Texture evolution and enhanced grain refinement under high-pressure-double-torsion

https://doi.org/10.1016/j.msea.2014.05.081Get rights and content

Abstract

We present a severe plastic deformation process called high-pressure-double-torsion (HPDT) for grain size refinement in metallic polycrystals. Like standard high-pressure torsion, HPDT monotonically imposes extreme plastic strains (⪢10) but via rotating both ends of the sample rather than one. Commercial purity Cu was subjected to HPT and HPDT for 1, 2, and 4 turns. The grain structure, hardness, and crystallographic texture were examined by transmission electron microscopy (TEM), Vickers microhardness tests, and X-ray diffraction (XRD) in both processes and compared in the mid-radius and at the edge of disks. We report that HPDT leads to finer grain sizes and higher hardness than HPT for the same number of turns. The measured textures exhibit typical shear components, which continuously strengthened with the plastic strain and also weakened with extreme grain refinement. These measurements also indicate that the texture gradients are lower in HPDT than HPT.

Introduction

Severe plastic deformation (SPD) techniques have proven a valuable method to transform coarse-grained metals into ultra-fine-grained and, in some cases, nanograined materials [1], [2], [3], [4], [5], [6]. They have been successfully applied to a variety of metallic systems, including both single-phase and dual-phase metals [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. The final SPD-processed materials with finer grains tend to be much stronger than their coarse-grained counterparts, following trends established by Hall–Petch [19], [20], [21], [22], [23], [24].

There is a suite of SPD techniques, including equal channel angular pressing or extrusion (ECAE) [25], accumulative roll bonding (ARB) [3], [5], [15], high-pressure torsion (HPT) [26], and wire drawing and bundling (WDB) [27], [28]. These SPD processes also differ in the geometry of the sample that they produce. In its original invented form, ECAE produces rods of square or round cross-section; ARB produces sheets; WDB makes wires; and HPT fabricates disks. Later, modified versions of these techniques were introduced in which the product geometry was altered. For instance, continuous ECAE was designed to apply the concept of ECAE to sheets [29], [30] and HPT-twist (HPTT) was created to use the HPT idea to hollow tubes [31]. With the exception of HPT, these processes are capable of making ultra-fine grained or nanocrystalline material in large quantities. While HPT makes relatively small amounts of material, it has proven invaluable for basic research studies on material behavior in extreme strains [26], phase transformations [32], [33], mechanical mixing [34], [35], and plasticity in nanoscale crystals [36], [37]. There are different types of HPT processing: unconstrained, constrained [26] and, now predominantly, quasi-constrained HPT [38]. In the constrained case, the initial sample dimensions do not change drastically or at least the diameter of the sample does not change since the technique confines the materials in a container [39]. In the unconstrained method, the material is free to move vertically and laterally [40]. In quasi-constrained HPT, like constrained HPT, there is some small outward flow between the upper and lower anvils [41].

Recently, much attention has been paid to increasing the efficiency of SPD techniques in refining the microstructure. The efficiency at which these techniques can reduce the grain size depends on material system, as well as processing parameters and conditions. For the same material, HPT tends to be more efficient compared to ECAE and ARB [42]. For instance, grain sizes in Cu can be refined to 150 nm by HPT [43], whereas it saturates to 270 nm after 6 ECAE passes [44]. To enhance refinement with fewer processing steps, researchers have combined more than one SPD technique, such as ECAE followed by HPT [24] or ARB followed by HPT [45].

In the present work, we introduce an extension of the HPT process called high pressure double torsion (HPDT), as a simple way of increasing the amount of strain imposed per turn and hence the grain size refinement efficiency compared to standard HPT. The main difference is that in HPDT both sides of the disk are being rotated, but in opposite directions. In this way, HPDT applies theoretically twice as much plastic strain per revolution than HPT. The higher the strain generated in the material the finer the grain size is expected to be until a threshold grain size is reached. To demonstrate, we subject the same commercially pure Cu to either unconstrained HPT or unconstrained HPDT. Using transmission electron microscopy (TEM) and hardness measurements, we show that grain refinement is enhanced in HPDT compared to HPT. We also reveal that HPDT leads to lower texture gradients per turn than HPT.

Section snippets

High-pressure-double-torsion

Fig. 1 shows schematics of the unconstrained HPT and HPDT tooling and process. A machine, able to perform both HPT and HPDT, was custom built. As shown, the primary difference is that in HPDT, the top and bottom anvils are both turning, whereas in HPT, only one is turning while the other is fixed. In HPDT, the top and bottom anvils are rotated in opposite directions, one counter-clockwise and the other clockwise.

Starting materials

To demonstrate the HPDT process, we use commercial purity Cu (99.9%), a material that has been widely used in many SPD processes [6], [18], [46]. The material was taken from an extruded rod and then annealed at 650 °C for two hours. Afterwards the material had weak texture and an initial grain size of approximately 30 µm (Fig. 2).

