FIELD OF THE INVENTION:
The present invention relates to determination of whirl direction of shaft. More
specifically, the invention relates to a method for determining the whirl direction through
motor current of the 3-phase induction motor which is operating the said shaft.
BACKGROUND OF THE INVENTION:
Conventionally, the identification of whirl direction of a shaft coupled with a 3-0
induction motor has been mainly performed using vibration signals. It has been observed
that shaft whirl results in shaft vibrations in the plane perpendicular to the shaft axis.
The analysis of the vibration in two mutually perpendicular directions may lead to
detecting whirl direction. However, condition monitoring using vibration signals has
numerous disadvantages such as signals background noise due to external excitation
motion, sensitivity to the installation position, and their invasive measurement nature.
Similarly, other whirl direction detection techniques are also based on vibration signals
and orbit analysis of the vibration signals using displacement probes or accelerometer
installed perpendicularly to the shaft. The two probes give measurement for two
mutually perpendicular directions. Further, signals processing method like full spectrum is
employed to determine the whirl direction. But these methods require additional effort
for mounting two proximity probes or accelerometer near or on shaft in mutually
perpendicular direction. Hence, they are not easy in implementation and are bit
expensive. Some prior art documents related to the present invention are

0049733 Al 2/2013 Neti et al.
5,519,337 5/1996 Casada et al.
5,864,058 1/1999 Chen et al.
6,092, 029 7/2000 Bently et al.
6,205,851 Bl 3/2001 Jogi et al.
7,676,285 B2 3/2010 Hoyte et al.
P. Goldman, A. Muszynska, "Application of full spectrum to rotating machinery
diagnostics," Orbit First Quarter, pp. 17-21, 1999.
Whirl detection using stator current signature analysis has been rarely proposed although
condition monitoring using stator current signals is widely used in industrial applications.
In general, current monitoring techniques are employed in the detection of mechanical
faults in electric machines as the load torque oscillations modulates the stator current,
whereby the stator current signature is analyzed for detection mechanical perturbations
due to fault.
Therefore, there exists a need for an improved method and system for the detection of
whirl direction which can offer significant economic savings and easy implementation.
Using motor current signature analysis as the tool for whirl direction identification has the
potential to serve both the purpose i.e. easy implementation and economic saving.

SUMMARY OF THE INVENTION:
In accordance with an embodiment of the invention, a method for detecting whirl
direction of a shaft coupled with a 3-phase induction motor is provided. The method
includes acquiring stator current signals of an induction motor. The method also includes
processing of the acquired stator current signals data to extract spectral information/The
method further includes detecting of whirl direction based on analyzing the spectral
information. In accordance with another embodiment of the invention, a system for
detecting whirl direction of a shaft coupled with induction motor is provided. The system
includes two current sensors for acquiring stator current signals of an induction motor.
The system also includes a data acquisition system (DAQ) for acquiring stator current
signals. The system further includes processing and analyzing of stator current data to
identify the whirl direction from the spectral information of the processed data.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS:
The features, benefits and advantages of the present invention can be better understood
through the detail block diagram and the schematic drawing presented above, wherein:
Fig.l is a block diagram representation of method for whirl detection through motor
current analysis.
Fig.2 is a schematic representation of experimental setup for acquiring motor current
signals.

Fig.3 is a flow chart representation of method for whirl detection through motor current
analysis.
Fig.4 is a modified full spectrum analysis algorithm for motor current signals considering
a 3-0 induction motor.
Fig.5 is a full spectrum plot for determining whirl direction using vibration signals.
Fig.6 is a modified full spectrum plot for.determining whirl direction using electrical motor
current signals.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT OF
THE INVENTION:
A whirl detection technique for a shaft coupled with a 3- induction motor using motor
current signature analysis is described herein. The technique is based on stator current
signature analysis. Whirl is the displacement of shaft along two axes, which results in an
orbital motion or whirl motion. If the direction of the whirl motion is same as the spin
direction of the shaft, the whirl motion is referred as forward whirl and if the direction of
whirl motion is opposite to the spin direction of shaft, the whirl motion is referred as
backward whirl.
One embodiment of the invention provides method for whirl direction detection. It
includes acquiring stator current signals from any two phase of a 3-0 induction motor.

