Proceedings in
MANUFACTURING
SYSTEMS
Proceedings in Manufacturing Systems, Vol. 5 (2010), No.
STUDY OF DYNAMIC BEHAVIOUR FOR A MOTOR SPINDLE BY EDDY CURRENT
MONITORING
Claudiu BISU, Andra LITA, Dorel ANANIA, Miron ZAPCIU, Constantin ISPAS, Constantin MINCIU
Abstract: Considering the severe conditions of productivity, quality, and wear a series of dynamic
analysis methods are developed for determining the dynamic behavior of machines and production
systems. The objective of this paper is to measure the displacements of a radial spindle of milling
machine tools using Eddy Current transducers and evaluated effects by signal processing and
analyzing and then determine the causes of these vibrations. To achieve analysis uses a number of
elements of computing and signal processing both waveform and frequency, and location to
identify any defects. To reduce the vibrations amplitudes of a circularity diagram on one or more
rotations is developed with the aim of intervention on the defect to extend the lifetime of the
vibration generating element.
Key words: vibrations, displacements, spindle motor, eddy-current.
1. INTRODUCTION
Appearance vibration is inevitable given the dynamic
milling process which is subject to system-piece
machine-tool in terms of productivity, quality and cost
more difficult. Vibrations analysis has long been used for
the detection and identification of the machine tool
condition. Predictive maintenance is directed towards
recognizing the earliest significant changes in machinery
condition [1]. Thin-walled work piece is greatly
influenced by the vibrations of the cutting process with
the risk strains wall piece. The machining of thin walls
generally generates milling chatter, that damage surface
roughness and manufacturing tools [2, 7].
Besides appearing as an important factor for
deterioration of vibration behavior of machines had to do
with displacements and deformations that appear while
the bearings due to the operation machine tools,
especially in dynamic conditions poor. Angular contact
ball bearings have been widely used in machine tool
spindles, and the bearing preload plays an important role
on the performance of the spindle. With the development
of high speed machining, for high speed milling, heavy
cutting at low speed and light cutting at a high speed are
often performed on a single machine tool spindle, thus,
high stiffness at low speed and low temperature rise at
high speed are required [6].
In this paper we made quantifying displacements and
gathering of data bearing dynamic effects in terms of
vibrations behavior of motor spindle by using eddycurrent transducer [3]. It will make a dynamic analysis of
frequencies measured in the stiffness influences the
quality of functioning of the motor shaft.
Vibration analysis is used to identify the causes or effects
of dynamic damaging the spindle motors or components,
especially the bearings.
2. OBJECTIVE PRESENTATION
A frequent problem in the manufacturing industry
today is the vibrations induced by metal cutting, turning,
milling and boring operations.
Vibrations in boring operations or internal turning
operations, for example, are inevitable and constitute a
major problem for the manufacturing industry.
Tool vibrations in metal cutting affect the result of the
machining, in particular for the surface finish.
Furthermore, tool life is correlated with the degree of
vibration and acoustic noise introduced [5].
Vibration and noise occurring in the operation of
machine-tools and those existing in the environment have
negative influence on safety in the operation of these
machines, the productivity and precision mechanical
processing and humans.
Noise and vibration are factors which are becoming
increasingly important to purchasers who might want a
quality product or have a requirement to minimize the
risk to health [4]. Vibrations from a piece of industrial
equipment can be not only a nuisance to employees, but
can also lead to future machine problems or failure.
Implementing a customized vibration control
treatment is a simple and affordable venture that can
improve equipment operation while contributing to a
more pleasant, safe workplace and improved employee
satisfaction. Specialized machinery and equipment
utilized in industrial environments often endure
inadvertent abuse caused by environmental vibrations.
A piece of machinery may produce vibrations that
affect its own stability, or may be affected by vibrations
transferred from other equipment. Regardless of the
source, persistent vibration can have a negative effect on
machine performance and create a noisy and distracting
Fig. 2. Tests device.
Fig. 1. Spindle motor presentation.
environment for personnel. Presented in Fig. 1 is a
spindle motor model, which identifies parts, especially
ceramic ball bearing. This paper aims at characterizing
the dynamics of the electric motor to determine the
causes of vibration and resolve their effects.
