Microstructure and Mechanical Properties of Bulk Nanocrystalline Al88Mm5Ni5Fe2 Alloy Consolidated at High Pressure

Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 1 Microstructure and Mechanical Properties of Bulk Nanocrystalline Al88Mm5Ni5Fe2 Alloy Consolidated at High Pressure H. Dimitrov1, J.S. Blázquez2*, J. Latuch1, T. Kulik1 1Faculty of Materials Science and Engineering, Warsaw University of Technology, Wołoska 141, 02-507 Warsaw, Poland 2 Departamento de Física de la Materia Condensada, Instituto de Ciencia de Materiales, CSIC-Universidad de Sevilla, Apartado 1065, 41080 Sevilla, Spain Abstract: Bulk nanocrystalline Al88Mm5Ni5Fe2 alloys have been produced by consolidation of pulverised melt-spun ribbons at high pressure. Different production procedures were explored to improve the quality of compaction of the resulting bulk samples. Quality of compaction of samples pressed at room temperature is clearly improved by increasing applied pressure from 2 to 7.7 GPa. All hot compacted samples had good quality of compaction. Onset of crystallisation shifts to higher temperatures as the applied pressure increases. Nanocrystalline powder fails to be compacted at room temperature even at 7.7 GPa. Mechanical properties were studied in terms of Vickers microhardness. Relationship between microhardness and microstructure of bulk samples was studied in the frame of two different theoretical models. Keywords: A-aluminides (miscellaneous), C-nanocrystals, F-mechanical testing. * Corresponding author: J. S. Blázquez Tel.: +34-954559541 Fax: +34-954612097 e-mail: [email protected] Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 2 1 Introduction Amorphous Al-based alloys are interesting for their excellent mechanical properties, superior to those of the commercial crystalline Al-based alloys (e.g. tensile strength in amorphous alloys is twice higher compared to commercial crystalline ones) [1]. However, mechanical properties can be still improved in some compositions, exhibiting a primary crystallisation in which nanocrystalline microstructure is developed [2]. This microstructure consists of dispersed nano-sized crystals (~20 nm) of -Al embedded in a residual amorphous matrix. One successful way for achieving nanocrystalline microstructure is the controlled crystallisation of a precursor amorphous alloy. This amorphous microstructure can be achieved by rapid quenching techniques, e.g. melt-spinning technique, applied to quaternary systems containing Al, rare earth (RE) or Y, and two transition metals (TM) [3]. Although the main interest of Al-based alloys applicability is as bulk samples, the typical procedures for obtaining amorphous or nanocrystalline alloys lead to thin ribbons (melt-spinning, typically tens m thick) or powder samples (atomisation or mechanical alloying). Therefore, much effort has been devoted to powder consolidation by different procedures, as hot extrusion [4-6], hot pressing [7,8] and cold consolidation using severe plastic deformation [9]. On the other hand, relationship between microhardness and microstructure of nanocrystalline Al-based alloys is not fully understood yet, and different models predict different dependence of microhardness on crystalline volume fraction [10-13]. In this work, different ways of production of bulk nanocrystalline samples were explored and the relationship between hardness and nanocrystalline microstructure was studied. Al88Mm88Ni5Fe2 alloy was chosen for this study, where Mm denotes Mischmetal, which is a mixture of different lanthanides. The use of Mm highly reduces the cost of the alloy, without big effect on the crystallisation process, with respect to compositions Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 3 containing pure rare earth elements [14]. Previous work on the studied composition [15] showed that devitrification process occurs in three different stages. Nanocrystalline microstructure is developed during primary crystallisation (crystallisation onset temperature at 40 K min-1: 514 K), -Al nanocrystals are embedded into a residual amorphous matrix. A new metastable phase appears during the second crystallisation process and, after the third crystallisation event, final stable phases found are: -Al, Al11Mm3 and Al3Ni [16]. 