Académique Documents
Professionnel Documents
Culture Documents
= =
= = =
1 1
1
1
1 1 1 1
2
2
cos
cos
N f
U
U
t U t N
dt
d
N u
, (1)
where sin
1 1
= ,
1
^
=S
Fe
*B
Fe
is the peak-value
of magnetic flux coupling the primary winding, =2f -
is the pulsation and f is the frequency.
The primary power will be:
( )
1 1 1
1 1
1 1 1
2
I N f
I U
I U P ==
= = . (2)
Fromthis relation we can see that to increase
the power the best way is to increase the frequency
because the increasing of magnetic flux lead to large
transversal section area of magnetic core and the
increasing of magnetomotive force (N
1
I
1
) lead to
increasing the copper mass. The increasing of the copper
mass means the increasing of magnetic core window and
hence and of the core mass.
Another problemis the presence of a very large
air-gap of the magnetic core, about g=5 mm. This large
air-gap requires a very large magnetizing current,
according to the Ampere law written below:
1 1
0 0
1
i N
g B
l B
g H l H l d H F
g
rFe
Fe Fe
g Fe Fe
C
=
= + = =
, (3)
where: N
1
is the number of turns of primary windings;
i
1
- is the transformer magnetizing current; g is the air-
gap length; B
g
is the magnetic flux density in the air-
gap; B
Fe
is the magnetic flux density in the ferrite
magnetic core; l
Fe
is the length of magnetic flux line in
the ferrite core;
0
is the air magnetic permeability and
rFe
is the relative magnetic permeability of ferrite
material.
Because of the high value of the
rFe
(>1000)
the first termis negligible. The secondary term, usually
do not appear in a classical transformer, but now it has a
high value. This value requires a high magnetizing
current. In such a way we must to enlarge the
dimensions of primary coil.
If we imposed the values for frequency f, the
air-gap length g, the magnetic flux density in the air-gap
B
g
and in the core B
Fe
, we can calculate the cross-section
area of the magnetic core S
Fe
with relation (2) and (3).
The diameter D
1
of the shaft will be calculated
frommechanical resistance reasons and then knowing
the S
Fe
we can calculate the diameter D
2
.
It should be known that the primary current has
two components:
1 11 1
i i i + = , (4)
where i
11
is the util current and i
1
is the magnetizing
current, but they are adding vectorial. Because the first
current is almost active and the latter is pure inductive
and the magnetizing current is of the same value as the
util current we can assume that the overall current is:
1 1
2
I I . (5)
From (1), if we impose the primary voltage U
1
,
we can calculate the primary numbers of turns N
1
, and
then the total primary magnetomotive force:
1 1 1 1
2
F I N F = = . (6)
But:
1 1 1 1
2
f a w
k J S F = , (7)
where S
w1
is the primary window area; J
a1
is the
permissible current density in primary conductor and k
f1
is the filling factor of primary window. Fromhere we
can calculate the primary window area and then the
diameter D
3
if we choose the window height h
1
+h
2
.
The diameter D
4
will be calculated in order to
assure for the outer yoke the magnetic core cross-section
area S
Fe.
The secondary window can be a little small than
the primary because of the lack of magnetizing current,
but fromsymmetry reasons we can take the secondary
side of the same shape and dimensions as the primary
side.
4. APPLICATION
The design methodology presented before will
be used to electromagnetic design of a rotating
transformer excitation system(RTES) for synchronous
motors of 300KW rated power.
4.1 Magnetic core design
Taking in account that the needed excitation
power of a synchronous electric motor is about 2% from
the main machine power, the rated power of RTES will
be:
kW P 6 300
100
2
1
= = . (8)
A lot of initial data will be also imposed:
U
1
=560V;
U
2
=323V;
Frequency f=20 kHz;
B
Fe
=0.2T;
B
g
=0.05T;
g=5mm;
k
f1
=0.2, it take in consideration and skin effect;
J
a1
=3 A/mm
2
.
Fromthe relation (3) we can calculate:
A g
B
F
g
200 10 5
10 4
05 . 0
3
7
0
1
=
(9)
According to (6) the overall magnetomotive
force will be:
A F F 8 . 282
1 1
=
. (10)
Fromrelation (2) we can calculate the primary
coupling magnetic flux:
mWb
F f
P
34 . 0
8 . 282 20000
6000
1
1
1
=
= =
. (11)
Now, we can calculate the magnetic core cross-
section area:
2 4
5
1
10 17
2 . 0
10 34
m
B
S
Fe
Fe
= .(12)
If we impose the shaft diameter D
1
=30mm from
mechanical resistance reasons (its calculations is not the
subject of this paper), we can calculate the next
diameter:
mm
S D D
Fe
56 1700
4
30
4
2
2
1 2
= +
= + =
. (13)
Now we will calculate the primary window aria:
2
1 1
1
1
334
2 . 0 3 2
8 . 282
2
mm
k J
F
S
f a
w
=
=
= (14)
If we impose the height of primary window
h
1
+h
2
=20+5=25mm, we can calculate the width of
primary window:
mm
h
S
a
w
17
20
334
1
1
1
= = . (15)
The next diameter will be:
mm a D D 90 17 2 56 2
1 2 3
= + = + = . (16)
The outer diameter of magnetic core will be:
mm
S D D
Fe
102 1700
4
90
4
2
2
3 4
= +
= + =
. (16)
And yoke width will be:
mm
D
S
h
Fe
10
56
1700
2
3
=
= =
(17)
4.2 Transformer coils design
Now, we can design and the primary and
secondary windings.
