb. The glide velocity between super-jogs is linear as a function of the total driving stress (applied stress minus line-tension stress due to dislocation curvature), in agreement with the solute dragging mechanism. The origin of the stress–strain rate dependence with an exponent larger than unity is then discussed.

Introduction

The origin of the high-temperature mechanical properties of Zr-α and Zr alloys (T > 200 °C) are still the subject of deep confusion and intricate controversies, although abundant experimental results are available and look most often fairly consistent.1 The reason for this unsatisfying situation is obviously that the dislocation mechanisms likely to occur above room temperature are not sufficiently understood. Different interpretations found in the literature are all based on the analysis of the temperature dependence (activation energy), stress dependence (activation volume, power-law exponent) of the deformation rate and post-mortem TEM observations [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21].

The temperature dependence will not be discussed here. Indeed, the activation energies measured in creep and constant strain-rate tests are very scattered, and no reliable conclusion can emerge from their comparison with various diffusion energies (also not accurately known). The stress dependence is much more informative, and most interpretations are actually based on the study of either activation volumes or stress exponents of the stress–strain rate dependence.

As shown below, two types of interpretations emerge from two types of experiments: whereas thermally activated glide is the natural interpretation of constant strain-rate tests, either climb-controlled recovery or thermally activated glide is proposed in creep tests. In other words, thermally activated glide may be the only mechanism involved in constant strain-rate and creep if both tests are considered to be not fundamentally different (apart from their different strain rates), but climb-controlled recovery may be rate-controlling in creep in the alternative case. The detailed results of both kinds are listed below.

  • Relaxation tests in constant strain-rate experiments yield activation volumes which all have the same variation as a function of temperature: an increase from ∼25 b3 at 200 °C to a peak at ∼350 °C, followed by a decrease to ∼50 b3 at 400 °C [1], [2], [3], [4], [5]. The peak, to which corresponds a plateau in the stress versus temperature curve, is unanimously interpreted by the occurrence of dynamic strain aging (DSA) [1], [2], [3], [4], [5], [6], [7], [8]. Note that several creep studies also conclude that DSA takes place in the same temperature range ([9], [10], see below). Below the peak, the controlling mechanism has been proposed to be the crossing of solute atoms (solid solution strengthening) according to [1], [2]. Above the peak, the controlling mechanism is unclear but has been proposed to be the climb of jogs on screw dislocations by [1].

  • Creep tests yield the stress exponent of the minimum creep rate versus stress dependence, n  6, in average. However, two sets of measurements and corresponding interpretations can be distinguished: (i) n independent of stress [11], [12], [13], interpreted in terms of classical climb-recovery controlled mechanisms, and (ii) n increasing with increasing stress [14], [15], interpreted by some thermally activated glide mechanism, either non-specified [14], or assumed to be the climb of jogs or super-jogs on screw dislocations [15], [16]. This second category also includes the results of Rupa [10] showing that n  3.7 at 300 °C, increases to ∼6.6 at 350 °C, and then decreases to ∼5.2 at 400 °C, in accordance with the peak of activation volume discussed above.

In brief, two types of mechanisms are likely to occur: (i) the thermally activated glide of dislocations, including DSA at about 350 °C, most often assigned to the climb of jogs/superjogs on screw dislocations, and (ii) the classical recovery-controlled creep where the climb-driven annihilation of dislocations in excess is rate-controlling.

If one considers that the occurrence of DSA is well established, there is however one surprising inconsistency in the corresponding models. Indeed, DSA which is the dynamic interaction between gliding dislocations and mobile solute atoms (the latter tending to catch and pin the dislocations), should be followed by the dragging of solute atoms at higher temperatures. This thermally activated mechanism has however never been assumed to play any role above the DSA domain, probably because its stress dependence (n = 1) does not fit with the experimental one (n  6). It should however not be completely rejected, and a minima be included in, or mixed with other mechanisms.

All these possible mechanisms are so different that they should be easily distinguished from each other, especially in in situ straining experiments where both the geometry and the kinetics of dislocation motion can be recorded as a function of stress and temperature. In particular, it should be fairly easy to identify glide or climb motion, as well as the most efficient obstacles to this motion, e.g., pinning points like jogs or solute dragging. Such experiments have been carried out between 250 °C and 450 °C, and the results are presented below.