2 Experimental Ingots of quaternary Al88Mm5Ni5Fe2 alloy were prepared from pure elements by arc melting in argon atmosphere. Composition of Mischmetal (Mm), in at. %, was: Ce-50.3, La43.5, Pr-5.9 and Nd-0.3. During melt-spinning, the melt is ejected from the crucible onto a rotating copper wheel at peripheral speeds of 30-40 m s-1. By this technique it is possible to quench the melt at a rate of 105-106 K s-1. The resulting ribbons were typically 2-3 mm wide and 30-40 m thick. The melt-spun amorphous ribbons were ball milled in a Fritsch P5 planetary ball mill for 90 min in argon atmosphere at a rotational speed of 250 rpm. Ball to powder ratio was 20:1. In order to obtain nanocrystalline powder, amorphous ribbons were vacuum-sealed in quartz crucibles and annealed at 508 K in a laboratory tube furnace for 20 min. The resulting nanocrystalline ribbons were ball milled as the amorphous melt-spun ones but for 60 min and the same ball to powder ratio and rotational speed. A high-pressure toroidal-cell press was used for compacting powders. The powder was initially pre-pressed uni-axially at 100 MPa and the resulting green compact was inserted into the toroidal cell. The shape and the material of the gasket ensure that the compacting conditions are close to isostatic ones. The applied pressures were 2 and 7.7 GPa. Temperatures applied in our experiments were: room temperature (300 K), 548 K and 573 K. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 4 Samples were hold at maximum pressure and room or elevated temperatures for 5 min. Resulting bulk samples are limited to cylinders of 5 mm diameter and a maximum of 5 mm high. Vickers microhardness measurements were performed on a Zwick hardness tester with a load of 100 g. To improve statistic of the results, 10 measurements were performed on each tested bulk sample (5 on both faces). The surface was previously polished by hand. The error was estimated from the deviation of the set of measurements to be ~25 HV. During the experiments, the size of the indenter mark was >15 m. Quality of compaction of the as-pressed bulk samples was determined by scanning electron microscopy (SEM), in a Hitachi S-3500N equipment, as well as density measurements. SEM observations were performed after hardness tests on polished surface. Besides some of the samples with bad quality of compaction (preventing hardness testing) were also observed by SEM. In such cases, the surface was not polished. Density of compacted bulk samples was determined by Archimedes method using a Gibertini E154 balance with a set for solid state density measurements. Before the samples were subjected to weight measurements, they were kept in boiling water. Differential scanning calorimetry (Perkin-Elmer DSC7) was used in argon atmosphere for characterising the crystallisation behaviour of amorphous ribbons and powder samples. The same equipment was used for annealing of the bulk samples. The heating rate used was 40 K/min. Phase composition and microstructure were determined using X-ray diffraction technique (XRD) in a Philips diffractometer operated with Cu K wavelength. Crystalline volume fraction (Vcr) and average crystalline size were estimated by this technique. To obtain Vcr, a deconvolution procedure was performed to separate crystalline peaks and amorphous halo in a 2 range from 27.5 to 47.5 degrees involving the amorphous halo and the crystalline Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 5 diffraction maxima of the -Al phase (111) and (200) [16]. Crystalline peaks were fitted using Lorentzian function, assuming the small grain size as the main effect contributing to the broadening of the diffraction peaks. Amorphous halo was fitted using a Gaussian function. Vcr was estimated from the area ratio between the (111) peak and the sum of (111) peak and the amorphous halo. This area ratio was corrected taking into account the different scattering factor of the Al (which can be assumed as the only element in the crystalline phase [17]) and those of the elements composing the residual amorphous matrix (enriched in Mm, Ni and Fe elements and impoverished in Al), as it is detailed elsewhere for other nanocrystalline systems [18]. The main advantage of this measurement of Vcr with respect to the widely used enthalpy fraction from DSC is that, in calorimetric measurements, the difference in the enthalpy associated to the formation of different phases is neglected and possible recrystallisation processes can lead to misleading of the results. The procedure for Vcr measurement detailed here is expected to give