You are on page 1of 3

JOURNAL OF APPLIED PHYSICS 97, 1 2005

Microstructure and magnetic properties of sintered NdFeB magnets with improved impact toughness
J. F. Liu,a P. Vora, M. H. Walmer, and E. Kottcamp
Electron Energy Corporation, 924 Links Avenue, Landisville, Pennsylvania 17538

S. A. Bauser, A. Higgins, and S. Liu


University of Dayton, 300 College Park, Dayton, Ohio 45469

Presented on 9 November 2004 An effort to increase the impact toughness of NdFeB sintered magnets by adding small amounts of Al, Nd, Ga, Cu, and Nb was successful. No signicant compromise to magnetic properties occurred. Based on this work, a series of sintered NdFeB magnets with improved toughness was developed, which we call ToughNEO. Small precipitates, which may contribute to the improvement of toughness, were observed using scanning electron microscope for all samples with improved toughness. Tumbling and drilling tests further veried the improved toughness of these developed ToughNEO magnets. 2005 American Institute of Physics. DOI: 10.1063/1.1847215
I. INTRODUCTION

Sintered NdFeB-type permanent magnets are based on Nd2Fe14B intermetallic compounds, which are intrinsically very brittle. Generally, machining techniques are limited to grinding, lapping, and electric discharge machining EDM . Even grinding and lapping these magnets to the required dimension often result in chipping, cracking, and fracture, which leads to production losses ranging from 10%30%. Chipping, cracking, or fracture can also occur during assembly and operation. Past efforts to increase the toughness of sintered NdFeB magnets by adding ne bers, such as Al2O3 or SiC, failed because both Al2O3 and SiC are very brittle although they have high strength.1,2 Most of the research effort on NdFeB-type magnets was focused on the magnetic properties and microstructure analysis in the past, although some work was done to characterize the mechanical properties of the commercial magnets. Withey et al.3 studied the effect of quench rate on the brittleness of sintered NdFeB-type magnets and concluded that fast quenching from 1097 C by immersion in liquid nitrogen or oil causes signicant deterioration in both magnetic and mechanical properties. Horton et al.4 found large variations in the fracture toughness KIC of commercial NdFeB-type magnets from various commercial sources.

hysteresigraph. Impact toughness was measured using the Charpy impact method with a Bell-Laboratory-type impacttesting machine. ASTM C1161, ASTM E8-00b, and ASTM C1424 were used as the testing standards for the exural, tensile, and compressive strengths, respectively.

III. RESULTS AND DISCUSSION A. Magnetic properties

Based on the results reported earlier about the effect of individual elements on the mechanical and magnetic properties,1,2 magnet samples with a nominal composition of Nd, Pr 16.5Dy0.5Fe76 Nb, Ga, Al, Cu, Mn, Ti 1B6, which will be called ToughNEO magnets in this paper, were chosen to be compared to commercial NdFeB-type magnets. The typical magnetic properties of ToughNEO magnets are Br = 1.23 1.35 T, Hci = 1040 1520 kA/ m, and BH max = 288 336 kJ/ m3. A typical demagnetization curve of a

II. EXPERIMENT

Alloys of RExFe94xy M yB6 were prepared by vacuum induction melting, where M is one or more element s selected from the group consisting of Al, Nb, Zr, Cu, Ga, Co, V, and Mn, and RE is one or more rare-earth element s selected from the group consisting of Nd, Dy, and Pr. Sintered magnets were made using a conventional powder metallurgy process. Magnetic properties were tested using a KJS
a

Electronic mail: j@electronenergy.com 97, 1-1

FIG. 1. Typical demagnetization curve of ToughNEO magnets. 2005 American Institute of Physics

0021-8979/2005/97 10 /1/0/$22.50

1-2

Liu et al. TABLE I. Average mechanical properties of ToughNEO and commercial NdFeB magnets. ToughNEO magnets 46.2 kJ/ m2 84.4 347.5 MPa 1051.5 MPa 161.0 MPa Commercial magnets 27.3 kJ/ m2 87.5 299.2 MPa 1139.1 MPa 163.4 MPa

J. Appl. Phys. 97, 1 2005

Mechanical properties Impact toughness HR15N Flexural strength Compressive strength Tensile strength

Diff. 69% 4% 16% 8% 1%

Standard

Number of parts tested 5 5 parts at 5 locations each 5 6 5

ASTM E18-00 ASTM C1161 ASTM C1424 ASTM E8-00b

ToughNEO magnet with Br = 1.27 T, Hci = 1240 kA/ m, Hc = 944 kA/ m, and BH max = 304 kJ/ m3, is shown in Fig. 1. A commercial N45M sintered magnet was used for the comparison of mechanical properties. This N45M magnet has the following nominal magnetic properties: Br = 1.35 T, Hci = 1120 kA/ m, Hc = 968 kA/ m, and BH max = 360 kJ/ m3.

B. Mechanical properties

The average mechanical properties of both ToughNEO and commercial magnets were tested and listed in Table I. As can be seen, the impact toughness is 46.2 kJ/ m2 for the ToughNEO magnets, but only 27.3 kJ/ m2 for the commercial N45M sintered NdFeB magnets. This represents a 69% improvement on average in impact toughness. The Rockwell hardness of the ToughNEO magnets is lower compared to commercial NdFeB magnets, as shown in Table I. This is expected because, typically for metallic materials, hardness decreases as toughness increases. The exural strength of the ToughNEO magnets is 16% higher than the commercial NdFeB magnets, which is also consistent with the Charpy impact test and the hardness measurements. For the ToughNEO material, the compressive strength is slightly lower while the tensile strength is almost identical compared to the commercial NdFeB magnets.