3. EXPERIMENTAL DEVICE
In order to make dynamic measurements, we want to
obtain the dynamic behavior of the main spindle while
moving, using displacement transducers without contact
"Eddy current" positioned at 2 mm from the master tool
(high accuracy balanced tool). Tests are made on a
milling machine (with HSK A63), 5-axis numerically
controlled with a motor of 34 kW and a range of speed
from 50 up to 13000 min-1. The speed of rotation and
phase angle are measured with a laser tachymeter. The
signal is acquired through an acquisition card and then
processed on a specialized program of vibrations and
signal processing. Tests were carried out starting from a
speed of 500 min-1 up to 12 000 min-1 level with 500 min1
. Signal acquisition was achieved through the acquisition
of signal Digitline Fastview.
An eddy current is a local electric current induced in
a conductive material by the magnetic field produced by
the active coil. This local electric current in turn induces
a magnetic field opposite in sense to the one from the
active coil and reduces the inductance in the coil. When
the distance between the target and the probe changes,
the impedance of the coil changes correspondingly. This
change in impedance can be detected by a carefully
arranged bridge circuit. The eddy currents are confined to
shallow depths near the conductive target surface. Their
effective depth is given by,

1
,
  f   
where f is the excitation frequency of the circuit, µ is the
magnetic permeability of the target material, and σ is the
conductivity of the target material. The target material
must be at least three times thicker than the effective
depth of the eddy currents to make the transducer
successful. This is because the transducer assumes that
the eddy currents are localized near the surface of a semiinfinite solid, and the actual eddy current amplitude
decreases quadratically with distance.
4. DYNAMIC RESULTS
Further, will be presented the vibrations
measurements made for a speed of 2000 min-1 and 12000
min-1. Comparing the results it can be seen that the
amplitudes for 2000 min-1 are about 0.023 mm (fig.3)
while for 12000 min-1 are global amplitudes (overall)
about 0,039 mm (fig.4). With the wave form signal, the
response in time of the displacements we can determine
the spectral response for a velocity of 2000 min-1 and
12000 min-1. After obtaining the FFT transformation
could find that at 2000 min-1 the main frequency
corresponding to the rotation speed is 33.33 Hz (fig.5),
where we have a peak of amplitude of 0,0012 mm. In
case of 12000 min-1 the displacements amplitude is about
4 times larger, having a fundamental frequency
corresponding to rotation speed of 200 Hz (Fig. 6).
Fig. 3. Amplitude measurement on 2 000 min-1
(1)
Fig.4. Amplitude measurement on 12000 min-1.
F1. Surname1 et al. / Proceedings in Manufacturing Systems, Vol. 5 (2010), No. 1 / 110 (8 pt, Italic)
3
Fig.5. Frequency spectrum for 2000 min-1.
Thus comparing the measured values that the
machine tool must have (max 0.0015 mm) to a proper
functioning, we note that 0.004 mm is very high.
Beyond, this magnitude develops the exact frequency of
order 1, which is the fundamental frequency. Following
the analysis phase to determine the angular position
vector of vibration remains constant with variations of
±2º.
Fig.6. Frequency spectrum for 12000 min-1.
As such it requires a frequency analysis of bearing
elements because they are prone to defects or
deformations. Figure 8 shows the configuration of a
bearing on the frequency calculations and identification
of fault frequencies.
5. EXPERIMENTAL ANALYSIS
The results obtained from measurements have
revealed a number of dynamic features of the car, that the
main shaft. As shown in Figure 5 and Figure 6 shows
significant vibration amplitudes increase with increasing
rotational speed which requires the assumption of the
existence of bearing excited faults. Therefore our
analysis will deal primarily by emphasizing dynamic
phenomena caused by bearing defects, dynamic
phenomena monitored by eddy current type transducer.
These sensors can provide information necessary to
evaluate real travel games or beatings but also for the
analysis of dynamic frequency spectra. Figure 7 shows
the trend of order 1 harmonic, obtained as a result of the
change from 2000 to 12.000 min-1. Thus we can
distinguish the difference amplitudes at maximum speed
and minimum speed and also top the 5600 min-1.
Our attention is directed to the peaks shown in Figure
7 thus indicating the importance of proper harmonic
vibration of order 1. This observation requires two
assumptions taking into account the existence of an
imbalance of the spindle (motor spindle) and the second
related to a strain off the end of the main shaft.