absolute values of this magnitude with an error below 10 %. Average grain size was estimated using Scherrer’s formula. In the present work, the microstrain effects and the possible presence of a grain size distribution were neglected. However, as the approximations are imposed for all the samples studied in the same way, the obtained values are trustworthy as an estimation of the order of the nanocrystal size and an error bar of 5 nm was assumed. Bulk samples had a surface smaller than the illuminated area by the X-ray beam. Therefore, special care was taken for positioning of the samples. Before the deconvolution procedure was applied, background was substracted using the registered XRD pattern of the empty holder, specially designed to match the shape of the bulk samples. 3 Results 3.1 Production procedures Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 6 Figure 1 schematically shows the three ways of production of bulk nanocrystalline material proposed in this work, each of them starting from amorphous ribbons obtained by melt-spinning technique. Figure 2 shows the XRD pattern and DSC scan corresponding to ascast amorphous ribbons. The three procedures are listed below. Bulk samples from amorphous powder pressed at room temperature. Amorphous powder was obtained by ball-milling of amorphous ribbons. Figure 3 shows the XRD pattern and the DSC scan of the amorphous powder obtained by milling. The powder was pressed at room temperature at different applied pressures (2 and 7.7 GPa, maximum pressure was applied for 5 min), to obtain bulk amorphous samples. Nanocrystalline microstructure was achieved by subsequent annealing at 508 K (about 5 K below the onset of crystallisation measured at heating rate of 40 K/min [15]) for different times (from 5 to 60 min). Figure 4 shows the XRD patterns of the final nanocrystalline bulk samples after the different annealing processes. The a priori advantage of this procedure is that the amorphous state of the powder is preserved during pressing and the nanocrystalline microstructure is achieved by a well controlled annealing process. Bulk samples from amorphous powder pressed at high temperature. The procedure followed is similar to the previous one, but instead of cold compaction (pressing at room temperature), hot compaction was performed. Therefore, two parameters were changed in these experiments: pressure and temperature of compaction. The time at maximum pressure, which is coincident with the time the sample was submitted to isothermal annealing, was 5 min. Figure 5 shows the XRD patterns of the different samples obtained by this procedure. It can be seen that not all the samples developed a nanocrystalline microstructure, although the pressing temperature was as high as 35 K over the onset temperature registered by DSC at 40 K/min (see Fig. 2b and Fig. 3b). From figure 5 it can be concluded that the onset of crystallisation shift to higher temperature as the applied pressure increases: sample pressed at Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 7 548 K and 7.7 GPa has an amorphous XRD pattern, meanwhile sample pressed at the same temperature at 2 GPa evidences clear crystalline peaks corresponding to the -Al phase, which was analysed after subtraction of the baseline as it is described in the Experimental section. The bulk sample obtained in amorphous state by this procedure was subsequently annealed in the same way as the cold compacted samples to achieve a controlled nanocrystalline microstructure. The main a priori advantage of this procedure is that one extra process could be avoided if the nanocrystalline microstructure is achieved (annealing and pressing are done at the same time). The annealing conditions at which the nanocrystalline microstructure is developed are less controllable than in a furnace. Another important advantage that might be pointed for the hot compacted samples with respect to the cold compacted samples is that high temperature pressing might enhance the quality of compaction of the resulting bulk material, enhancing the binding among powder particles. Bulk samples from nanocrystalline powder pressed at room temperature. In this case, nanocrystalline microstructure was achieved by annealing of amorphous ribbons at 508 K for 20 min. The nanocrystalline microstructure developed was characterized by a crystalline volume fraction of 0.3 and an average grain size of 19 nm. Afterwards, nanocrystalline ribbons were milled in order to produce nanocrystalline powder, which was compacted at room temperature at two different pressures, 2 and 7.7 GPa. This procedure presents the same advantage as the first procedure and the nanocrystalline microstructure can be finely controlled by annealing the amorphous ribbons. Besides, a new a priori advantage appears, because nanocrystalline ribbons present a higher thermal stability than that of the amorphous structures, facilitating that the milling process does not affect the microstructure. 