PERMANENT MAGNET MATERIALS & LINEAR INDUCTION MOTOR

PERMANENT MAGNET MATERIALS & LINEAR INDUCTION MOTOR

The market for hard magnetic materials is nowadays dominated by two families of materials, the hexagonal ferrites and Nd–Fe–B (Fig. 13.1(a)). Their production volumes and costs are quite different, but each holds roughly half the market.

Ferrites are produced throughout the world, but Nd–Fe–B production is concentrated in China, where there are abundant rare-earth reserves. Most magnets are produced as sintered blocks or other simple shapes, but increasing quantities of polymer-bonded material, suitable for making complex shapes by injection moulding, are manufactured. The maximum energy product doubled roughly every 12 years over the course of the twentieth century (Fig. 1.13).

Consequences for permanent magnet device design have been dramatic, as illustrated in Fig. 13.4. The devices shrink as higher energy-product magnets become available, their configuration changes and the number of parts can be reduced. The advantage of magnets over coils in small structures can be appreciated by comparing a small discshaped magnet with a coil having the same magnetic moment. A magnetized cylinder with diameter 8 mm and height 2 mm made of a material with M = 1 MA m−1 has m ≈ 0.1Am2. The equivalent current loop, m = IA would require 2000 ampere-turns, an impossible demand in such a small space! There has been no further doubling of (BH)max since 1996, and there seems to be little scope for further dramatic improvements of bulk material, barring the development of practical oriented hard/soft nanocomposites. Properties of some typical magnets are shown in Table 13.3.

The data are for dense oriented magnets, made from sintered uniaxial ceramic ferrite or metallic alloy powder. These magnets have a microstructure in which the c axes of the individual crystallites are aligned by applying a magnetic field during processing. All except alnico are true permanent magnets, in the sense that the magnetic hardness parameter κ, defined by (8.1), is greater than unity. The (BH)max values achieved fall short of the theoretical maximum µ0M2 r /4 because of nonideal loop shape. Two values of coercivity are listed, one is the ‘intrinsic’ coercivity iHc measured on the M(H) loop, the other, BHc, is the coercivity measured on the B(H) loop.

Ferromagnetic microstructures are illustrated in Fig. 13.5. Sometimes the additional process steps needed to produce an oriented magnet cannot be justified economically. It may be cheaper to make more of an isotropic magnet with inferior magnetic properties. The c axes of the individual crystallites are random and, if interactions between crystallites are neglected, the remanence is Ms cos θ = 1 2Ms, where θ is the angle between the c axis and the direction of magnetization. Components made of ceramic or sintered metal are shaped by slicing or grinding. Polymer bonding allows for more versatility in shaping magnetic components. A coercive powder is mixed with a binder, with a fill factor fm of 60–80 vol%. The mixture is die-pressed, injection moulded, extruded or rolled into the required shape. The c axes of the crystallites in the powder may be oriented in the binder by using a magnetic field in order to augment the remanence. Since Mr is now fmMs cos θ , where θ is the angle between the c axis and the aligning field, even a fully aligned powder with fm = 0.7 has an energy product that is less than half of the bulk value. Table 13.4 illustrates the properties of differently processed magnets based on SrFe12O19. The magnets in electrical machines can be subject to temperatures in excess of 100 ◦C during routine operation. Magnetization and coercivity naturally decline as the Curie point is approached.

Temperature coefficients around ambient temperature are listed in Table 13.5. Not all the loss is recoverable on returning to ambient temperature; irreversible losses are associated with thermal cycling. Maximum service temperatures for different materials are included in the table. Ferrites and rare-earth magnets with wide, square hysteresis loops have the property that the field of one magnet does not significantly perturb the magnetization of a neighbouring magnet.

This is because the longitudinal susceptibility is zero for a square hysteresis loop, and the transverse susceptibility is Ms/Ha, which is only of order 0.1, since the anisotropy field Ha = 2K1/µ0Ms is much greater than the magnetization. Rigidity of the magnetization means that the superposition of the induction of rare-earth permanent magnets is linear and the magnetic material is effectively transparent, behaving like vacuum with permeability µ0.

Transparency and rigidity of the magnetization greatly simplify the design of magnetic circuits. There is no need to worry about the effect of one magnetic segment on another. Magnet cost is often critical for applications. Ferrite and Nd–Fe–B magnets are both produced in large quantities – roughly 1 million tonnes of ferrite and 50 000 tonnes of Nd–Fe–B in 2008. The properties of the two are quite different, and numerous grades of Nd–Fe–B are available, featuring high coercivity or high remanence, with energy products ranging from 250 to 450 kJ m−3. A very rough guide to cost is $1 per joule of stored energy. Hence, from the data in Tables 13.3 and 13.4, the costs of oriented ferrite and Nd–Fe–B work out at about 7$ kg−1 and 40$ kg−1, respectively. Ferrite is even cheaper than this but the cost of sintered Nd-Fe-B is about twice that estimate.