Experimental parameters influencing grain refinement and microstructural evolution during high-pressure torsion
Introduction
Considerable interest has recently developed in processing bulk materials through the application of severe plastic deformation (SPD) [1], [2]. These procedures are attractive because, unlike alternative processes involving nanopowder compaction or gas condensation, SPD processing is capable of producing large bulk samples that are free from any residual porosity. The basic principle of SPD processing is that a bulk material is subjected to a very intense plastic strain without introducing any significant changes in the overall dimensions of the work-piece. Thus, SPD processing differs in a very important way from more conventional metal-working processes, such as rolling, drawing and extrusion.
It is now well established that SPD processing can lead to a very significant refinement in the grain size of a wide range of materials including pure metals, metallic alloys and intermetallics [3], [4]. Furthermore, these ultrafine grain sizes, which are typically in the submicrometer or nanometer range, produce new and unusual physical properties, such as decreases in the elastic moduli, decreases in the Curie and Debye temperatures, increases in the rates of diffusion and improved magnetic properties [2]. If the ultrafine grain sizes are reasonably stable at elevated temperatures, there is a potential for achieving superplastic ductility at both unusually low testing temperatures and exceptionally rapid strain rates [5]. Very recent results have shown that SPD processing is capable of producing materials that combine both high strength and high ductility [6] and this unusual combination has been attributed specifically to the development of unique nanostructures in SPD.
Two different procedures are generally used in SPD processing. The first, known as equal-channel angular pressing (ECAP), involves pressing a bar or rod through a die within a channel bent into an L-shaped configuration [7]. The second, known as high-pressure torsion (HPT), involves subjecting a sample, in the form of a thin disk, to a high pressure and concurrent torsional straining [8]. Earlier experimental evidence suggested that HPT may be more effective than ECAP in producing exceptionally small grain sizes: for example, experiments on an Al–3% Mg solid solution alloy gave a grain size of ~90 nm when processing using HPT at room temperature [9] whereas a similar alloy processed by ECAP at room temperature yielded a grain size of ~270 nm [10]. The greater grain refinement achieved in HPT has been confirmed recently in two different ways. First, by processing the same alloy using both procedures where there are reports of grain sizes of ~150 nm after HPT and ~600 nm after ECAP in an Al–2% Fe alloy [11] and ~170 nm after HPT and ~350 nm after ECAP in high purity Ni [12]. Secondly, by subjecting samples of pure Ti to HPT after processing by ECAP where it was demonstrated that there was an additional reduction in the grain size from ~300 to ~200 nm [13]. In addition, whereas the grain sizes attained in ECAP are generally in the submicrometer range of ~100–1000 nm, numerous reports are now available documenting grain sizes of <100 nm in materials processed by HPT: for example, nanometer grains sizes have been reported in various Al alloys [14], [15], [16], a Cu nanocomposite [17], various Ni alloys [18], [19], Ni3Al [20], [21], [22], Ti–6% Al–4% V [23] and various steels [16], [24], [25], [26], [27], [28].
Although these results confirm the potential for producing remarkable microstructures using HPT, no systematic investigations have been conducted to date to evaluate the precise microstructural characteristics developed in HPT over a range of experimental variables (for example, as a function of the applied pressure and the total strain). This contrasts with processing by ECAP where there are detailed descriptions of the development and evolution of ultrafine-grained microstructures [29], [30]. Generally, it is found in HPT that an applied pressure of ~5 GPa and more than 5 rotations of the sample in torsion are sufficient to produce reasonably homogeneous microstructures throughout the sample with as-processed grain sizes of ~100 nm or smaller [2], [31], but the precise relationships linking the pressure, total strain and location within the sample have not been examined in any detail. The present investigation was initiated to address this deficiency by conducting a series of careful experiments on pure nickel subjected to HPT processing and by examining, throughout the deformed disks, the variation of the microhardness and the characteristics of the microstructure using transmission electron microscopy (TEM) and orientation imaging microscopy (OIM).
An important unresolved problem in experiments using HPT concerns determining the precise level of strain at any selected point within the sample. This problem is addressed in Section 2 and the subsequent sections describe the experimental procedures and results.
