SciELO - Scientific Electronic Library Online

 
vol.10 issue3A study of the consolidation method with albumin to obtain porous beta-TCP ceramicsThermodynamic simulation as an assistant tool in the development of iron based hardfacing materials author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Materials Research

Print version ISSN 1516-1439On-line version ISSN 1980-5373

Mat. Res. vol.10 no.3 São Carlos July/Sept. 2007

http://dx.doi.org/10.1590/S1516-14392007000300018 

A study of high-energy milling for the production of sintered PrFeB magnets

 

 

Elio Alberto PérigoI, *; Edson Pereira SoaresI; Ricardo Mendes Leal NetoI; Cláudio Costa MottaII; Rubens Nunes de Faria Jr.I

IInstituto de Pesquisas Energéticas e Nucleares – IPEN/CNEN, Av. Prof. Lineu Prestes, 2242, Cidade Universitária, 05508-000 São Paulo - SP, Brazil
IICentro Tecnológico da Marinha em São Paulo – CTMSP, Av. Prof. Lineu Prestes, 2468, Cidade Universitária, 05508-900 São Paulo - SP, Brazil

 

 


ABSTRACT

Sintered Pr16Fe76B8 magnets have been produced using the hydrogen decrepitation (HD) process and high-energy planetary ball milling. Investigations have been carried out to evaluate the influence of the milling speed and time. The best magnetic properties obtained were Br = (1020 ± 20) mT, µ0iHc = (1420 ± 30) mT and (BH)max/= (200 ± 4) kJm-3, for a magnet prepared with the alloy milled at 200 rpm for 4.5 ks. Magnets prepared from this powder exhibited a superior intrinsic coercivity compared to that of magnets produced using low-energy ball milling. However, the remanence and energy product of the latter were somewhat lower. An important feature was the dramatic reduction in the processing time (about 90%). Microstructural observation have shown that increasing the milling time and keeping constant the rotational mill speed caused an exponential grain size reduction in the sintered magnet. Increasing the milling speed also reduced the grain size and influenced both remanence and intrinsic coercivity.

Keywords: high-energy milling, hydrogen decrepitation, PrFeB


 

 

1. Introduction

The production of REFeB (RE = Pr, Nd) sintered magnets involves a multistep process1. It is possible to identify at least seven steps: melting of the alloy, homogenization, milling, powder magnetic alignment and compaction, sintering and post-sintering heat treatment. Good HD sintered REFeB magnets have been produced using the "roller" ball milling (RBM) from alloys comminuted between 64.8 and 72.0 ks (18 and 20 hours, respectively)2,3. However, milling is time consuming, even using the hydrogen decrepitation (HD) process. If production costs are considered, reducing the processing time is essential, without changing significantly the magnetic properties.

Few works report about high-energy milling technique to prepare sintered magnets4-7. However, none of them reports the use of planetary ball milling.

This paper addresses this aspect and reports work carried out to produce PrFeB sintered magnets from HD powders milled for a short time using high-energy milling.

 

2. Experimental

A commercial Pr16Fe76B8 (wt. (%): 34.24Pr-64.45Fe-1.31B) alloy in the as-cast state was used in this study. To produce the sintered Pr-based magnets using the HD process, (14.000 ± 0.002) g of the bulk ingot was placed in a stainless steel hydrogenation vessel which was evacuated to backing-pump pressure. Hydrogen was then introduced to a pressure of 2 bar which resulted in decrepitation of the bulk material. The decrepitated hydride material was transferred to a planetary ball mill Fritsch Pulverisette 5 and milled at several rotational speeds and for several times using cyclohexane as the milling medium. A preliminary evaluation showed that hydrogen is not an effective milling atmosphere due to the formation of agglomerates. Rotational speeds from 150 to 300/rpm with steps of 50 rpm and milling times from 1.8 to 5.4 ks with steps of 0.9 ks were used. Milling times longer than 5.4 ks were not practical due to difficulty to handle the extremely fine pyrophoric powder. On the other hand, it was verified that speeds below 150 rpm yield a coarse powder, useless to produce good sintered REFeB magnets. No contamination has been found in any sample from the milling spheres or jar. Based on the magnetic properties obtained from the hydrogen decrepitated Pr16Fe76B8 alloy milled for 1.8 and 2.7 ks, samples produced with longer times were investigated only for the speed of 200 rpm. Ball-to-powder weight ratio was kept constant for all experiments (10:1). The resultant pyrophoric powder was dried for 1.5 ks and transferred to a small cylindrical rubber tube under a nitrogen atmosphere. The resultant fine powder was then pulsed at a magnetic field of 6T, isostatically pressed at 200 MPa, vacuum sintered at 1333 K for 3.6 ks (60 minutes) and furnace cooled.

