An amorphous core transformer: design and experimental performance

MATERIALS SCIEWCE & ENGINEERIRG Materials Science and Engineering A226228

(1997) 1079-1082

An amorphous core transformer: design and experimental performance Benedito Antonio a Department b Materials

of Electrical Engineering

Engineerillg, Departnwnt,

Luciano a, Claudio Shyinti Kiminami

Federal University Federal Univexity

b

of Paraiba, PO Box 10105, 58109 970 Campina Grande, Brazil of SZo Carlos, PO Box 676, 1356.5 950 Sn”o Carlos, Brazil

Abstract The main aspectsrelated to the designand experimentalimplementation of a 1 kVA, 220 V/l10 V, 60 Hz amorphousmetal core transformer (prototype) are presented.The core material is the amorphousalloy Fe,,B,,Si,. This as-castcore wasannealed at 380°C for I h in an argon atmosphere,under a DC magnetic field applied along the ribbon length to induce an uniaxial anisotropy. After having been annealed,samplesof the core material were examinedby differential scanningcalorimetry (DSC) and by X-ray diffraction and no crystallization was observed.Experimental tests are presentedto comparethe amorphouscore transformer performance with that of a commercialtransformer with the same ratings values, but built with a steel core. In comparativeterms, the amorphouscore transformerpresentedlower active lossesand lower excitation power than the commercial steel core transformer. 0 1997Elsevier ScienceS.A. Keywords;

Amorphousmetal;Coretransformer;Annealing

1. Introduction The core of the first practical transformer, built by William Stanley in 1885, was made of carbon steel sheets [l]. Later, these materials were substituted by silicon steels and today most distribution and power transformer cores in service are of grain oriented silicon steel laminations; but this trend has been changed since the introduction of the amorphous alloys in the transformers of the U.S.A, Japan and European power distribution systems [2]. Amorphous metals are metallic materials with a noncrystalline liquid-like molecular structure. These metals are formed in long and thin strips by rapidly cooling molten metal to prevent crystallization of the material during solidification. Typical amorphous alloy compositions used for transformer cores contain about 80 at.% of iron and about 20 at.% of a mixture of metalloids such as boron and silicon [3]. However, a wider actual use of amorphous alloys in power transformers requires the solution of some seripus manufacturing problems, which arise because of the small ribbon thickness and the fact that it is very hard. Despite these adverse factors, since 1980, designers in companies have been working on new methods of 0921-.5093/97/%17.00 0 1997 Elsevier Science S.A. All rights reserved. DTr cl-lo,,

and secondary (Y,) voltages: s, = s/w,

(1)

s, = S/JVs

(2)

The number of winding turns (N) is obtained by the equation N = V/4.44fB,Sm,,

(3)

where Y is the rated rms voltage, B, is the peak flux density, Smag is the magnetic core cross sectional central lag, and 4.44 is the form factor. Because of the m.m.f balance (N,I, = NZIZ) the primary and secondary windings each occupy about onehalf of the available area so that the primary current linkage is NrIr = i&A, J and the rms value of the current is given by: Lx = KAv JW

(4)

By Faraday’s law an electromotive induced in each turn of the coil.

force (en@ will be

Vnn, = 4.44fls,,,B,

(5)

Therefore, the transformer

volt-ampere

rating is:

S = 2.22fB, Jk’,A,S,,,

(6)

Writing the total H,I,, where I, is around the magnetic the N1 turns of the

m.m.f for a peak flux density as the mean length of the flux path circuit, the magnetizing current in primary windings is

ignoring harmonics. Then, using Eq. (5) for the primary voltage, the magnetizing reactive power (in var) is given by [Xl:

Q, = VJo, = 4.44f~ms,,,Hm~ml~

(7)

B.A.

Luciano,

C.S. Kimbzami

/Materials

Science

and Engineering

A226-228

(1997)

1079-1082

1081

As SmasZmis the core volume and the density of the amorphous alloy is 7.18 g cmm2, the specific magnetizing reactive power (in var kg-‘) is q0 =

4.36fBmHm x 1o-4

(8)

with the total core loss P,, the active component of the no-load current 1, is loa = PO/ VI, and the total nodoad current is the phasor sum of the active and reactive current components.

3. The annealing treatment Before annealing, the as-cast structure of the amorphous material was examined by differential scanning calorimetry (DSC) and by X-ray diffraction and no crystallization was observed. The core material is the amorphous alloy Fe,,B,,Si,. This as-cast core was annealed at 380°C for 1 h in an argon atmosphere, under a 800 A m-l DC magnetic field applied along the ribbon length for inducing an uniaxial anisotropy. This magnetic field intensity was chosen to ensure core magnetic saturation during the annealing process. After having been annealed, samples of the core material were examined by differential scanning calorimetry (DSC) and by X-ray diffraction and no crystallization was observed. The purpose of the annealing treatment is to minimize the level of stress which is caused by the ribbon casting and the core winding process. The combination of temperature and time is chosen such that is sufficient to relieve as much stress as possible without causing crystallization of the amorphous alloy, During the annealing the temperature must be kept constant. The magnetizing field is used primarily for inducing a uniaxial anisotropy along the ribbon direction [9]. Fig. 2 shows the experimental arrangements used for annealing, and in Fig. 3 its total anneal cycle is illustrated. Fig. 4 shows the DSC curves obtained for as-cast and annealed sample of Fe,,B,,Si,.