3.2 Quality of compaction Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 8 The quality of compaction of the as-pressed samples was checked using SEM and density measurements. Three are the main parameters which could affect the quality of compaction: microstructure of the powder, temperature of pressing and applied pressure. Figure 6 shows the SEM micrographs of the bulk samples obtained by cold compaction of amorphous powder (the scratches are due to the hand polishing of the samples as they were prepared for hardness testing). Density is also indicated. In this case, only compaction pressure varies. Some individual smashed particles could be observed on the sample pressed at 2 GPa, whereas, for the sample compacted at 7.7 GPa, homogeneous flat surface, indicating a good quality of compaction was observed. Density is also lower in the sample compacted at 2 GPa than in the sample compacted at 7.7 GPa (3.06 and 3.18 g cm-3, respectively). From these results it can be concluded that, in the region explored, pressure has a big effect on quality of compaction, being noticeable the effect of increasing pressure up to 7.7 GPa with respect to 2 GPa for a good compaction at room temperature. This conclusion is worth noticing because the applied pressures are typically below 1 GPa [7,8]. However, sample pressed at 300 K and 2 GPa could be successfully prepared for hardness measurements (it did not become ruined by polishing its surface) and, therefore, it was considered well compacted for further studies. Figure 7 shows densities and SEM micrographs of the bulk samples obtained by hot compaction of amorphous powder. In this case, the effect of pressing temperature and applied pressure could be studied. All the samples obtained by this procedure showed high density (~3.2 g cm-3) and a flat surface in which the individual particles of the former powder were indistinguishable, as it was observed in the sample compacted at room temperature and 7.7 GPa. The main difference observed was the resulting microstructure of the hot compacted bulk samples. As it was mentioned above, the increase of the applied pressure shifts the onset of crystallisation to higher temperatures: for the same temperature of pressing, 548 K, sample Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 9 pressed at 7.7 GPa presented a fully amorphous microstructure, whereas samples pressed at 2 GPa developed nanocrystalline microstructure with ~0.4 crystalline volume fraction and 15 nm of average grain size (see Fig. 5). On the other hand, samples pressed at 2 GPa and different temperatures (548 and 573 K, respectively) for 5 min, show no difference in the final microstructure achieved, both cases presenting the same values of Vcr and average grain size. It is worth noticing that, although nanocrystalline microstructure is expected to be denser than amorphous one, the bulk sample which results in amorphous state after hot compaction shows even a slightly higher density than those resulting in nanocrystalline state. Therefore, it might be inferred that the bulk amorphous sample obtained by hot compaction shows the best quality of compaction among the studied samples. Figure 8 shows densities and SEM micrographs of the bulk samples obtained by cold compaction of nanocrystalline powder. Both samples presented bad quality of compaction and it is possible to distinguish the individual particles of the nanocrystalline powder. Moreover, it was not possible to prepare the samples for hardness measurement, being ruined during surface polishing. However, SEM micrographs and density measurements clearly show that a better compaction, although not successful, could be obtained in the sample pressed at 7.7 GPa. This sample showed higher density in comparison with that exhibited by the sample pressed at 2 GPa and at room temperature from amorphous powder, which was considered to be a successful bulk sample. However, the higher density of the sample obtained from nanocrystalline powder pressed at 7.7 GPa could be explained by its different microstructure, because crystalline structure has a higher density than amorphous one. In conclusion, nanocrystalline powder, which is supposed to be harder than the amorphous one, fails to be compacted at room temperature in the region of pressures explored in this work, meanwhile the less hard amorphous powder could be successfully compacted even at 2 GPa and room temperature. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 16 different quality of compaction achieved for the different samples. In fact, it was clearly shown in this work that a change of density in amorphous samples strongly affects the microhardness and therefore special attention might be devoted to enhance the quality of compaction of the consolidated powder to optimise the hardness of the system. Recently, very high values of hardness (>900 HV=8.83 GPa) have been reported for nanocrystalline intermetallic (AlCu)3Zr and (AlMn)3Zr after consolidation in pore free bulk samples by spark plasma sintering [20] whereas less compact samples (77 % of density) of similar alloys exhibit only values <350 HV=3.43 GPa) [21]. A larger number of works have been devoted to ribbon samples than to bulk samples of ternary Al-RE(Y)-TM or quaternary Al-RE(Y)-TM1-TM2 alloys [10,11,19,22-27]. In our case, measurement of the hardness of as-cast amorphous ribbons leads to a value of 34025 HV (3.30.2 GPa), which is similar to the values obtained for the amorphous bulks with densities about 3.2 g cm-3 (see Fig. 9). Hardness values above 500 HV~5 GPa were reported for nanocrystalline ribbons with compositions Al-Mm-Ni [19], Al-Y-Ni [10] and AlGd-Ni [26] alloys. These facts allow us to consider the proposed production procedure promising and worth further developments. 5. Conclusions In this study, several ways of production of bulk nanocrystalline alloys by high pressure compaction were explored. The effect of different compaction parameters on the characterisation and properties of the final bulk samples were studied. Bulk nanocrystalline alloys were successfully produced with high hardness of above 500 HV~5 GPa. Several conclusions can be established:  Best compacted samples are those compacted at high temperature.  Nanocrystalline powder fails to be compacted at room temperature even at 7.7 GPa. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 17  Pressure (in the range from 2 to 7.7 GPa) clearly affects the quality of compaction and the microhardness of samples pressed at room temperature.  Crystallisation onset is shifted to higher temperatures as the pressure is increased from 2 to 7.7 GPa.  Microhardness increases as the density of the amorphous bulk increases.  Microhardness increases as the crystalline volume fraction of -Al phase increases, reaching a maximum at about 0.25-0.3 crystalline volume fraction.  Maximum microhardness was achieved in samples compacted at 548 K and 7.7 GPa (which results in an as-pressed amorphous bulk) after annealing for 20 min at 508 K (540 HV=5.3 GPa). This value is comparative to the maximum values reported for Albased alloys. References [1] A. Inoue, K. Ohtera, A. P. Tsai, T. Masumoto, J. Jpn. Appl. Phys. 27 (1998) L280. [2] Y. H. Kim, A. Inoue, T. Masumoto, Mater. Trans. JIM. 31 (1990) 747. [3] J. Latuch, H. Matyja, V. I. Fadeeva. Mat. Sci. Eng. A179-180 (1994) 506. [4] J. Q. Guo, K. Kita, N. S. Kazama, J. Nagahora, K. Ohtera, Mat. Sci. Eng. A203 (1995) 420. [5] K. I. Moon, K. S. Lee, J. All. Comp. 291 (1999) 312. [6] S. J. Hong, B. S. Chun. Mat. Sci. Eng. A348 (2003) 262. [7] Y. Kawamura, H. Mano, A. Inoue, Scripta Mater. 44 (2001) 1599. [8] I. Börner, J. Eckert, Scripta Mater. 45 (2001) 237. [9] A. R. Yavari, W. J. Botta-Filho, C. A. D. Rodrigues, C. Cardoso, R. Z Valiev, Scripta Mater. 46 (2002) 711. [10] Z. C. Zhong, X. Y. Jiang, A. L. Greer, Mat. Sci. Eng. A226-228 (1997) 531. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 18 [11] H. S. Kim, P. J. Warren, B. Cantor, H. R. Lee, Nanostr. Mater. 11 (1999) 241. [12] A. L. Greer, Mat. Sci. Eng. A304-306 (2001) 68. [13] H. S. Kim, Mat. Sci. Eng. A304-306 (2001) 327. [14] J. S. Blázquez, E. Fazakas, H. Dimitrov, J. Latuch, L. K. Varga, T. Kulik, J. NonCryst. Solids, 351 (2005)158. [15] H. Dimitrov, J. Latuch, T. Kulik, Solid State Phenom. 