Ijertv12n7 01

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp.

929-935
© International Research Publication House. http://www.irphouse.com

Balancing of Multi-Stage Pump Using the Coupling Hub


Karamoko Diarrassouba1 and Marco Cammalleri2
Ph.D, Department of Engineering, University of Palermo, Italy.
1

2 Associate Professor, Department of Engineering, University of Palermo, Italy.


1
ORCID: 0000-0003-2129-8179 2ORCID: 0000-0002-4958-6482

Abstract The goal of the present study is to propose a method for


temporary on-site balancing of this kind of pump, in order to
In this paper, we present an investigation to perform a
bring back the vibration magnitude below the alarm threshold
temporary on-site balancing of machines which rotor is not
in a little bit time, waiting for the scheduled general review of
accessible, such as multi-stage oil centrifugal pumps, in order
the machine.
to bring back the vibration magnitude below the alarm
threshold, waiting for the scheduled general review of the In the first part of this paper, we describe the problem and the
machine. theoretical approach for its resolution. The key assumption of
the proposed method is that, in presence of particular
We have shown that if the vibration measured on the Driver
conditions, the coupling hub fitted to the pump (CHP) is used
End Support (DE) is very higher than that measured on the Non
as a single balancing plane of a multi stage centrifugal pump
Driver End Support (NDE), we can consider the Coupling Hub
(MSCP). In the second part, we perform the experimental
fitted to the Pump (CHP) a correction plane in order to apply
verifications on a multistage centrifugal pump of an industrial
on site balancing by means of the Influence Coefficients
plant by use of the Influence Coefficients Method (ICM),
Method. This permits to overcome the problem of the
which is recognized for its efficiency and for its ease of
inaccessibility of the rotor and to solve, in the shortest possible
implementation in industrial context. The pump considered in
time, the unbalance occurred during the working. We have
this study is used in the process of heavy petrochemical
experimentally verified this hypothesis in a multistage pump of
products. Indeed, the product in question is a residue (vacuum
a petroleum industrial plant by two different approaches. The
residue) of a first stage of refining crude oil, which is reused for
first performs Single Plane Balancing (SPB) and vibration
feeding a second phase. The product is very viscous and is fed
measurements on two supports, acquired with eddy probe
at a temperature of 306 °C. For local unfavourable conditions,
sensors. The second approach, which may be more simple and
a dirt adheres to rotor impellers causing a phenomenon of
easy to apply, aims to apply the SPB using vibration
unbalance of the rotor with consequent unacceptable levels of
measurements on only one support acquired with an
vibration. The pump and the measurement instruments are
accelerometer. The results of the different approaches have
descripted in details and the balancing process based on the
been compared.
coupling hub is shown. We have used the ICM with two
Keywords: Coupling Hub, Influence Coefficients Method, On- approaches. The first performs Single Plane Balancing (SPB)
site balancing, Vibration alarm threshold. by means of vibration measurements on two supports; they are
acquired with fixed eddy probe sensors. The second approach
aims to apply the SPB using vibration measurements on only
I. INTRODUCTION one support, acquired with a removable accelerometer. The
latter has the advantage to be relatively simpler and more easily
The rotating machinery vibration problems remain a concern, applicable whereas many machines, for cost reason, are not
despite progress in recent years in their design. Wear, equipped with fixed sensors. The results of the different
deformations, shifts of the assembled elements, deposition of approaches have been compared.
material (dirk) on the rotor cause a change in the state of their
equilibrium, which induces a deterioration of vibration levels
[1-6]. The imbalances are an important cause of vibration.
II. THEORETICAL APPROACH
Particularly, the unbalance caused by deposition of material on
the rotor happens very often in the multi-stage centrifugal The shaft of a rotor can be considered, at same time, as rigid, if
pumps used in heavy residual oils refining process and, it is operating much below its first critical speed, or flexible,
generally, it represents the more important part of their total when it is operating near or above the first critical speed. An
unbalance. Normal balancing practices need the total shutdown eccentricity of the center of gravity of a rigid rotor causes a
of the machine, removal the rotor and balancing it on a static unbalance. On the contrary, a uniformly distributed
balancing machine. This operation is very expensive and unbalance along the length of a rigid rotor causes a couple
induces additional costs of missed production caused by the unbalance. The latter cannot be detected without to rotate the
long downtime, mostly if it is unscheduled. The literature shaft. The effect of the static and of the couple unbalance
review highlights the interest of designing balancing together is named dynamic unbalance, which is what occurs
approaches that reduce the number of test runs required and the generally in industrial machines. However, once the rotor
machines downtime so that to minimize the production loss approaches a critical speed, its centerline bends and whirls
generated in industrial plants [7-16]. around and new centrifugal forces set-up. In this case,