Fig.8. Frequency spectrum for 5600 min-1 and 1200 min-1
In formula 2, 3 and 4 presents the calculation of
frequencies runway, the outer ring and inner ring. In
figure 9 are given values of frequencies calculated for
rolling elements, cage, inner ring, outer ring and dynamic
frequencies [8]. Where Fc is the rotational frequency of
cage, Fe is frequency of outer inner and Fi is frequency of
intern inner, with Fa - the rotational frequency of spindle,
d - diameter of the ball, dm - medium diameter of the
bearings, α - contact angle and n - the number of balls. In
Figure 8 we see and what is the frequency peaks around
1500 Hz which lies between the harmonic frequencies,
leading to the hypothesis of a defect or a game revolving
working capital increased.
Fig.7. Trend of dynamic measurement
Fig.9. Bearing configuration, default frequency.
Fig.11. Circularity diagram of waveform analysis
Fig.10. Frequency calculation
Fc 
1
 Fa
2


d
 1 
 cos 
d
m


Fe  n  Fc
Fi  n  Fa  Fc 
(2)
(3)
(4)
From measurements made could find that the entire
range of speeds of rotation used this frequency of 1500
Hz (± 30 Hz) was found on all frequency spectrums, with
the remark that the amplitude is corresponding third
harmonic amplitude of order 1.
Following these frequencies on the graph in Figure 9 we
can see the validation of frequencies especially common
inner ring. Given the magnitude of this frequency is
much smaller magnitude than the fundamental frequency
amplitude we use dynamic analysis of existing vibration
fundamental frequency.
Based on measurements
acquired using eddy-current transducers and the signal
processing program circularity diagram was obtained
equivalent to a complete rotation. Such a rotation
corresponding vibrations are around 0.03 mm. Analysis
allowed determining a rotation angle where maximum
displacement is being compared to 270º and zero position
tachometer. By this analysis could highlight existing
defect on the motor spindle shaft and then following its
settlement. In the first phase of the game and taking
bearings tightening could improve the behavior of the
machine tool spindle motor shaft, reducing the
amplitudes to half of global travel.
5. CONCLUSIONS
This paper presents an experimental study of the
dynamic behavior for a CNC milling machine with 5
axes where using eddy current transducers were
measured the real displacements of the tool during
cutting process. This information is very important for
determining the vibrations source. The frequency
comparison made at 2000 min-1 and 12000 min-1 allowed
the validation of the imbalance vibrating effect caused by
the main shaft bearing wear, especially the front bearings
that are in the tool direction. In the perspective, an
analytical study will be conducted to determine the
influence of displacement measured at 12000 min-1 in the
goal of cutting processes. Finally solving the problem by
increasing the lifetime of the main shaft bearings before
changing.
This method allows verifying the machines-tools
performances from dynamic point of view and can be use
F1. Surname1 et al. / Proceedings in Manufacturing Systems, Vol. 5 (2010), No. 1 / 110 (8 pt, Italic)
to determine the state of different part of the machine
tool (like main spindle bearings) in the working
condition.
5. REFERENCES
[1] M. Zapciu, J.Y. K’Nevez, R. Laheurte, Ph. Darnis,
Dynamic characterisation and Predictive Maintenance
Concept of Machine Tool Spindle, ICSAAM- Structural
Analysis of Avanced Materials, 7-10 Sepetember 2009,
Tarbes
[2] T. Wehbe, G. Dessein, L. Arnaud, Experimental Study of
Thin Part Vibration Modes in Machining, Proceedings of
the 17th International Conference on Manufacturing
System – ICMAS, pages 247-252, 13-14 November 2008.
[3] Devillez A., Dudzinski D., Tool vibration detection with
eddy current sensors in machining process and
computation of stability lobes using fuzzy classifiers, In
Mechanical Systems and Signal Processing, S.G. Braun,
Volume 21, Issue 1, Pages 441-456, Elsevier.
[5] Guillem Quintana , Joaquim Ciurana and Daniel
Teixidor, A new experimental methodology for
identification of stability lobes diagram in milling
operations, vol.48, Issue 15, 1637-1645, 2008,
International Journal of Machine Tools &
Manufacture, Elsevier.
[6] Shuyun Jiang, Hebin Mao, Investigation of variable
optimum preload for machine tool spindle, vol.50,
19-28, 2010, International Journal of Machine Tools
& Manufacture, Elsevier.
[7] S. Khachan and F. Ismail, Machining chatter
simulation in multi-axis milling using graphical
method, vol.47, Issue 9, 1455-1473, 2007,
International Journal of Machine Tools &
Manufacture, Elsevier.
[8] www.skf.ro
5