Professional Documents
Culture Documents
R
h
xL
R
e
xL
R
r
xL
1
The loss tangent due to the winding resistance is:
tand
dc
q
c
l
w
10
9
xA
w
F
w
F
q
A
l
k
1
f
2
where q
c
is the copper resistivity, l
w
is the mean turn length, A
w
is
the cross sectional area of the winding space in the coil former, F
w
is
the copper space factor of the winding, A
l
is the inductance factor in
lH for 1 turn, f is the frequency and k
1
is the winding loss coef-
cient [7].
The hysteresis loss tangent is usually calculated by the Rayleigh
model at a weak magnetic eld as follows:
tand
h
4m
^
b
3pl
0
l
2
a
g
b
^
b k
2
3
where m is the Rayleigh coefcient, l
0
is the permeability of vac-
uum, l
a
is the amplitude permeability at a low applied eld and
k
2
is the hysteresis loss coefcient [7,18]. Hence, for a constant
maximum induction
^
b, the hysteresis loss factor is a constant inde-
pendent of the frequency.
The loss factor corresponding to the eddy currents without con-
sideration of the magnetic domain structure is:
tand
e
l
0
lD
2
x
2qb
k
3
f 4
where D is core diameter, b is the shape factor, k
3
is the eddy cur-
rent loss coefcient and q is the specic electrical resistivity [7].
As the equation indicates, the eddy current loss factor has a linear
relationship by frequency with the slope corresponding to the eddy
current loss coefcient.
The residual or excess contribution of the total loss factor can be
determined by the extrapolation of the maximum induction to
zero and subtraction of the eddy current loss from the total core
loss [3]. The excess loss has different contributions depending on
the working frequency. At high frequency range, the main contri-
butions to the excess loss factor arise from the ferromagnetic res-
onance and domain wall resonance [3]. However, at lower
frequencies, the main contribution of the excess loss originates
from the micro eddy currents produced at the domain walls which
could be neglected compared to the eddy current loss factor (Eq.
(4)) [7].
The total loss factor is obtained by adding the mentioned com-
ponents as [7]:
tand
tot
tand
dc
tand
h
tand
e
tand
r
5
or
tand
tot
k
1
f
k
2
k
3
f k
4
f 6
Fig. 7 shows the variations of the total loss factor of the as-pre-
pared and annealed samples versus frequency. It is obviously seen
that the total loss factor suddenly decreases to a minimum value at
low frequencies and then increases gradually by increasing the fre-
quency. At low frequencies, the winding loss which is proportional
to reciprocal of the frequency is dominant (Eq. (6)). However, this
part of the total loss factor is the same for all samples because of
the same winding parameters used in all of the measurements.
At higher frequencies, the total loss is dominated by one or more
of the other components.
Fig. 8 depicts the variations of core loss factors of the as-pre-
pared and heat treated compacts with frequency. To calculate the
core loss factor, the winding loss factor which is prevailing at very
low frequencies can be subtracted from the total loss factor.
According to Fig. 8, the core loss factor exhibits nearly a linear rela-
Fig. 6. The SEM image of the composites after annealing at 873 K for 30 min (a) and the X-ray maps of aluminum (b) and oxygen (c) at the cross section.
Fig. 7. Total loss factors of the as-prepared and annealed samples at different
frequencies.
296 M. Yaghtin et al. / Journal of Alloys and Compounds 581 (2013) 293297
tion with frequency for all samples. Regarding to this fact and the
linear relationship of the eddy current loss factor with frequency
(Eq. (6)), the slope of each line in Fig. 8 may indicate the eddy cur-
rent loss coefcient (k
3
). In addition, by extrapolation of the men-
tioned lines to zero frequency, the hysteresis part of the core loss
can be determined.
Table 1 lists the loss coefcients of the as-prepared and an-
nealed samples at different temperatures. According to Table 1,
the hysteresis loss coefcient is lower for the annealed SMCs and
its value decreases with increase of the annealing temperature.
The suppression of the hysteresis loss coefcient results from the
stress relieving and decreasing the dislocation density by the
annealing process [5]. From Table 1, the annealed pure iron com-
pacts exhibit a lower hysteresis loss coefcient compared to the
SMCs annealed at the same temperature. The existence of the alu-
mina insulation even after annealing at 873 K (see Fig. 6) as a non-
magnetic phase increases the internal stray eld and consequently
the hysteresis loss coefcient. The results are in a good agreement
with the permeability variations at very low frequencies (Fig. 5). As
can be seen, the pure iron compact after annealing at 873 K has the
highest permeability besides its lowest hysteresis loss coefcient.
In contrast, the annealing process increases the core loss factor at
higher frequencies (Fig. 8) and the core loss factor enhancement in-
creases with increase of the annealing temperature. At high fre-
quencies, the core loss factor can be described by the eddy
current loss factor which has an inverse relation with electrical
resistivity (Eq. (4)). The increase of the core loss by annealing at
high frequencies results in increasing the eddy current loss
coefcient according to Table 1. As explained before, the annealing
treatment can decrease the particles distortion produced during
the compaction step and consequently decrease the electrical
resistivity. This effect leads to increasing the eddy loss coefcient
of the SMCs after the heat treatment. Table 1 shows that the an-
nealed SMCs exhibit a lower eddy current loss coefcient in com-
parison with the heat treated pure iron compacts. This
additionally proves that the alumina insulation remains intact
upon the annealing (see Fig. 6) and contributes effectively to
decreasing the eddy current loss factor of the composites.