929
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

eccentricity and its angular orientation may change in three- them together into a larger matrix [C]. However, most of the
dimension continuously from one end of the shaft to another. machines has a single operating speed.
Basic principles of rigid and flexible rotor balancing are quite
different. Various types of practical balancing techniques have In order to cancel the vibrations of the rotor, a suitable system
been proposed [17-32]. One of the most representative of correction masses is needed such that it is
balancing method is the influence coefficient method (ICM),
which is largely employed in industrial sector. This method is
the more appropriate to perform on site-balancing because it 𝒘𝑖 + ∑𝑛𝑗=1 𝒄𝑖𝑗 𝑚c𝑗 = 0 (2)
uses only experimental information; indeed, it has the
advantage of not requiring a mathematical modelling of the and in matrix form
system, with the consequent uncertainties related to the
discretization of the system itself. Consequently, it can apply to
rotors assumed to be rigid or flexible. We summarize it in the {𝑾} + [𝐂]{𝑴𝐜 } = {𝟎} (3)
in the next section.
II.I ICM- Influence Coefficients Method However, if the measurement plans p are more than the
corrective ones n, the matrix [C] is rectangular, thus not
The machine is described by p measurement plans and n invertible and eq. (3) cannot be solved. The least square method
corrective plans (or balancing plans). This method involves the (or another equivalent) is needed to minimize the residual
system linearity assumption and it is based on the construction vibrations [33-37]. We used the method proposed in [34] which
of a matrix modelling the system, made of sensitivity to minimizes the quadratic norm of the residual vibration. In this
unbalance experimentally measured. The vibration case, the vector of the correction masses {𝑴𝒄 } is expressed by
measurements can be carried out either by arranging two sets the following equation (Eq.2):
of p transducers, in correspondence with the p measurement
planes, on two planes orthogonal to each other and containing
the rotation axis, or by arranging a single series of p coplanar {𝑴𝒄 } = −{[𝑪∗ ]𝒕 [𝑪]}−𝟏 [𝑪∗ ]𝒕 {𝑾} (4)
transducers and an indicator of phase.
For a given rotation speed, the coefficient 𝒄𝑖𝑗 means how where the symbol * is the conjugate.
much the vibration on the sensor i is influenced by the
imbalance on the plane j, i.e.
II.II ICM applied to on-site balancing of a multi stage
centrifugal pump
𝒘𝑖𝑗 −𝒘𝑖
𝒄𝑖𝑗 = (1) Fig.1 schematics the pump: it has 12 impellers divided into two
𝑚t𝑗
groups with opposing flows. This design aims primarily to
reduce the axial thrust. Two cylindrical hydrodynamic bearings
where: 𝒘𝑖 is the vector of initial vibration measured by the
bring the shaft. The red point in Fig. 1 represents the dirt of
sensor i in amplitude and phase, 𝒘𝑖𝑗 is the vector of vibration
mass m, which causes a change in the state of equilibrium of
measured by the same sensor i but after adding a trial mass 𝑚t𝑗 the rotating machine.
on the plane j. Each test finds one column of the matrix [C].
Then it is needed as many experimental tests, as they are the
imbalance plans. If a machine operates at different speeds, it is
necessary to construct a matrix for each speed, thus grouping

1. Support Driver End


2. Support Non Driver End
3. Shaft
4. Impellers (number 12)
5. Pump Hub
6. Mass m (dirt, cause of
unbalance)
7. Correction mass
8. Aspiration
9. Delivery

