Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract
Determination of the temperature rise as well as the power dissipated inside induction machines as heat is a matter of interest to both
customers and manufacturers. This is important since it affects the insulation materials, the cooling systems and the efficiency of the
machine. In this paper, some of established methods of full-load temperature rise testing of induction machines, such as direct loading
test, back-to-back test, phantom loading test, forward short circuit test, variable inertia test and mixed-frequency test, have been
reviewed. The experimental and simulation results of the mixed-frequency test using an inverter have then been analysed in detail.
Introduction
Review of the established methods for the fullload temperature rise testing of induction
machines
The most accurate and basic heat run of the induction machine is
to load the machine shaft directly with a mechanical or an
electrical load. This test is capable of producing the full-load
current flowing into the machine and the full-load mechanical
losses occurring inside the machine as well as the full-load rotor
current flowing in the rotor bars at rated rotor speed. However
coupling of a load to the machine shaft is not easy and hence the
test is expensive due to the following reasons:
The back-to-back test is one of the testing methods for the fullload heat run of an induction machine. This test has two distinct
merits, which are the economy and the accuracy. Because, in this
test, the full-load size plant is not needed for loading and the
major part of the test power is circulated rather than dissipated.
Further, the loss is measured as a net input rather than as the
small difference between two separately measured large powers.
However, it requires coupling of two induction machines which
will run at different speeds when generating or motoring. Thus
direct coupling of two induction machines is not possible if two
induction machines are operated from a constant frequency
supply.
Christofides and Adkins [1] suggested a method of back-to-back
test to overcome this problem as shown in Fig. 1, in where, the
two induction machines are fed from the same power supply and
coupled to two DC machines which are fixed on a stationary
bedplate. As the stators of the two induction motors are strapped
together and are free to rotate on their own bearings, the
induction machine 1 is driven as generator at speed of G while
the induction machine 2 runs as a motor at speed of M.The
speed of these two induction machines can be adjusted until the
stator frames have no tendency to rotate, which means that the
torques in two couplings are equal. They concluded that, the
input power Pac supplies the stator losses and very nearly the
whole of the load losses.
Apart from the direct loading test, the back to back test, the
phantom loading test, the forward short circuit test and the
variable inertia test also require the mechanical coupling of the
test machine to any other load machine. But the mixed-frequency
testing method does not require any mechanical connection of
EPE Journal Vol. 6 no 1 May 1996
37
DC supply
3 - Phase supply
Pac
P
P2
DC
Induction
Induction
DC
Machine
Motor 1
Motor 2
Machine
Floating Badplate
Floating gearbox
with a torque boost
Induction
Induction
Machine 1
Machine 2
AC
Power Supply
DC Power
Supply
AC
Power Supply
IM-1
IM-2
AC
Power
Induction
Motor
DC
DC
Motor
Power
Supply
Supply
1
Induction
Motor
Jact
Non-constant
Speed-ratio
Gearbox
38
Review of the Testing Methods for Full-Load Temperature Rise Testing of Induction Machines
circuit test". Finally, it is reported that this test produces a
temperature rise which is very close to that obtained by direct
loading test at the rated stator current and the rated rotor speed.
The "variable inertia test" introduced by Garvey et al. [6] for
temperature-rise testing of induction machines has an interesting
set of advantages and disadvantages. The principal disadvantage
must be that it does require mechanical coupling of the test
machine to a load. Against this, however, is the fact that the
"load" is very simple and compact compared with an electrical
load machine and the only net electrical power drawn from the
supply is that which appears as losses in the induction machine.
By having a large free-running inertia, the variation in rotor
speed can be very small indeed. The physical layout of this new
test is shown schematically in Fig. 5. It comprises three essential
elements:
(a)
(b)
(c)
39
Two Supplies
Alternator
Mains
Auxiliary Alternator
Inverter
Mains
Auxiliary Inverter
Transformer
Main Supply
Main Supply
Auxiliary Supply
w here
d (t)
dt
t
sin( 0 t)
oralternatively (t)= 0 (t)dt= 0 t+
( )
40
where
Vsa
V0
0
b
= Resultant voltage,
= Amplitude of base voltage,
= Angular base frequency,
= Angular beat frequency,
= Ratio of beat frequency over base frequency (b/0),
= Degree of either voltage modulation (V0/V0) or
frequency modulation (f0/f0).