4. Conclusion
In this work, effect of heat treatment on the microstructure and
magnetic properties of the soft magnetic composites (SMCs) with
the alumina insulation was investigated. Furthermore, different
components of the magnetic loss factor were calculated by a loss
separation method for the as-prepared and annealed samples.
The following conclusions could be drawn:
1. Coating of the iron particles was carried out successfully by the
solgel method. The EDS, XRD, and FTIR spectra proved that the
iron particles were uniformly covered with alumina insulation.
2. Results of the FTIR and EDS analyses showed that alumina insu-
lation had a high thermal stability and the SMCs produced with
this insulation could be annealed at high temperatures without
any considerable degradation of the insulating layer.
3. Annealing treatment increased the permeability of the compos-
ites at low and medium frequency ranges. The annealed com-
posites exhibited signicantly higher frequency stability of the
magnetic permeability compared to the heat treated pure iron
compacts.
4. The results of the loss separation indicated that the annealing
treatment decreased the hysteresis loss coefcient while it
enhanced the eddy current loss coefcient of the SMCs.
5. The preservation of the alumina insulation upon the heat treat-
ment resulted in lower eddy current loss and higher hysteresis
loss coefcients of the annealed SMCs in comparison with the
heat treated pure iron compacts.
References
[1] A.H. Taghvaei, H. Shokrollahi, M. Ghaffari, K. Janghorban, J. Phys. Chem. Solids
71 (2010) 711.
[2] A.H. Taghvaei, H. Shokrollahi, K. Janghorban, H. abiri, Mater. Des. 30 (2009)
39893995.
[3] A.H. Taghvaei, H. Shokrollahi, K. Janghorban, Mater. Des. 31 (2010) 142148.
[4] H.R. Hemmati, H. Madaah Hosseini, A. Kianvash, J. Magn. Magn. Mater. 305
(2006) 147151.
[5] A.H. Taghvaei, H. Shokrollahi, K. Janghorban, J. Alloys Comp. 481 (2009) 681
686.
[6] B. Dong, C.R. Li, M.K. Lei, J. Lumin. 126 (2007) 441446.
[7] E.C. Snelling, A.D. Giles, Ferrites for Inductors and Transformers, Research
Studies Press, Letchworth, New York, 1983.
[8] L. Xu, H. Song, L. Chou, Mater. Chem. Phys. 132 (2012) 10711076.
[9] K. Velez, J.F. Quinson, J. SolGel Sci. Technol. 19 (2000) 469472.
[10] J.A. Wang, X. Bokhimi, A. Morales, O. Novaro, T. Lpez, R. Gmez, J. Phys. Chem.
B 103 (1999) 299303.
[11] V. Jayaraman, T. Gnanasekaran, G. Periaswami, Mater. Lett. 30 (1997) 157162.
[12] P. Padmaja, G.M. Anilkumar, P. Mukundan, G. Aruldhas, K.G.K. Warrier, Int. J.
Inorg. Mater. 3 (2001) 693698.
[13] R. Urlaub, U. Posset, R. Thull, J. Non-Cryst. Solids 265 (2000) 276284.
[14] A.G. Alejandre, M.G. Cruz, M. Trombetta, G. Busca, J. Ramirez, Mesoporous
Mater. 23 (1998) 265275.
[15] A. Raveh, Z.K. Tsameret, E. Grossman, Surf. Coat. Technol. 88 (1997) 103111.
[16] S. Yilmaz, Y. Kutmen-Kalpakli, E. Yilmaz, Ceram. Int. 35 (2009) 20292034.
[17] S. Wu, A. Sun, F. Zhai, J. Wang, Q. Zhang, W. Xu, P. Logan, A.A. Volinsky, J. Magn.
Magn. Mater. 324 (2012) 818822.
[18] M. Xingyu, X. Feng, G. Wenli, Z. Hongru, D. Youwei, Mater. Sci. Eng. A 396
(2005) 155158.
Fig. 8. The variations of the core loss factor with frequency for the as-prepared and
annealed samples.
Table 1
The hysteresis and eddy current loss coefcients of the as-prepared and annealed
compacts at different annealing temperatures.
SMCs Hysteresis loss
coefcient, K
2
Eddy current loss
coefcient, K
3
Pure Fe, annealed at
873 K
0.0012 4.16 10
6
Coated Fe, without
annealing
0.0026 1.95 10
6
Coated Fe, annealed at
673 K
0.0019 3.47 10
6
Coated Fe, annealed at
873 K
0.0016 3.66 10
6
M. Yaghtin et al. / Journal of Alloys and Compounds 581 (2013) 293297 297