Section snippets
Estimating the strain in HPT
The principles of HPT processing are depicted schematically in Fig. 1. Fig. 1(a) shows the sample, in the form of a disk, located between two anvils where it is subjected to a compressive applied pressure, P, of several GPa at room temperature and simultaneously to a torsional strain which is imposed through rotation of the lower support. Surface frictional forces, therefore, deform the disk by shear so that deformation proceeds under a quasi-hydrostatic pressure. Fig. 1(b) depicts the
Material and HPT processing conditions
High purity nickel (99.99%) was selected for use in this investigation. There were two reasons for this choice. First, pure Ni was used in two earlier investigations documenting the characteristics of HPT [12], [31]. Second, it was established recently that pure Ni represents an ideal model material for investigations involving processing by SPD because the stacking fault energy, which is lower than for pure Al but higher than for pure Cu, leads to a significantly smaller grain size than is
Microhardness as a function of applied pressure and strain
Fig. 4 shows the influence of the applied pressure on the microhardness profiles for disks subjected to 5 whole revolutions under applied pressures of 1, 3, 6 and 9 GPa. For comparison, the influence of the magnitude of the torsional strain is recorded in Fig. 5, where the applied pressure was maintained constant at 6 GPa and the disks were subjected to 0.5, 1, 3 and 7 whole revolutions. The entire data are plotted in the form of 3-D meshes and the local hardness values within these meshes are
Discussion
This investigation leads to two important conclusions regarding the processing of samples by HPT. First, HPT is extremely effective in producing an exceptionally small grain size in bulk metals: in pure Ni, the present experiments give an average grain size of ~170 nm under optimum processing conditions, whereas the grain size achieved in the same pure Ni using ECAP was reported earlier as ~350 nm [12]. Second, the results provide very clear evidence that a reasonably homogeneous microstructure
Summary and conclusions
- 1.
Processing by HPT produces a significant increase in hardness and a very substantial grain refinement in bulk metals: in the present experiments on pure Ni, the hardness was increased from ~1.4 to >3 GPa and the grain size was reduced from ~100 μm to ~170 nm.
- 2.
Using the three separate procedures of microhardness measurements, TEM and OIM, it is shown that HPT may be used effectively to develop an essentially homogeneous microstructure throughout the sample provided the applied pressure and the
Acknowledgements
This work was partially supported by the US Army Research Office under Grant No DAAD19-00-1-0488. One of the authors (A.P.Z.) thanks the DGR of Generalitat of Catalonia and the CICYT (MAT2001-2555) for financial support.
References (41)
- et al.
Prog Mater Sci
(2000) - et al.
Mater Sci Eng
(1991) - et al.
Mater Sci Eng
(1993) - et al.
Scr Mater
(2002) - et al.
Nanostruct Mater
(1999) - et al.
Mater Sci Eng
(2001) - et al.
Mater Sci Eng
(2001) - et al.
Acta mater
(1999) - et al.
Acta mater
(2001) - et al.
Mater Sci Eng
(2001)
Nanostruct Mater
Acta mater
Acta mater
Scr Mater
Acta mater
J Mater Process Tech
Mater Sci Eng
Ultramicroscopy
Scr Mater
Mater Sci Eng
Cited by (754)
Phase decomposition behavior and its impact on mechanical properties in bulk nanostructured Cu-20 at.%Fe supersaturated solid solution
2024, Journal of Materials Science and TechnologyDiscrete modelling of continuous dynamic recrystallisation by modified Metropolis algorithm
2024, Computational Materials ScienceInterfacial characteristics of multi-material SS316L/IN718 fabricated by laser powder bed fusion and processed by high-pressure torsion
2024, Journal of Manufacturing ProcessesFrom unlikely pairings to functional nanocomposites: FeTi–Cu as a model system
2023, Materials Today AdvancesDevelopment and applications of aluminum alloys for aerospace industry
2023, Journal of Materials Research and TechnologyPassivation properties of ultrafine grain 316L stainless steels manufactured by ball milling and spark plasma sintering
2023, Materials Chemistry and Physics