Fromrelation (1) we can write:
. 18
10 34 . 0 20000 2
560 2
2
2
3
1
1
1
turns
f
U
N
=
=
(18)
The primary magnetizing current will be:
A
N
F
I 1 . 11
18
200
1
1
1
= = =
. (19)
The primary util curent will be:
A
U
P
I 7 . 10
560
6000
1
1
1
= = = . (20)
It can be seen that the assumption that these both
currents are of the same value but of the different phases
is correct.
The effective primary conductor section area is:
2
2 2
1
2
1
2
1
1
5
3
1 . 11 7 . 10
mm
J
I I
s
a
Cu
=
+
=
+
=
(21)
and the diameter of this conductor will be:
mm
k
s
d
f
Cu
2 . 3
6 . 0
5 4 4
1
1
=
=
. (22)
These conductors can be put in four layers of five
conductors on a layer in the frame of 17x20 mm
2
.
If we dont use litz-wire, we must take in
account the skin effect. According with this
phenomenon, the high frequency current will path only
through a thin layer placed at the surface of conductor
and having the penetration thickness:
mm
f
r
5 . 0
20000 1
10 0172 . 0
503
503
6
1
=
. (23)
The diameter of full conductor will be:
mm
s
d
Cu
2 . 3
5 . 0
5
1
1
=
=
. (24)
It can be seen that in this case the diameters are similar.
The secondary winding design can be made on
the same way.
5. FEM SIMULATION OF RTES MAGNETIC
FIELD
The goal of magnetic field simulation by Finite
Elements Method (FEM) is to check the correctness of
magnetic flux density imposed in the magnetic core
B
Fe
=0.2T and in the air gap B
g
=0.05T.
It has been used the FEM computational
program EMFCAD, elaborated in the frame of
Changwon University, Korea [7].
The results are presented in figures 5,6 and 7.
Fig. 5 The magnetic flux flow-chart of the designed
RTES
B B
A
A
Fromthis picture can be seen the high level of
primary stray magnetic flux. From15 primary magnetic
flux lines only 4 is coupling and the secondary windings
that means a dispersion factor:
75 . 3
4
15
= =
=
u
t
k
(25)
very close to the value imposed at the design:
4
05 . 0
2 . 0
= =
=
g
Fe
g g
Fe Fe
u
t
B
B
S B
S B
k
,(26)
Fig, 6 The magnetic flux density B
y
, in the middle of the
air-gap (section A-A)
It can be seen that in ihe inner air-gap the
medium value of the magnetic flux density is about
0.06T, and in the outer air-gap is only about 0.03 T, that
means an overall average value of 0.045T very close to
the value take in account at design (0.05T).
Fig.7 The magnetic flux density B
y
in the core,
(section B-B)
It can be seen, that the magnetic flux density B
y
in the inner core is about 0.19T and in the outer core
about 0.15T, very close to 0.2T, value used in the
analytical design.
6. CONCLUSIONS
In the first paper approaching this research
field, our team presented the main features, the
methodology design and the FEM simulation of the main
part of Contactless Excitation Systemfor Synchronous
Machines (CESSM) Rotating Transformer (RT). The
rest of CESSM will be presented in a future paper.
Because of the using of high frequency, the
dimensions of a transformer of significant power (6
kVA) are very small (outer diameter 102 mmand length
75 mm). The rotor rectifier is also of the small
dimensions because of the actually advances in the field
of power electronics, and this rectifier will be controlled,
the command signals coming by another pair of rotating
transformer data windings. So this excitation system
will be very fast and maintenance free, fitted for all types
of synchronous machines with requirements of excitation
forcing and voltage or power factor regulations.
The contactless excitation systems presented in
[9] and [10] are conceived with older electronics
components and for special applications.
Can be imagined and hibrid systems [9]:
contactless excitations system for a.c. exciter of an
brushless synchronous motor.
REFERENCES
1. M. J ufer, L. Cardoletti, Induction Contactless
Energy Transmission System for an Electric
Vehicle, Procc. Of ICEM96, Vigo, Spain, pp. 343
347.
2. J urgen Meins, Contactless Power Supply for
Transport Systems, Procc. Of MAGLEV98,
Damanashi, J apan, pp. 268-273.
3. http://www.vahle.de/
4. http://www.wampfler.com/
5. M. J ufer, M. Perrottet, "Contactless Energy and
Information Transmission for Electric Drives", First
International Symposium on Advanced
Electromechanical Motion Control Systems,
Electromotion'95, Cluj-Napoca, Romania, May 25-
26, 1995, pp. 1-8.
6. D. Finney, The Power Thyristor and its
Applications, McGraw-Hill, 1980, London, UK.
7. G.R. Slemon, A. Straughen, Elrctric Machines,
Addison-Wesley Publishing Company, 1982.
8. Electromagnetic-Field Computation Program, EMF
CAD Ver. 1.0, CMS Technology Inc., Korea.
9. M. L. Waters, Field Excitation System for
Synchronous Machines Utilizing a Rotating
Transformer Brushless Exciter Generating
Combination, patent US3908161/1975, Applicant
GEN ELECTRIC.
10. Stich Frederik, Brushless Exciter for Controlling
Excitation of a Synchronous Machine, patent
US4659973/1987, Applicant SIEMENS
ENERGY&AUTOMAT (US).