Magnetic characterization of the HD sintered Pr16Fe76B8 permanent magnets was carried out using a permeameter. Remanence (Br), intrinsic coercivity (µ0iHc) and maximum energy product ((BH)max) have been obtained from the second quadrant hysteresis curve. Microstructural observations were carried out using a scanning electron microscope. Grain size measurements were carried out using image analysis. Samples for grain analysis were etched with a solution of 25%H2O - 50%HCl - 25%HNO3 (%vol.) in order to reveal the grain boundaries.

 

3. Results and Discussion

The magnetic properties for the magnets produced from the alloy hydrogen decrepitated and milled for 1.8 ks (30 minutes) are shown in Figure 1. Remanence reached a peak of 860 mT for a milling speed of 200 rpm. This initial improvement in Br with the milling speed has been attributed to better alignment degree of the Pr2Fe14B (f) grains due to the reduction of the mean particle size. Rotational milling speed superior to 200 rpm diminished the remanence with a minimum at 250 rpm, possibly due to formation of agglomerates of the fine powder. A number of variables affect the magnet production and the variation on the magnetic properties at 250 rpm has been attributed to a processing drawback of this particular magnet. The magnet produced using a milling speed of 300 rpm showed the best µ0iHc. This has been attributed to the smaller grain size obtained at higher speeds, as shown in Figure 2. It has also been verified that low milling times led to heterogeneities of particle size (even at high speeds), as can be seen by the large standard deviations of the grain sizes for these times in Table 1. In this case, the magnet exhibits very small grains (3 µm) and also large grains (~45 µm). This effect is clearly seen in Figure 3a and most certainly influences the remanence since larger grains have a multi-domain structure.

 

 

 

 

 


 

Figure 4 shows the variation of the magnetic properties with milling speed for magnets produced using a constant milling time of 2.7/ks (45 minutes). Again, the better remanence and energy product (see Table 2) were obtained for the sintered magnet milled at 200/rpm. The best µ0iHc was obtained for the magnet prepared from the alloy milled at 250 rpm, although the intrinsic coercivity for the alloy milled at 300 rpm is quite similar. Figures 2 and 3b show a reduction of the mean gain size (about 15%) compared to the previous case, but the heterogeneity of the microstructure (Figure 3b) is very evident, as in the previous case (Figure 3a).

 

 

The magnetic properties for the Pr16Fe76B8 magnet prepared from the decrepitated alloy milled for 3.6 ks (60 minutes) are presented in Figure 5. There was a steady improvement in remanence up to 4.5 ks. The intrinsic coercivity showed a gradual improvement followed by a slight decrease. It was verified that the mean grain size was also reduced with longer milling time, and that the homogeneity was improved. This is clearly seen in the microstructure shown in Figure 6a. Furthermore, it was verified that the reduction grain size profile has a particular feature (an exponential decay), as presented in Figure 7.