Figure1: Schematic representation of Multi-stage Centrifugal Pump unbalanced

930
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

Fig. 2 presents the schematic representation of the forces not have the desired effect but even worsen the situation.
involved in the pump balancing. F represents the unbalancing
force created by the mass m, while 𝑹𝑁𝐷𝐸 and 𝑹𝐷𝐸 are the
reaction forces on the supports to this unbalance. The rotor is
supposed to be balanced in absence of the mass m.
The reaction forces change in 𝑹′ 𝑁𝐷𝐸 and 𝑹′ 𝐷𝐸 if we add a
correction force 𝒁 ∝ 𝑚c𝑧 in opposite versus of F at the end of
the rotor (i.e. on the CHP). Indeed, applying Z creates a couple
unbalance, but it can reduce the static unbalance; our aim is to
define in which condition a proper balancing force Z can bring
the measured vibration magnitude on the supports below the
threshold alarm.

Figure 3: Trend of the reaction forces as a function of the


correction force if 𝑤𝐷𝐸 > 𝑤𝑁𝐷𝐸
𝑤𝐷𝐸
However, the occurrence > 2 is common in multi-stage
𝑤𝑁𝐷𝐸
pumps unbalances, so we have verified experimental the
possibility to use the CHP as unique correction plane. We
followed two different approaches by means of the ICM.
The first performs balancing by vibration measurements on
both the supports. In order to simplify the notation, in the
following, we indicate with 1 the mid plane of the support NDE,
with 2 the one of the support DE and with z the one of the CHP.
The matrix of influence coefficients [C] = [𝑐𝑐11 ] is rectangular
21
Figure 2: Forces involved in the pump balancing and his size is (2 × 1). Applying eq (4), the correction mass
𝑚c𝑧 is:

𝒄∗𝟏𝟏 𝒘𝟏 +𝒄∗𝟐𝟏 𝒘𝟐
The following equations give the reaction forces of the supports 𝑚c𝑧 = − (7)
𝒄∗𝟏𝟏 𝒄𝟏𝟏 +𝒄∗𝟐𝟏 𝒄𝟐𝟏
to the initial unbalance F and to the correction force Z:

𝐹𝑑2 𝑑3 The second approach aims to use vibration measurements only


𝑅′ 𝑁𝐷𝐸 = + 𝑍 (5) on the support DE, basing on the observation that 𝑅𝐷𝐸 is very
𝑙 𝑙

higher than 𝑅𝑁𝐷𝐸 . In this case, it is 𝑐21 =0 and the correction
𝐹𝑑1 (𝑙+𝑑3 ) mass will result simply
𝑅′𝐷𝐸 = − 𝑍 (6)
𝑙 𝑙
𝒘𝟏
𝑚c𝑧 = − (8)
F, 𝑑1 and 𝑑2 are unknown, but they are fixed data and linked to 𝒄𝟏𝟏
the state of unbalance; on the contrary, 𝑑3 and l are known from
𝑑
the machine design. In general, the ratio 3 ≈ 0.1, this means
𝑙 III. EXPERIMENTAL TESTS
that 𝑅′ 𝑁𝐷𝐸 grows slowly in function of Z, while 𝑅′ 𝐷𝐸 decreases
(𝑙+𝑑3 ) An electric three-phase motor, which rotation speed is 1492
faster (indeed, ≈ 1.1).
𝑙 rev/min, drives the pump via a speed multiplier (see Fig. 4),
If we now assume the vibration magnitude on a support is which has a transmission ratio of approximately 3.22.
proportional to the reaction force on the same support, then it Therefore, the pump rotation speed is equal to 4798 rev/min.
𝑤 𝑅 𝑑
is 𝐷𝐸 = 𝐷𝐸 = 1 , where wDE and wNDE are the vibration The shaft of the pump is connected to the gearbox by a Meta-
𝑤𝑁𝐷𝐸 𝑅𝑁𝐷𝐸 𝑑2
amplitudes due to the initial unbalance F. Therefore, if it is stream coupling (CHP) (see Fig.5). This hub is connected to the
𝑤𝐷𝐸 ≫ 𝑤𝑁𝐷𝐸 at first, then 𝑅′𝐷𝐸 will decrease with the increase coupling block by 10 bolts at regular distance and angle (36
of Z, and bring back the vibration magnitude in the normal on degrees).
the DE support, before that𝑅′ 𝑁𝐷𝐸 increase enough to cause The length of the steel shaft of the multistage pump is l=3297
vibration higher than the alarm threshold (see Fig.3). On the mm. The rotor weighs 218 kg. The pumped liquid is vacuum
contrary, if it is 𝑤𝐷𝐸 < 𝑤𝑁𝐷𝐸 , a corrective action on CHP could residue, its capacity 120 m3.

