Introduction
Rare-earth permanent magnets based on NdFeB or SmCo have applications in efficient motors and generators with a high energy density. While NdFeB-based magnets are cheaper to manufacture than SmCo-based magnets and have superior magnetic properties at room temperature, they suffer from a loss of magnetic properties at elevated temperatures.
The highest temperatures at which NdFeB magnets can currently operate usefully is ~200ºC. By contrast, recently-developed SmCo-based magnetic alloys can retain useful magnetic behaviour to temperatures of ~550ºC, making them useful for aero-engine applications such as frictionless bearings or co-axial starter-motors/generators.
However, limited previous work has indicated that at these high temperatures oxidation may become a problem. The aims of the present research were to undertake an extensive study of the oxidation properties of current SmCo alloys and to explore coating methods to improve their resistance to oxidation. The project was undertaken in collaboration with Rolls-Royce plc, Indestructible Paints Ltd., Precision Magnetics Ltd. and Less Common Metals Ltd.
Project results
The mechanism by which SmCo-based magnets oxidise has been identified. Several processes have been shown to occur but the most significant effect is the diffusion of oxygen into the magnet producing an internally oxidised zone (IOZ) as a layer growing from the surface into the magnet.
Within this IOZ the Sm is preferentially oxidised and the SmCo-based phases, that produce the good permanent magnetic properties, are consumed to give a mixture of Sm oxide in a matrix of CoFe phase. This transformation destroys the permanent magnetic properties of the magnet within the IOZ, hence progressively eating away at the magnet with increasing time at temperature.
An extensive analysis of the growth rate of the IOZ over the temperature range 300-600 degree C has provided the necessary information to allow the lifetime of a SmCo-based magnet to be predicted when operating in air over that temperature range (previous studies had shown insignificant oxidation below 300 C.)
This analysis was carried out by exposure of magnet samples to air at elevated temperature for specific times. The magnets were then sectioning and examined by SEM to measure the thickness of the IOZ. The SEM also revealed microstructural and chemical changes that had not previously been observed.
A range of coatings have been produced and assessed for their efficiency at protecting the magnets from oxidation. As the coating could only protect the magnet by providing a barrier to prevent the oxygen getting to the magnet then it is essential to produce a well-adhered continuous coating.
The coatings that were chosen were sputtered silica, sputtered platinum and a range of proprietary coatings supplied by Indestructible Paints Ltd., one of the project collaborators.
All of the coatings showed some degree of protection at the test temperatures of 450 and 550ºC. At 450ºC the Pt coating and the Industrial Paints' IP-9184-R1 coating show 'complete' protection with no measurable internally oxidised zone at times up to 650 hours, where in the uncoated samples the IOZ had reached ~100 microns thick. At 550ºC after 350 hours, when uncoated samples show an IOZ of ~250 microns, none of the coatings showed complete protection.
However, the IP-9184-R1 coated sample showed an IOZ of only ~20 microns after 200 hours at 550ºC, which did not grow measurably further up to 350 hours, suggesting that this could be a suitable coating for use at these temperatures.
The project was funded by EPSRC, with the support of:
-
Less Common Metals Ltd
-
Indestructible Paints Ltd
-
Precision Magnetics Ltd
-
Rolls-Royce PLC
Detailed Background
Permanent magnets find applications primarily in devices that convert mechanical energy into electrical energy or vice-versa, i.e. generators/sensors and motors/actuators. There are various types of permanent magnet materials each having their particular advantages and disadvantages. The most widely used material is ferrite, which is a hexagonal mixed oxide phase based on either Ba or Sr with Fe oxide. This material is low cost, but has fairly poor magnetic properties compared to the more modern rare-earth transition metal magnets, based on NdFeB or SmCo.
The other significant class of permanent magnet material is Alnico, which is Fe-based containing additions of Al, Ni and Co. Although the Alnico magnets have a high Curie temperature (above which magnetism is lost) they have a very low coercivity (resistance to demagnetisation) and their use is limited to applications where demagnetising fields are low, but high field and stability are required, e.g. speedometers.
Both NdFeB- and SmCo-based permanent magnets find applications in efficient motors and generators with high 'energy density', i.e. power to weight ratio. NdFeB-based magnets are cheaper and have superior magnetic properties at room temperature but experience a significant loss of magnetic properties at elevated temperature. Some grades of NdFeB magnets with additions of Dy and Co (which reduce the remanence and energy product) can operate at temperatures up to ~200ºC.
