Special Feature
U
nderstanding Complete Tests
Performed On Induction Motors
by
William R. Finley, Mark M. Hodowanec, Khursheed S. Hussain, and John Larabee
Siemens Energy & Automation, Inc.
Introduction
Locked-Rotor Test at Rated Frequency
When new motors are purchased, complete tests
can be conducted to verify performance and integrity. These tests are not standard and may add to the
motor cost without adding significant value to the
purchaser. The value of the tests depends on criticality
of the application, the user’s experiences, motor size,
motor voltage, etc.
There are many standards regarding testing of induction motors. Standards such as NEMA MG1, IEEE
112, IEC 60034-01 & -02, API 541, and IEEE 841 make
recommendation as to what tests are required and how
they should be performed. There are many different
specified methods to performance test induction motors, all requiring that the motor be loaded (i.e., heat
run). The different test methods do not necessarily
produce the same results.
Two common test methods are used today to load
motors: coupled load test and dual frequency method. The coupled load test requires that the motor be
coupled to a load machine and placed under rated
load. The dual frequency test involves applying both
50- and 60-hertz power to the motor at the same time,
simulating full-load heating. It is important the user
understands the test employed and, if comparing motors from different vendors, that the motors need to
be tested using the same test method in order to keep
values comparable.
The locked-rotor test at rated frequency is used
to determine the locked-rotor torque (LRT) and current (LRA). In order to determine the values at rated
voltage, at least three test points of voltage versus
current, watts, and sometimes torque are taken to as
high a voltage as possible and then extrapolated to
rated voltage on log-log graph paper to establish the
desired values.
A complete test uses a number of individual tests.
These individual tests include:
• Locked-rotor test at a rated frequency
• Speed-torque curve
• No-load saturation curve
• Dual-frequency heat run or coupled heat run
Speed-Torque Curve
On large motors it is difficult and costly to measure
directly torque versus speed at rated voltage. In such
cases the test is run first to determine the shape of
the speed-torque curve. The curve is then calibrated
utilizing the test results from the locked-rotor test to
establish the actual speed-torque curve of the motor.
A typical speed-torque trace is shown in Figure 1.
Figure 1 — Torque versus Speed Trace
These four parts are detailed as follows.
Summer 2004
1
The speed-torque trace is normally determined
utilizing a tachometer to measure motor speed as it
accelerates to its no-load speed. The output of the tachometer is fed into a computer, where it is recorded
as a function of time. The output is then differentiated
with respect to time to arrive at the rate of change of
speed versus time, which is the angular acceleration
of the motor.
This test is normally done at reduced voltages so as
not to damage the test equipment and to get a good
sampling. The resulting curve represents the shape
of the speed-torque curve but does not yet establish
absolute torque.
The speed-torque trace is then calibrated using the
locked-rotor torque value obtained in the locked-rotor test described above. By assigning this value to
the curve at zero speed, a speed-torque curve is now
accurately defined in absolute values at all speeds.
Dual-Frequency Heat Run
Dual-frequency heat run is a temperature test of an
induction machine under simulated load conditions.
The test involves using two separate sources of power
with two separate frequencies: a primary source of
rated frequency and a secondary source generally
10 hertz below the rated frequency. The two sources
are set up to supply power simultaneously to the test
machine by being connected either in series or superimposed by use of a series transformer. These are
shown schematically in Figure 3 and Figure 4.
No-Load Saturation Curve
This test is performed to determine the windage,
friction, and core losses in a motor. The saturation
curve is taken with the motor running without any
load. The test is usually performed after half an hour
or more of the no-load run, to ensure the bearings have
run in and input values have stabilized.
At rated frequency, the line voltage on the motor
is varied in steps from 125 percent of rated down to
a value where further voltage reduction results in a
disproportionate increase in the current. Voltage, current, power, and winding temperatures are recorded
at each step.
To segregate the losses, power input minus the
stator I2R loss is plotted versus voltage, and the curve
extended to zero voltage. Refer to Figure 2. The intercept on the power axis is the windage and friction
loss. Core loss at rated voltage and frequency can then
be obtained by subtracting the value of the windage
and friction loss from the total loss from the curve at
rated voltage point.
Figure 2 — Determination of Windage and Friction Losses
2
Figure 3 — Dual-Frequency Power Sources, 50 and 60 Hertz
Connected in Series
Figure 4 — Dual-Frequency Power Sources, 50 Hertz Power
superimposed on 60 Hertz by Use of a Coupling Transformer
The frequency that the motor sees changes completely 10 times per second. This continuous change
causes the revolving magnetic field inside the motor to
change its synchronous speed between that of 50 hertz
and 60 hertz. When the motor is under the influence
of the 60-hertz supply, the motor accelerates towards
the 60-hertz synchronous speed, drawing current in
the process to achieve the acceleration and operating as a motor under high slip. However, because of
its rotor inertia, it cannot reach that speed instantly.