Review of the Testing Methods for Full-Load Temperature Rise Testing of Induction Machines
TEST RESULT
SIMULATION RESULT
TEST RESULT
SIMULATION RESULT
Airgap
93
Core
55 C
Room
23 C
End windings
90 C
Rotor
104 C
Airgap
96
Core
61 C
Room
26 C
End windings
88 C
Rotor
104 C
TEST RESULT
SIMULATION RESULT
TEST RESULT
SIMULATION RESULT
TEST RESULT
10
10
Is(rms) = 7.201 A
Is(rms) = 7.08 A
-5
-5
-10
0
0.5
0.1
t(s)
0.15
0.2
TEST RESULT
303.5
302.5
302
0.5
0.1
t(s)
0
302.65
302.6
302.55
302.5
302.45
303
301.5
0
-10
0.15
0.2
302.4
302.35
0
0.5
0.1
t(s)
0.15
0.2
SIMULATION RESULT
0.5
0.1
t(s)
0.15
0.2
41
from the direct loading test but the mechanical coupling of the
test machine to any other load machine is still necessary in these
tests. The mechanical connection of the test machine to a load
machine can be eliminated when the mixed-frequency testing
method is used. However, the conventional mixed-frequency test
requires the use of two different power supplies of different
frequencies and amplitudes. Different supply arrangements for
the mixed-frequency testing have been given as a single supply
or two supplies. Arrangements of these supplies include an
alternator or an inverter or a combination of an alternator and an
inverter. Finally, as mentioned in the previous sections, the use
of an inverter as a mixed-frequency power supply has some
advantages over the use of an alternator such as eliminating the
use of extra DC motors to drive the alternator and the use of
extra power supplies of the alternator and DC motor. On the
other hand, the use of inverter increases the core losses over the
full-load value due to the harmonics in the inverter output
voltage, but it is possible to include some form of filter between
the inverter and the induction machine to eliminate these
harmonics.
References
Fig. 12 Inverter output voltage and its harmonic spectrum for
case 2 (fb = 10 Hz, fswitching = 3 kHz, fsampling = 40 kHz).
Conclusions
Different established methods for the full-load temperature rise
testing of induction machines have been reviewed. It has been
mentioned that, the most accurate and basic heat run of the
induction machine is to load the machine shaft directly with a
mechanical or an electrical load. However coupling of a load to
the machine shaft is not easy and hence the test is expensive.
Apart from the direct loading test, the back to back test, the
phantom loading test, the forward short circuit test and the
variable inertia test also provide the full-load temperature rises
of induction machine which are very close to those obtained
42
1.
Review of the Testing Methods for Full-Load Temperature Rise Testing of Induction Machines
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Ilhami OLAK
Ilhami OLAK has received his BSc and
MSc degrees in The Department of
Electrical Engineering from Technical
Education Faculty of Gazi University
ANKARA-TURKEY in 1985 and 1988
respectively. He worked as a research
assistant in the same department from
1986 to 1989. Then he has his MPhil in
Electrical and Electronics Engineering Department from
Birmingham University and his PhD in Electrical Engineering
Department from Aston University in Birmingham, UK in 1991
and 1994 respectively. His main interests are electrical machines,
control and power electronics. (Gazi niversitesi, Teknik Egitim
Fakltesi, Elektrik Egitimi Blm, Besevler Ankara-Turkey.
Tel.:+90 312 2123962, Fax.: +90 312 2120059).
Gngr BAL
Gngr BAL has received his BSc and
MSc degrees in The Department of
Electrical Engineering from Technical
Education Faculty of Gazi University
ANKARA-TURKEY in 1983 and 1988
respectively. He worked as a research
assistant in the same department from
1985 to 1989. Then he has his PhD in
Electronics and Electrical Engineering Department from
Strathclyde University in Glasgow, UK in 1993. His main
interests are electrical machines, AC drives and control. (Gazi
niversitesi, Teknik Egitim Fakltesi, Elektrik Egitimi Blm,
Besevler Ankara-Turkey. Tel.:+90 312 2123962, Fax.: +90 312
2120059).
etin ELMAS
etin ELMAS has received his BSc and
MSc degrees in The Department of
Electrical Engineering from Technical
Education Faculty of Gazi University
ANKARA-TURKEY in 1986 and 1988
respectively. He worked as a research
assistant in the same department from 1987
to 1989. Then he has his PhD in School of
Electronic and Electrical Engineering Department from
Birmingham University, UK in 1993. His main interests are
electrical machines, control, power electronics and artificial
intelligent. (Gazi niversitesi, Teknik Egitim Fakltesi, Elektrik
Egitimi Blm, Besevler Ankara-Turkey. Tel.:+90 312
2123962, Fax.: +90 312 2120059).
The authors:
43