Fig. 2. Schematic representation of the experimental arrangement for annealing.

Fig. 3. The anneal cycle for amorphous alloy core.

Fig. 5 shows the specific heat difference (AC,) vs temperature for the as-cast and annealed samples determined by DSC. The curves show that the as-cast sample presents a maximum point about 480 K and from it the specific heat (C,) decreases due to the structural relaxation. This behavior is not verified at curve 2 (annealed sample) which confirms the structural relaxation. Welldefined peaks are observed in curves 1 and 2 at about 650 K which means the Curie temperature T,.

4. Performance transformers

of amorphous metal and silicon steel

Experimental tests are presented to compare the amorphous core transformer performance with that of

Fig. 4. DSC for as-cast and annealed sample.

1082

B.A. Luciano,

C.S. Kiminami

/Materials

Science

' 300

400

600 TEMPERATURE~K) son

700

11226-228

- 0.4 0.00

I 0.1

nnd Engineering

(1997)

l”“l”“l”“1”“~ 0 01

1079-1082

0.02

0 03

0.04

800 % 100

Fig, 5. The specific heat difference (AC,) versus temperature: 1, as-cast; 2, annealed.

a commercial transformer with the same rating values, but built with a steel core. The performance of an amorphous metal core transformer (prototype) is compared with a typical silicon iron transformer in Table 2. with a steel core. To compute the conventional efficiency (the ratio of power output to power input), the no-load losses (core losses) and the short-circuit losses (copper losses) are taken in account, and the percentage efficiency (17) may be defined as follows: Output power x 100% I’ = Output power + cores losses + copper losses

(9)

The exciting current is equal to the primary current when the secondary is open-circuited. This exciting current is non-sinusoidal, due to the non linear relationship between 3 and H for the amorphous alloy, and contains a large third-harmonic component as shown in Fig. 6. Although it is unimportant, the waveform of the exciting current must be reckoned with in numerous problems. For example, distortion introduced by exciting current is important in communication transformer. Harmonics in the exciting current of power-system transformers often have an important effect on interference between adjacent power lines and communication circuits [lo].

0 05 t (31

50 II

I

J+-b-Tm 0

5

10

15

Fig. 6. The exciting current and their harmonic components.

Core loss is approximately 87% lower for the amorphous unit, with a significantly lower exciting current (41% lower). When compared to silicon steel core, amorphous metal core exhibits reduced losses during magnetization. So, in comparison with conventional core transformer, amorphous core transformer shows radically reduced no-load losses and excitation power, which yields considerable energy saving.

Acknowledgements The authors are grateful to Allied-Signal Inc., Metglas Products, USA, for providing the amorphous alloy and to Dr. R. Hasegawa for the donation of this material. The present investigation was sponsored by FAPESP and CNPq-RHAE (Brazil).

References

5. Conclusions

Some aspects of amorphous Fe,,B,,Si, low power transformer core design, construction and experimental results have been presented. Table 2 Comparison of amorphous metal and silicon steel transformers: 1 kVA, 220 V/l10 V, 60 Hz Test

Amorphous metal

Silicon steel

No-load losses, W Short-circuit losses, W Exciting current, A Efficiency, %

2.5 41.5 0.14 95.6

19.0 44.0 0.24 93.7

[l] S.J. Chapman, Electric M~chhcry Fz&nmentnls, McGraw-Hill, 1991, p, 716. [2] R. Hasegawa, Techniche Rzmdsc/~n~~, 1991. [3] H.W. Ng, R. Hasegawa, A.C. Lee and L.A. Lowdermilk, Proc. IEEE, 79 (11) (1991) 1610. [4] D. Raskin and L.A. Davis, IEEE Spectrum, 18 (11) (1981) 28. [5] R. Hasegawa, J. Mngn. Mngn. Mater., 41 (1984) 79. [6] G.E. Fish, Pwc. IEEE, 78 (6) (1990) 958. [7] SK. Das, N.J. DeCristofaro and L.A. Davis, in S. Steeb, H. Warlimont (eds.), Rapidly Qzrenched Metals, 1985, p. 1621, [8] M.G. Say, Ahenzating CEtn.ent Mncliirzes, Pitman,. London, 1977. [9] D.M. Nathasing and H.H. Liebermann, IEEE Trans. Power Deiiv.,

P WRD-2

(3) (1987)

843.

[lo] M.I.T. Electrical Engineering Staff, Mngnetic foimers, Wiley, New York, 1954, p. 718.

Ciuits

nnd Trans-