One tenth of a second later the motor sees 50-hertz
power. The motor then decelerates towards the 50hertz synchronous speed. The slip being negative,
the motor now generates current and feeds it back to
supply lines as an induction generator. With proper
adjustments to input parameters, a steady operating
NETA WORLD
condition can be achieved wherein the motor sees
rated root-mean-square (rms) voltage and rated rms
line current. The wave shapes are not sinusoidal, but
tests show that they produce similar heating in the
motor. Table 1 lists comparative test results reported
by various manufacturers.
Source
HP
Volts
Poles
Hz
A
B
B
A
A
C
450
4000
3000
1500
4000
1340
3300
4000
4000
4160
6900
6000
2
4
6
8
10
18
60
60
60
60
60
50
Rise by
Resist.
Load
Dual
56.3
64.3
62.85 63.0
51.5
52.6
62.8
71.1
45.6
49.1
44.3
45.0
Table 1 — Comparative Temperature Rise Between
Dual Frequency and Load Test
The rated condition is generally reached when
the 50-hertz input voltage reaches 20 to 30 percent
of the 60-hertz rated voltage as measured at V1 and
V2, respectively, in Figure 3.. During the duration
of the heat run, the terminal voltage and current
of the motor are maintained at their rated 60-hertz
values. Volt, ampere, and kilowatt readings at the
motor terminals are recorded along with the motor
temperatures. After the machine temperatures (as
indicated by stator resistance temperature detectors
or auxiliary thermocouples) have stabilized, the voltages of the auxiliary power and the prime power are
reduced. After the motor is stopped and all breakers
are opened and locked out, resistance is measured to
evaluate temperature rise.
During the heat run, the motor is being supplied
from two power sources at different frequencies,
and is subjected to the oscillatory torques associated
with these frequencies. Consequently, the vibration
will be abnormal during this condition and may not
meet the normal limits of vibration. For this reason a
no-load cold vibration is measured at rated voltage
before the application of the auxiliary power. Then,
at the end of the heat run after the temperatures on
the machine have stabilized, the auxiliary power is
removed, and the vibration at rated frequency and
voltage is measured again to determine the vibration
of the machine at normal running temperature. This
is done without stopping the test motor, which allows the hot vibration to be recorded quickly since
the machine — especially an open machine — cools
Summer 2004
down rapidly after the auxiliary power is removed.
IEEE 112 recommends that the vibrations should be
measured while the motor is within 25 percent of the
normal operating temperature.
Dual-frequency load testing is a cost-effective
method for temperature testing of general purpose
and vertical induction motors. The test setup is simple
— no test coupling, rigging, or alignment is required. It
takes 50 to 60 percent less time to rig and test the motor
than by the conventional coupled load method.
Coupled Heat Run
Coupled heat run is the direct-loading method for
temperature testing of electric motors. The test machine is coupled to a dynamometer or a load machine.
The load on the dynamometer or the load machine is
increased until the test machine reaches rated load.
In a coupled load test, the test machine is installed
with a test coupling and then rigged and aligned
to the load machine as shown. It is important that
the test motor be set firmly on a stiff test bed. If it is
raised on rails or blocks to match the shaft height of the
load machine, the rails and blocks must be perfectly
squared off and have adequate stiffness. Similarly, the
test couplings, the center spool piece, the coupling on
the load machine, and the load machine itself must be
well balanced and aligned accurately to assure that no
vibration is introduced as a result of inferior couplings
or rigging. The setup on the test floor is normally temporary since motors of all sizes are tested in the same
location. The hot vibration readings need to be taken
while the motor is still hot from the loaded run but uncoupled to remove the effects of misalignment, etc.
IEEE 112 recommends that the vibrations be measured while the motor is within 25 percent of the normal operating temperature. This is sometimes difficult
to do for a coupled heat run test. In this case it can be
necessary to stop the machine and uncouple it from
the load, then start it up again to measure the vibration. This, of course, is not necessary if the vibration
and vectorial change from cold to hot is good while
coupled.
Determine Performance Characteristics
There are many different ways to determine the
performance characteristics of an induction motor.
These characteristics include efficiency, power factor,
load current, and speed. In North America motors are
tested in accordance with IEEE 112, although even
within that there are still many different methods to
use, including methods B, C, E, E1, F and F1. These
methods do not require the coupled heat run method,
but it is logical that, if a motor is coupled up for a method B or E test, it should be a coupled heat run. In addition, if a method F efficiency test is to be performed
(where it is not necessary to couple up the motor), a
dual-frequency heat run should be performed. Heat
3
run and efficiency methods should never be implied or
assumed. Performance will vary significantly depending on the method used. Efficiency will be accurate and
higher if utilizing methods B or E, whereas method
F will normally provide slightly less accurate and
lower efficiency. Nevertheless, economics frequently
outweigh the concern for accuracy, and thus, method
F tests are commonly performed.
In addition to IEEE 112 test methods, the IEC and
JEC also have methods for testing induction motors.
These test methods differ from one another in their details and arrive at different results. Cummins, Bowers,
and Martiny in 1981 compared in detail these various
methods for testing induction motor efficiency. The
JEC and IEC methods tend to be less rigorous, and
provide less accurate results when compared to IEEE
methods, albeit they are less expensive to conduct.