931
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

Multi-stage
CentrifugalPump
Gear box

Electrical
Motor

Figure 4: Global view of motor, gearbox and pump

This pump is monitoring online with eddy current probes which Table 1 reports the values of the pump allowable vibration,
measure shaft motion in axial direction z on the support NDE defined basing on historical data [39-40].
and in direction X’ and Y’ for each of the two supports as shown Fig.7 shows the vibration spectrum on DE support in the initial
in Fig.5 and Fig. 6. Other four eddy current sensors are installed conditions, measured by the eddy current probes. The peaks
on the gearbox supports in vertical direction. An addiction detected at the frequency corresponding to the rotation speed
accelerometer with magnetic base is fixed only on the pump are very high (61.69 µm in direction X’ and 72.78 µm in
support DE in the horizontal direction. A key phasor (laser direction Y’) than any other peak. This proves that the state of
device) is installed under the rotor to connect sensors signals the machine vibration is inacceptable and that the main cause
and phase. All the sensors are connected to a data collect SKF of high vibration is really due to the unbalance; we have
Microlog Analyzer GX. Displacements of the shaft are confirmed it also by the vibration measured in velocity with the
measured in micro with detection peak-to-peak, while the accelerometer.
signal from the accelerometer is acquired in velocity by time
integration. The acquisition of the measures has been done
simultaneously from all the sensors [38].

Figure 5 Pump Support DE and coupling hub CHP, devices


for vibrations and phase measurement. Figure 6 Pump Support NDE, devices for vibrations
measurement.

Table 1 Pump allowable vibration. In order to apply the balance with one-plane of balancing and
two planes of measurement, it is sufficient one point of
Allowable vibration acquisition for each support. Since we have a pump that
Normal 40 µm peak to peak equipped with four sensors (two for support), we profit to
compare the results of the two-way acquisition X’ and Y’.
Alarm 54 µm peak to peak
Tables 2 reports the results, independently for the directions X’
Shutdown 70 µm peak to peak and Y’: initial measurement, measurement with a trial mass and
with the related correction mass calculated by eq. (7). It is
worth to note that, in the initial conditions, it is 𝑤𝐷𝐸 = 𝑤1 ≫

932
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

𝑤2 = 𝑤𝑁𝐷𝐸 . On the contrary, Table 3 reports the results using


the vibration measurements of the support DE only, with the
addition of the accelerometer in the direction X.

Eddy current probe X' Eddy current probe Y'


70 70
micron Peak to Peak

micron Peak to Peak


60 60 Amplitude 72.78 mm; Frequency: 80 Hz
50 Amplitude 61.69 mm; Frequency: 80 Hz 50
40 40
30 Overall Trend: 85.29 mm; 30 Overall Trend: 85.55 mm;
20 Running speed: 4798 RPM 20 Running speed: 4798 RPM
10 Eddy current probe X’ 10 Eddy current probe X’
0 0
0 500 1000 1500 2000 0 500 1000 1500 2000
Frequency -Hz Frequency -Hz

Figure 6: Vibration Spectrum on support-DE showing the initial unbalance.

Table 2: Vibrations measured on both supports by eddy currents sensors.