The acquired time domain signals is further transformed to frequency domain signals
through FIT. The transformed signals are further processed' using modified full spectrum
analysis. The modified full spectrum analysis yields forward and backward amplitude.
Then the modified full spectrum plot with forward amplitude on positive frequency side
and backward amplitude on negative frequency side is plotted. The forward amplitude
and backward amplitude corresponding to IX harmonic component is compared to
determine the whirl direction of the shaft.
FIG. 1 is a block diagram representing of the steps involved for the whirl direction
detection using electric motor current signals. In first step, motor current signals which
act as a representative of a 3-0 induction motor (10) is acquired. In the present
invention, current signals are considered as the representative signals. Further, in second
step the time domain current signals are transformed to frequency domain based on Fast
Fourier Transforms (20). In next step, the frequency domain signals are further
processed using modified full spectrum analysis (30). Further based on full spectrum
analysis, the forward amplitude (12) and backward amplitude (14) is determined. Finally
in last step, whirl direction is determined by comparing the relative magnitude of forward
and backward amplitude.
In another embodiment, a system for detecting whirl direction of a shaft coupled with an
induction motor is provided. The system comprises of two current sensors deployed for
acquiring stator current signals which acts as a representative of an induction motor

(10). The acquired stator current signals are recorded using a data acquisition system
(DAQ) (40). The recorded data is further processed with data processing module. The
processing of data is done in order to extract Spectral information. The system comprises
of a 3-Φ induction motor (10) coupled with a shaft (11), wherein the motor is the driver
member while shaft is the driven member.
Referring to FIG. 2, a schematic representation of experimental setup for acquiring stator
current signals is shown. The setup includes a 3-Φ induction motor provided with a
power supply (8). The induction motor is coupled with a shaft using coupling. The shaft
is supported on bearing (11'). Current probes (32) are provided for acquiring stator
current data. The acquired data is recorded using DAQ (40). The recorded data is further
feed to a data processing module to extract the spectral information.
Induction motors are electromagnetic rotating machine which converts electrical energy
into mechanical energy. There exists various multi-phase induction motors, but a 3-Φ
induction motor is widely used in industrial applications. The windings of the three
phases in a 3-Φ induction motor are placed 120° apart.
The modified full spectrum (30) is based on the motor current signals from any two
phase of a 3-Φ induction motor is at a phase of 120° or
-120° w.r.t An(t). The modified full spectrum plot gives an insight about whether the
rotor orbit frequency components are forward or backward in whirl, in relation to the

rotor spin direction. The process of computing modified full spectrum starts from
digitizing the current waveforms. The signals obtained from the current sensors
constitute two individual signals and thus there is one waveform from each channel.
Signal processing based on FFT is independently carried out for each waveform. But the
signal processing based on FFT.doesn't provide the relative phase correlation between
the two signals.
Modified full spectrum overcomes this drawback by retaining the directional information
of each frequency components. Modified full spectrum is obtained from the FFT of the
stator current signals acquired from two current sensors. The stator current signals from
two sensors are sampled and are put into as the input for FFT. Considering one stator
current signal as the base signal however; the other signal is split in two mutually
perpendicular components, with one of the components along the base signal. The other
component of the signal perpendicular to the base signal is considered as the complex
signal. Further, the positive and negative frequency component of.the FFT is separately
considered. The positive and negative frequency component is further split into direct
component and quadrature component. The direct component and quadrature of the two
signals are combined to form direct output and quadrature output respectively. The
direct and quadrature output is used to compute forward and backward amplitude and is
plotted with respect to frequency to get the modified spectrum plot. The right half of the
modified frequency plot represents the amplitude of the forward whirling frequency
component also known as positive frequency components of signals, whereas, the left

part represents backward whirling frequency component also known as negative
frequency component of the signals.
FIG. 3 is a flow chart representation of the steps involved in determining whirl direction
of a shaft (11) coupled with an induction motor (10) through motor current signature
analysis in accordance with one embodiment of the invention. The method includes
acquiring (100) a set of stator current signals from any two phase of a 3-Φ induction
motor with known phase difference. Noise reduction (120) for the acquired sample is
done to remove noise. In the next step modified full spectrum analysis (140) is done for
the acquired signals. Further, the modified full spectrum analysis (30) yields forward and
backward amplitude. In the final step the relative magnitude of the forward and
backward amplitude is compared to determine (160) the whirl direction. If the magnitude
of forward amplitude (12) is greater than backward amplitude (14) then the whirl is
considered to be forward whirl (16) else backward whirl (18).
FIG. 4 is a flowchart representing modified full spectrum analysis algorithm for motor
current signature considering a 3-0 induction motor. In first step, stator current signals
from any two phases of a 3-0 induction motor is acquired, with the phase difference
between the two signals being 120°. Further, in second step FFT of the stator current
signals are done to transform the time domain signals to frequency domain signals. In
third step, the FFT signals corresponding to positive and negative frequency are
separately mentioned for both the signals. Then the direct and quadrature component of

the FFT signals is separated. The direct and quadrature component of each signals is
clubbed together to form direct output and quadrature output corresponding to positive
and negative frequency. The forward amplitude and backward amplitude is determined
using the direct output and quadrature output.
Mathematical formulation for modified full spectrum analysis for a 3-0 induction motor
coupled with shaft can be given as, Let, *.<» and B^W be two current signals in time
domain with a known phase difference. Since, the number of phases in a 3-

Rω-, then the whirl is said to be forward whirl but if Rω+