Table 2 illustrates the differences in results when the
efficiency of a single machine was evaluated per various methods.
IEEE-112
JEC-37
IEC 34-2
ANSI-C50-41
Stator I2R, kW
13.9
13.1
13.9
13.9
Rotor I2R
11.4
11.4
11.4
11.4
Core Loss kW
8.0
8.0
8.0
8.0
Wind & Fr., kW
4.0
4.0
4.0
4.0
Stray, kWQ
13.2
0
4.7
11.2
Total, kW
50.5
36.5
42.0
48.5
Output, kW
932.5
932.5
932.5
932.5
Input, kW
983.0
969.0
974.5
981.0
Efficiency
94.9
96.2
95.7
95.1
Table 2 — Table of Stray versus Test Method
M=Measured
(1)- in this case stray is a percentage of the input, which
makes the levels a little higher.
A point of interest is the variation in stray load loss
used in the different methods. Please see Table 3. In
IEEE 112, one has the option either to test for the load
loss (such as in method B or the Morgan test — also
known as reverse-rotation test — in methods E and
F), or use an assigned value to the load loss (such as
in methods E1 and F1). Tested values of stray load
loss provide the most accurate measurement of the
efficiency as compared to using assigned values, but
the cost could be prohibitive.
4
IEEE 122 Eff. Test Methods
HP
B C
E
<200
M
200-1500
>1500
E1
IEC (1) JEC
F
F1
M 1.6%
M
1.6%
.5%
0
M
M
1.25
M
1.2%
.5%
0
M
M
.9%
M
.9%
.5%
0
Table 3 — Efficiency Different Test Method
Conclusion
No matter what tests are chosen, it is important to
understand what information is being obtained from
the tests specified. The benefit of having information
obtained from rigorous tests must be compared against
the additional testing cost. Critical applications that
have historically been problematic may benefit from
a complete test. Alternatively, noncritical, trouble-free
applications would add unnecessary cost to the motor if the same tests were specified. By understanding
the information delivered by the many different motor tests available, optimal test requirements can be
specified.
References
1. Cummings, P. G., Bowers W. D., and Martiny, W. J.,
“Induction Motor Efficiency Test Methods,” IEEE
Transactions On Industry Applications, Vol. IA-17, No.
3, May/June 1981.
2. IEEE 112 Test Procedure for Polyphase Induction Motors
and Generators, 1996.
3. ANSI C50.41-2000 Polyphase Induction Motors for
Power Generating Stations.
4. NEMA Standards Publication No. MG 1-1998 (Rev.
1) Motors and Generators, 2000.
5. IEEE 522-1992 Guide for Testing Turn-to-Turn Insulation on Form-Wound Stator Coils for Alternating-Current Rotating Electric Machines.
6. Finley, W. R., Hodowanec, M. M., Holter, W. G.,
“An Analytical Approach to Solving Motor Vibration Problems,” IEEE Transactions, Vol. 36, No. 5,
September/October 2000.
William R. Finley received his BS in Electrical Engineering
from the University of Cincinnati. Present responsibilities for
Siemens Energy and Automation include being the operations
manager for the NEMA product out of Little Rock, Arkansas, and
manager of engineering for the same NEMA product where he is
responsible for design, development, and quality assurance out
of Norwood, Ohio. He is a Senior Member of IEEE and has previNETA WORLD
ously published over 12 technical papers, which resulted in one
first place, two second place, and one honorable mention award.
Most of the papers were included in the IEEE transactions. He is
currently active in over ten NEMA and IEC working groups and
subcommittees. He is Chairman of NEMA’s Large Machine Group
and International Standardization Group.
Mark M. Hodowanec received BS and MS degrees in Mechanical Engineering from the University of Akron, Ohio. Currently, he
is the manager of mechanical engineering for ANEMA induction
motors built in the US at Siemens Energy and Automation, Inc.,
Cincinnati. For the past ten years he has worked in a variety of
engineering positions including design, product development,
order processing, shop testing, and field support. He is currently
active on various NEMA, IEEE, and API working groups. In
addition to his ANEMA motor experience, Mr. Hodowanec has
worked on a wide assortment of induction motors such as NEMA,
submersible, and MSHA motors. He is the author of numerous
published technical articles.
Summer 2004
Khursheed S. Hussain received his BS from University of
Poona, India, and his MS in Electric Power Engineering from
Rensselar Polytechnic Institute, Troy, NY. Currently, he is the
principal product engineer for ANEMA induction motors built
in the US at Siemens Energy and Automation, Inc., Cincinnati.
He has over thirty-five years of engineering and project management experience in motors and generators for industrial, nuclear,
and government applications, including ten years in application,
design, and development of ship service generators for the US
Navy. He is an IEEE member, and member of the working group
on IEEE Std. 112, Standard Test Procedures for Induction Motors and
Generators.
John A. Larabee received his BS in Electrical Engineering
from Florida International University, Miami. Currently, he is
manager of product engineering and testing for Siemens Energy
and Automation, Inc., Cincinnati. He has a background of design
engineering, process engineering, and information technology
within Siemens.
5