C. Tumbling and drilling tests

length of approximately 305 mm. About 900 g of steel balls with various diameters were loaded in the tumbler together with about 20 magnet samples. After tumbling for 5 min at 590 rpm, the magnet samples were taken out for comparison. As seen in Fig. 2, ToughNEO magnets had no visible chips, while the commercial NdFeB magnets have large chips on the edges and corners. These results substantiated the improved toughness measured by impact testing. The dependence of chip size on the tumbling time is shown in Fig. 3. For ToughNEO magnets, the chip size does not change much with tumbling time, while for commercial magnets the chip size increases signicantly with an increase in tumbling time. This is another indication that ToughNEO magnets are not as brittle as commercial NdFeB magnets. The weight loss after tumbling is shown in Fig. 4. ToughNEO magnets have less weight loss compared to commercial Nd FeB magnets after tumbling, which is consistent with the chip size measurement shown in Fig. 3. A drilling experiment was also conducted on both ToughNEO and commercial NdFeB magnets using a 1/8-in. drill. The rotational speed of the drill was xed at 1000 rpm. The drill feed rate was 0.135 mm per revolution. The magnet samples were put on a metal surface before drilling. Sintered ToughNEO magnets had no chips on edges after drilling, while the commercial NdFeB magnets broke into pieces every time, and were essentially undrillable. This is clearly shown in Fig. 5.

In order to further compare ToughNEO magnets and commercial NdFeB magnets, 120 rectangular samples of each were made and subjected to tumbling and drilling tests. All of the 120 samples have the same dimensions 5.33 6.86 7.87 mm3 . Tumbling tests were conducted using a steel cylinder with an inner diameter of approximately 152 mm and a

FIG. 2. Surface conditions of a ToughNEO and b commercial NdFeBtype magnets after tumbling at 590 rpm for 5 min.

FIG. 3. Chip size vs tumbling time for ToughNEO and commercial NdFeB-type magnets.

1-3

Liu et al.

J. Appl. Phys. 97, 1 2005

FIG. 4. Weight loss vs tumbling time for ToughNEO and commercial NdFeB-type magnets.

D. Microstructure

The SEM images of the fractured surface of both ToughNEO and commercial NdFeB magnets are shown in Fig. 6. Commercial magnets are free of second-phase precipitates, while ToughNEO magnets have many small precipitates. According to previous reports on the microstructure of NdFeB magnets,5 the addition of low melting point elements such as Al and Ga forms intergranular phases, but the addition of refractory metals such as Nb, Ti, Mo, and V would form precipitates within the grains. The EDS spectrum shows that these precipitates are rich in neodymium, niobium, gallium, aluminum, and other added elements. The neodymium content of these precipitates is always signicantly higher than that of the matrix although the contents of other added elements vary from area to area in the samples. These precipitates are very likely softer than the Nd2Fe14B matrix phase, which may contribute to the improvement of toughness. Preliminary analysis using optical microscope suggests that the fracture of both commercial and ToughNEO magnets is largely intergranular. Further analysis on a statistical basis is in progress. The data will be reported elsewhere.

FIG. 6. SEM micrograph showing the fractured surface of a ToughNEO and b commercial NdFeB-type magnets.

NdFeB-type magnets, while the magnetic properties are comparable. Tumbling and drilling tests intended to be more representative of manufacturing and service environments conrmed the superior toughness of these ToughNEO materials. The signicant improvement in the toughness of the ToughNEO magnets may be related to the small precipitates. This new material can provide more design exibilities in critical devices, especially high g-force military applications.

IV. SUMMARY

ACKNOWLEDGMENT

The impact toughness of ToughNEO magnets was found to be signicantly better than that of the commercial

The nancial support for this study by the Ofce of Naval Research ONR is gratefully acknowledged.
1

FIG. 5. Comparison between a ToughNEO and b commercial Nd FeB magnets after the drilling test.

J. F. Liu, K. McGinnis, E. Kottcomp, M. H. Walmer, S. Bauser, A. Higgins, and S. Liu, 18th International Workshop on High Performance Magnets and Their Applications, Annecy, France, 29 August-2 September 2004 unpublished . 2 S. Liu, D. Cao, R. Leese, S. Bauser, G. E. Kuhl, J. F. Liu, M. H. Walmer, and E. Kottcamp, Proceedings of the 17th International Workshop on RE Magnets and Their Applications, Newark, Delaware, USA, 1822 August 2002, edited by G. C. Hadjipanayis and M. J. Bonder unpublished , p. 360371. 3 P. A. Withey, H. M. Kennett, P. Bowen, and I. R. Harris, IEEE Trans. Magn. 26, 2619 1990 . 4 J. A. Horton, J. L. Wright, and J. W. Herchenroeder, IEEE Trans. Magn. 32, 4374 1996 . 5 C. H. de Groot, K. H. J. Buschow, F. R. de Boer, and K. de Kort, J. Appl. Phys. 83, 388 1998 .

You might also like