Section snippets

Experimental

In situ experiments were carried out in a JEOL 2010 HC transmission electron microscope, with a high-temperature straining device developed in the CEMES, and the motion of dislocations was recorded by a Megaview III camera working at 25 images per second.

Cladding tubes and strips of a M5®2 zirconium alloy were provided by AREVA-CEZUS. The tubes were obtained by cold pilgering rolling, followed by

Geometry and kinetics of glide as a function of temperature

The experimental results described in this section show that three temperature domains can be defined, the intermediate one corresponding to dynamic strain aging. They also give details on the geometry of the high-temperature mechanism, i.e. the multiplication at sources in the prismatic plane, and the pinning and multiplication at super-jogs formed by cross-slip. Lastly, they show that super-jogs move exclusively by glide, never by climb.

A first set of results have been obtained in sample 1 at

Discussion

The first important result of this study is the observation of DSA at the scale of individual dislocations. Indeed, several in situ measurements in samples with various grain orientations consistently show a particular kinetics of mobile dislocations at 350 °C. Whereas dislocations move homogeneously and steadily below and above this temperature, their motion is very irregular and either very slow or very fast, as if intermediate velocities were forbidden. This picture corresponds to what is

Conclusions

In situ straining experiments in a zirconium alloy have yielded the following results:

  • Dynamic strain aging (DSA) has been observed at the scale of individual dislocations, at the same temperature of 350 °C as in macroscopic mechanical tests. It is characterized by an inhomogeneous deformation by a series of very fast individual dislocation movements, mostly in prismatic planes.

  • Above and below this temperature, the deformation proceeds by a homogeneous and steady dislocation motion.

  • At 400 °C,

Acknowledgements

The authors are indebted to Philippe Pilvin, Dominique Poquillon and Jean-Marc Cloué for stimulating the studies of creep mechanisms in zirconium alloys. The experiments have been funded by AREVA NP through the French METSA program.

References (39)

  • J.L. Derep et al.

    Acta Met.

    (1980)
  • P. Delobelle et al.

    J. Nucl. Mater.

    (1996)
  • W.R. Thorpe et al.

    J. Nucl. Mater.

    (1978)
  • C. Nam et al.

    J. Nucl. Mater.

    (2002)
  • H. Siethoff et al.

    Scripta Met.

    (1987)
  • J.H. Moon et al.

    J. Nucl. Mater.

    (2006)
  • I. Charit et al.

    J. Nucl. Mater.

    (2008)
  • C. Grosjean et al.

    Mater. Sci. Eng. A

    (2009)
  • M. Rautenberg et al.

    Acta Mater.

    (2012)
  • M. Rautenberg et al.

    Nucl. Eng. Des.

    (2014)
  • M. Tremblay et al.

    Mater. Sci. Eng.

    (1973)
  • G. Molénat et al.

    Mater. Sci. Eng.

    (1997)
  • D. Caillard et al.

    Scripta Mater.

    (2015)
  • D. Caillard

    Acta Mater.

    (2013)
  • B. Li et al.

    Surf. Sci.

    (1995)
  • J. Cadek

    Mater. Sci. Eng.

    (1987)
  • J. Cadek et al.

    Mater. Sci. Eng.

    (1995)
  • D. Mills et al.

    Trans. Met. Soc. AIME

    (1968)
  • Z. Trojanova et al.

    Czech J. Phys.

    (1985)
  • Cited by (31)

    • A coupled vacancy diffusion-dislocation dynamics model for the climb-glide motion of jogged screw dislocations

      2023, Acta Materialia
      Citation Excerpt :

      Accordingly, the total plastic strain accumulation is determined by the truncation rate as well as the effective climb rate of small jogs. More recently, Caillard et al. [55] proposed another scenario based on their in-situ TEM observations. They argued that screw dislocations on the prismatic plane are pinned by super-jogs formed by cross-slip.

    • Prismatic-to-basal plastic slip transition in zirconium

      2023, Acta Materialia
      Citation Excerpt :

      This clearly reflects the small friction of prismatic slip and the thermal-athermal transition of basal dislocation motion. These findings are consistent with TEM observations of both prismatic and basal dislocations in Zr at various temperatures [2,6,13]. According to the results shown in Section 3.1, Zr pillars with B2 orientation are of particular interest, which exhibit the prismatic-to-basal slip mode transition with increasing temperature.

    View all citing articles on Scopus
    View full text