Processing

To systemically compare the HPT and HPDT processes, we applied the same initial pressure of 1.5 GPa and rotation rate ω of 0.2 rpm. Here in the HPDT technique, the

Finite element process simulations

Texture evolution can be linked to processing by performing finite element simulations. To better understand the measured texture gradients, we calculate strain distributions from a 3D finite element analysis using ABAQUS/Explicit [51]. Fig. 3 shows the meshed geometry of the sample and anvils. The model sample has the same relative dimensions as the actual sample. We elected to use 8-noded linear brick elements (C3D8R). Mesh sensitivity was tested for both HPDT and HPT after N=0.25 turns and

Microstructure

Higher strains can be expected to lead to faster grain refinement, at least in the first few turns under a saturated grain size is reached [62]. For a large-scale view, Fig. 4 presents a side-by-side comparison of the microstructure for HPT and HPDT from optical microscopy after N=1 turn. We see that the grains in the middle and edge areas of the HPDT disks are clearly smaller than those of the HPT disks. Fig. 5 shows TEM micrographs of the microstructure in the edge of the HPDT after N=2 and 4

Conclusions

In summary, we have presented an extended high-pressure torsion technique termed high-pressure-double-torsion (HPDT). We have shown that HPDT leads to faster grain refinement than standard high-pressure torsion (HPT). More efficient refinement can be linked to at least twice as much plastic strain imposed per turn in HPDT than HPT. The ability to impose higher strains can have a significant impact on studies of phase transformations, morphological evolution, interface stability, and nanoscale

Acknowledgments

MJ and MHP would like to acknowledge the financial support of Shiraz University through the grant no. 92-GR-ENG-16. IJB and SJZ gratefully acknowledge support by the Center for Materials at Irradiation and Mechanical Extremes, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, and Office of Basic Energy Sciences under Award Number 2008LANL1026. MK was supported by the University of New Hampshire faculty startup funds.

References (73)

  • R.Z. Valiev et al.

    Prog. Mater. Sci.

    (2006)
  • Y. Saito et al.

    Acta Mater.

    (1999)
  • K. Han et al.

    Acta Mater.

    (1998)
  • L. Jiang et al.

    Acta Mater.

    (2008)
  • Q. Xue et al.

    Acta Mater.

    (2007)
  • V.M. Segal et al.

    Mater. Sci. Eng. A

    (1997)
  • V. Vidal et al.

    Acta Mater.

    (2006)
  • S. Ohsaki et al.

    Acta Mater.

    (2007)
  • K. Yasuna et al.

    Mater. Sci. Eng. A

    (2000)
  • M. Al-Maharbi et al.

    Mater. Sci. Eng. A

    (2011)
  • Y.Z. Tian et al.

    Mater. Sci. Eng. A

    (2011)
  • M. Knezevic et al.

    Int. J. Plast.

    (2014)
  • M. Ardeljan et al.

    J. Mech. Phys. Solids

    (2014)
  • B. Mani et al.

    Powder Technol.

    (2012)
  • M. Jahedi et al.

    Mater. Sci. Eng. A

    (2010)
  • B. Mani et al.

    Mater. Sci. Eng. A

    (2011)
  • A.P. Zhilyaev et al.

    Progr. Mater. Sci.

    (2008)
  • L. Deng et al.

    Mater. Charact.

    (2013)
  • J.-C. Lee et al.

    Mater. Res. Bull.

    (2001)
  • J.-C. Lee et al.

    Acta Mater.

    (2002)
  • M. Arzaghi et al.

    Acta Mater.

    (2012)
  • K. Edalati et al.

    Mater. Sci. Eng. A

    (2009)
  • Y.F. Sun et al.

    Scr. Mater.

    (2011)
  • K. Edalati et al.

    Scr. Mater.

    (2012)
  • W. Skrotzki et al.

    Acta Mater.

    (2013)
  • R.B. Figueiredo et al.

    Mater. Sci. Eng. A

    (2011)
  • A.P. Zhilyaev et al.

    Acta Mater.

    (2003)
  • A. Vorhauer et al.

    Scr. Mater.

    (2004)
  • R.Z. Valiev et al.

    Prog. Mater. Sci.

    (2000)
  • Z. Horita et al.

    Mater. Sci. Eng. A

    (2005)
  • S.R. Kalidindi et al.

    Acta Mater.

    (2009)
  • H.Y. Um et al.

    Scr. Mater.

    (2014)
  • S.C. Yoon et al.

    J. Mater. Process. Technol.

    (2008)
  • M. Knezevic et al.

    J. Mech. Phys. Solids

    (2013)
  • M. Knezevic et al.

    Mater. Sci. Eng. A

    (2013)
  • M. Knezevic et al.

    Acta Mater.

    (2009)
  • Cited by (70)

    • Effect of powder reuse on tensile, compressive, and creep strength of Inconel 718 fabricated via laser powder bed fusion

      2022, Materials Characterization
      Citation Excerpt :

      Normally, the powder used in the LPBF process is spherical and the particle size distribution (PSD) is to be optimized for bulk density and fitting adequately within the layer for achieving a given layer thickness [31,32]. There are other important features like morphology, flowability, chemical composition, and humidity [32–36]. According to several studies [37,38], high density and repeatability of the finished parts can be achieved by using spherical powder with PSD of 10–45μm.

    View all citing articles on Scopus
    View full text