Therefore, SmCo-based magnets, despite their higher cost, are utilised in applications where the devices are likely to experience higher temperatures and / or require consistent magnetic properties (hence device performance) over a range of operating temperatures. Beginning in the late 1990s, there has been much research activity in the USA (primarily at the University of Delaware and the Electron Energy Corporation) and Europe (EU framework 5 programme: Hitemag) to develop grades of SmCo-based magnets that will operate at temperatures up to 550ºC, exceeding the conventional limit of 300ºC.
Permanent magnets that operate at these temperatures would find applications, for example, in aero-engines for frictionless bearings and generators that can be integrated into the main axis of the turbine. The advantages of these modifications would be:
-
simplified engine architectures - eliminates take-off shafts;
-
reduces engine net size with respect to aerodynamic drag;
-
allows energy transfer between spools;
-
improved efficiency;
-
reduced maintenance.
Figure 1 shows the maximum energy product, (BH)max, against temperature. The (BH)max is a figure of merit for permanent magnets and relates to the amount of work that can be done per unit volume of material. Each line represents a particular grade of magnet; shades of blue for NdFeB-type magnets, red/orange for conventional SmCo grades and greens for the new grades of SmCo of the type developed by the Hitemag programme.
There are two types of SmCo-based magnets: the 1:5 and 2:17 type, which are based on the SmCo5 and Sm2Co17 structures respectively. Of these types it is the 2:17-type that has been developed for use at elevated temperatures. The 2:17-type SmCo-based permanent magnets have the general composition Sm(Co,Fe,Cu,Zr)Z (where Z ~8.5) with Fe, Cu and Zr being minor additions relative to Co.
Permanent magnetic properties are achieved by careful control of the microstructure. The optimum magnetic properties can only be achieved in a particular crystal direction, the c-axis of the hexagonal crystal structure. Therefore, the cast ingot, which has an almost random orientation of grains, is broken down to a powder, which consists of single crystal particles that can be aligned in a magnetic field during pressing.
These powder compacts are then sintered into magnets and solution treated at ~1100ºC, where they are single phase. This homogenising stage is followed by several aging treatments at lower temperature where a cellular structure is formed. The cells are ~50-150 nm in diameter and based on the Sm2Co17 type phase, which is enriched in Fe.
The cell boundaries comprise of a layer, about 5-20 nm thick, of SmCo5 type phase, which is enriched in Cu. The Zr forms a Zr-rich lamella phase that cuts across the cellular structure and lies in the basal plane of the Sm2Co17 and SmCo5 type phases. It is believed that the Zr-rich phase acts as a rapid diffusion pathway and increases the kinetics of the formation of the cellular structure.
In order to achieve higher operating temperatures, the composition of the magnets has been modified by varying the concentrations of Co, Cu, Fe and Zr, as well as changing the Z-value (the ratio of rare earth to transition metal elements). These compositional changes create changes in the intrinsic properties of the phases.
The optimum heat treatments have also been modified and in conjunction with the compositional changes, this leads to a change in the microstructure.
The development of these compositions and the optimisation of the heat treatment and microstructure formed part of the work carried out in the Hitemag project. Permanent magnets have now been reported that have a maximum energy product of ~80 kJm-3 at 500ºC [1]. However, these magnets have been shown to suffer from magnetic losses due to oxidation [2], for example, ~10% after exposure to air at 500ºC for 360 hours and ~16% at 550oC after the same time.
No significant losses were observed at 300ºC, the previous maximum operating temperature of SmCo-based magnets, and this explains why there have been few studies of the oxidation of these materials. No explanation of the oxidation mechanism was provided but it was clear that the extent of attack was substantial with a sub-surface oxidation zone of 270 mm having formed after 360 hours exposure at 550ºC.
The current project at Birmingham will determine the kinetics and mechanism of this oxidation and investigate a method of protecting the magnets suitable for use in a gas turbine engine.[1] M.S. Walmer, C.H. Chen, M.H. Walmer, S. Liu and G.E. Kuhl, Proc. of 16th Int. Workshop on RE Magnets & their Applications, Sendai, 2000, p.41.
maxwithtemperatureforarangeofmagnetgrades.gif)
Figure 1: Variation of (BH)max with temperature for a range of magnet grades.
[2] C.H. Chen, M.S. Walmer, M.H. Walmer, S. Liu and G.E. Kuhl, IEEE Trans. Magn. 36 2000, 3291.