 

 

 



 

 

 

The best magnetic properties and microstructural homogeneity were obtained for the alloy decrepitated and subsequently milled for 4.5 ks (75 minutes). The remanence reached 1020 mT. This value is somewhat below to that (1200 mT) reported for a Pr16Fe76B8 magnet prepared using the "roller" ball mill8. On the other hand, m0iHc reached 1420 mT, 10% superior than the value reported for magnets produced using RBM8. The microstructure of this sintered magnet is presented in Figure 6b.

The last condition studied concerns magnets obtained from the decrepitated alloy milled for 5.4 ks (90 minutes). As can be seen in Figure 5, there was a reduction in Br after 4.5 ks, probably due to the reduction of the alignment degree of the Pr2Fe14B phase, which will be checked using X ray pole figure analyses. The microstructure of the magnet processed with the powder milled for 5.4 ks is shown in Figure 6c. There is a similarity of the grain structure of this sample compared to the previous case (4.5 ks milling), so that no significant change in the intrinsic coercivity would be expected, as verified in Figure 5.

It is worthwhile noting that increasing the milling speed, with constant milling time, results in the increase of the density of the magnets, as can be seen in Table 1, probably due to a better sinterability. A summary of the HD sintered Pr16Fe76B8 microstructural parameters and densities are presented in Table 1, and of the magnetic properties are presented in Table 2.

 

4. Conclusions

The present studies show clearly that, employing the appropriate parameters for high-energy milling, it is possible to produce good permanent sintered magnets. Satisfactory overall magnetic properties were obtained for the magnet prepared with the alloy hydrogen decrepitated and milled at 200 rpm for 4.5 ks (75 minutes). High-energy milling had the effect of improving the intrinsic coercivity of the PrFeB HD sintered magnet, but at the expenses of the remanence and energy product. The processing time of the magnets was dramatic reduced (around 90%) compared with the conventional "roller" ball milling.

 

Acknowledgments

The authors thank IPEN-CNEN and FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) for supporting this investigation. E. A. Périgo also thanks FAPESP for the provision of a scholarship (contract number 2005/04711-2).

 

References

1. Liu, S. Magnetic alignment in powder magnet processing. Journal of Applied Physics. 1994; 76(10):6757.        [ Links ]

2. Takiishi H, Lima LFCP, Costa I, Faria RN. The influence of process parameters and alloy structure on the magnetic properties of NdDyFeBNb HD sintered magnets. Journal of Materials Processing Technology. 2004; 152(1):1.        [ Links ]

3. Soares EP, Périgo EA, Takiishi H, Motta CC, Faria RN. A study of Pr-Fe-B magnets produced by a low-cost powder method and the hydrogen decrepitation process. Materials Research. 2005; 8(2):143-145.        [ Links ]

4. Kianvash A, Harris IR. The production of a Nd16Fe76B8 sintered magnet by the hydrogen decrepitation/ hydrogen vibration milling route. Journal of Alloys and Compounds. 1999; 282(1-2):213.        [ Links ]

5. Mottram RS, Williams AJ, Harris IR. The effects of blending additions of copper and cobalt to Nd16Fe76B8 milled powder to produce sintered magnets. Journal of Magnetism and Magnetic Materials. 2001; 234(1):80.        [ Links ]

6. Christodoulou CN, Schlup J, Hadjipanayis GC. Oxidation of Fe-R-B powders during preparation of permanent magnets. Journal of Applied Physics. 1987; 61(8):3760.        [ Links ]

7. Mcguiness PJ, Harris IR, Rozendaal E, Ormerod J, Ward, M. The production of a Nd-Fe-B permanent magnet by a hydrogen decrepitation/attritor milling route. Journal of Materials Science, 1986; 21(11):4107.        [ Links ]

8. Faria RN. The influence of zirconium addition and process parameters on the magnetic properties of Pr-Fe-B sintered magnets. Journal of Magnetism and Magnetic Materials. 2002; 238(1):56.        [ Links ]

 

 

Received: March 21, 2007; Revised: August 28, 2007

 

 

* e-mail: eaperigo@yahoo.com

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License