94 (2003) 71. [16] J. S. Blázquez, H. Dimitrov, J. Latuch, T. Kulik, Solid State Phenom. 101-102 (2005) 265. [17] K. Hono, Y. Zhang, A. P. Tsai, A. Inoue, T. Sakurai, Scripta Metall. Mater. 32 (1995) 191. [18] J. S. Blázquez, V. Franco, C. F. Conde, A. Conde, J. Magn. Magn. Mat. 254-255 (2003) 460. [19] W. S. Sun, M. X. Quan, Mater. Letters 27 (1996) 101. [20] S. H. Lee, K. I. Moon, K. S. Lee, Intermetallics 14 (2006) 1. [21] S. S. Nayak, S. k. Pabi, B. S. Murty, Intermetallics, in press, online 30 May 2006. [22] T. Gloriant, L. A. Greer, Nanostr. Mat. 10 (1998) 389. [23] E. S. Humphreys, P. J. Warren, J. M. Tichmarsh, A. Cerezo, Mat. Sci. Eng. A304-306 (2001) 844. [24] A. Inoue, H. Kimura, J. Light Metals 1 (2001) 31. [25] M. A. Muñoz-Morris, S. Suriñach, L. K. Varga, M. D. Baro, D. G. Morris, Scripta Mater. 47 (2002) 31. [26] B. C. Ko, P. Wesseling, O. L. Vatamanu, G. J. Shiflet, J. J. Lewandowski, Intermetallics 10 (2002) 1099. [27] M. A. Muñoz-Morris, S. Suriñach, M. Gich, M. D. Baro, D. G. Morris, Acta Mater. 51 (2003) 1067. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 19 Figure captions Figure 1: Scheme of the three different ways of production followed for obtaining bulk nanocrystalline alloys. Figure 2: (a) XRD pattern and (b) DSC scan, registered at 40 K/min, of amorphous ribbon of Al88Mm5Ni5Fe2 alloy. Figure 3: (a) XRD pattern and (b) DSC scan, registered at 40 K/min, of amorphous powder of Al88Mm5Ni5Fe2 alloy. Figure 4: XRD patterns, after substraction of the background (see experimental section), and fitting curves of Al88Mm5Ni5Fe2 alloy obtained by cold compaction at 7.7 GPa and different annealing times at 508 K. Figure 5: XRD patterns, before substraction of the background (see experimental section), of Al88Mm5Ni5Fe2 alloy obtained by hot compaction in their as-pressed state. Conditions of pressing are indicated for each sample. Figure 6: SEM micrographs of cold compacted Al88Mm5Ni5Fe2 alloy from amorphous powder. Density and conditions of pressing are indicated. Figure 7: SEM micrographs of hot compacted Al88Mm5Ni5Fe2 alloy from amorphous powder. Density and conditions of pressing are indicated. Figure 8: SEM micrographs of cold compacted Al88Mm5Ni5Fe2 alloy from nanocrystalline powder. Density and conditions of pressing are indicated. Figure 9: Vickers microhardness versus density for bulk amorphous samples. Figure 10: Theoretical relationship between the concentration of the different elements (Al, Mm, Ni and Fe) as well as the solute content (Mm+Ni+Fe) in the residual amorphous matrix and the crystalline volume fraction for the composition Al88Mm5Ni5Fe2. Pure Al was considered for -Al phase. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 20 Figure 11: Vickers microhardness versus (a) crystalline volume fraction and (b) solute content in the residual amorphous matrix. Solid lines in (a) represent the minimum limit predicted by the mixture model for each initially amorphous sample. Dashed lines are a guide to the eye. Figure 12: (a) Crystalline volume fraction versus the annealing time at 508 K and (b) Vickers microhardness versus annealing time at 508 K for bulk Al88Mm5Ni5Fe2 alloy. The samples were amorphous in as-pressed state. Lines are a guide to the eye. Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 21 Figure 1 MASTER ALLOY MELT SPINNING AMORPHOUS RIBBON MILLING AMORPHOUS POWDER COLD COMPACTION AMORPHOUS BULK BULK NANOCRYSTALLINE ANNEALING HOT COMPACTION AMORPHOUS BULK ANNEALING ANNEALING NANOCRYSTALLINE RIBBON MILLING NANOCRYSTALLINE POWDER COLD COMPACTION Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 22 Figure 2 25 30 35 40 45 50 55 400 500 600 700 heat flow temperature [K] 2 [degree] intensity Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 23 Figure 3 25 30 35 40 45 50 55 400 500 600 700 heat flow temperature [K] 2 [degree] intensity Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 24 Figure 4 30 35 40 45 5 min intensity 10 min 20 min 60 min 2 [degree] Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 25 Figure 5 30 40 50 60 70 80 548 K, 7.7 GPa intensity  - fcc Al     573 K, 2 GPa 548 K, 2 GPa 2 [degree] Intermetallics. Vol. 15. Núm. 7. 2007. Pag. 891-900 http://dx.doi.org/10.1016/j.intermet.2006.10.052 32 Figure 12 0 20 40 60 250 300 350 400 450 500 550 HV0.1 annealing time [minutes] 2 GPa, room temp. 7.7 GPa, room temp. 7.7 GPa, 548 K 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 V cr (a) (b)