Vibration measurement with trial Calculated correction
Initial vibration measurement
mass 𝑚𝑡𝑧 (10 g, θ=144°) mass
w1 w2
w1z w2z 𝑚c𝑧
(DE vibration) (NDE vibration)
Modulo Phase Modulo Phase Modulo Phase Modulo Phase Modulo Position
(µm) (deg) (µm) (deg) (µm) (deg) (µm) (deg) (g) (deg)
direction
61.69 128 13.72 308 31.45 129 10.5 308 20.64 145
X’
direction
72.11 218 12.01 38 36.97 218 15.55 38 19.98 145
Y’

Table3: Vibrations measured on support DE only by eddy currents sensors and accelerometer.
Vibration measurement Calculated correction mass
Vibration initial
with trial mass
measurement
𝑚𝑡𝑧 (10 g, θ=144°)
w1 w1z 𝑚c𝑧
Modulo Phase (deg) Modulo Phase (deg) Modulo (g) Position (deg)

Eddy current probes X’ 61.69 µm 128 31.45µm 129 20.39 145


Eddy current probes Y’ 72.11 µm 218 36.97µm 218 20.51 145
1.51
Accelerometer X 2.89 mm/s 83 83 20.93 145
mm/s

806 8
From Table 2 and Table 3, we can observe that the values of thus, it is 20.9𝑒 1000𝜋𝑗 = 20.43𝑒 10𝜋𝑗 + 0.62𝑒 𝜋𝑗 . Because 0.62
correction mass 𝑚c𝑧 calculated for the different approaches is negligible in respect of 20.43, as well as the difference
and different sensors are quite similar. Therefore, in order to between 145° and 144°, we chose to fix the correction mass to
balance the rotor, we considered the correction mass calculated 𝑚c𝑧 =20.9g at position 144°, so to use only one bolt as shown
with the second approach by the accelerometer vibration in Fig.7.
measurement (Table 3, 𝑚c𝑧 =20.9g, position 145°). Table 4 shows the residual vibrations, after the addition of the
The coupling hub is fitted to the pump shaft by 10 bolts at correction mass, measured on the support DE both by the
regular distance and angle (see Fig.5). Therefore, we thought to accelerometer and by the two eddy current sensors in directions
add the correction mass on the hub by increasing the weight of X’ and Y’.
two bolts closer, together with their screws. The bolts closer to
the calculated correction mass are at 144° (8/10π) and 180° (π),

933
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

REFERENCES
[1] McMillan RB (2004) Rotating machinery: practical
solutions to unbalance and misalignment. Fairmont
Press, Lilburn.
[2] Adams ML (2000) Rotating machinery vibration:
from analysis to troubleshooting, vol 131. CRC Press,
Boca Raton.
[3] Sorge, F., Cammalleri, M. On the beneficial effect of
rotor suspension anisotropy on viscous-dry hysteretic
instability. Meccanica, Springerlink, vol. 47, aprile
2012, ISSN 0025-6455, doi: 10.1007/s11012-012-
9549-y.
Figure 7: Addition of the correction mass (20.90g) to the [4] Goyal, D. & Pabla, B.S., The Vibration Monitoring
position 144 ° Methods and Signal Processing Techniques for
Structural Health Monitoring: A Review, Arch
Table 4: Residual vibration values Computat Methods Eng (2016) 23: 585.
measured after correction weight addition https://doi.org/10.1007/s11831-015-9145-0.
[5] M. Cammalleri, D. Rotella, Functional design of
Sensor Modulo wc1
power-split CVTs: an uncoupled hierarchical
Eddy Probe X’ 13.97 (µm) optimized model, Mech. Mach. Theory 116 (2017)
Eddy Probe Y’ 19.26 (µm) 294–309.
Accelerometer 0.51 ( mm/s) [6] W.C. Foiles, P. E. Allaire, and E. J. Gunter, Rotor
Balancing, Shock and Vibration, 1998, Volume 5,
Issue 5-6, pp. 325-336,
As it can be seen, the pump rotor balancing obtained by adding
http://dx.doi.org/10.1155/1998/648518.
a correction weight to the CHP has permitted to reduce
[7] Xiaofeng Li, Longxi Zheng, Zhenxia Liu First,
considerably the vibration level, which is turn to be normal.
Balancing of flexible rotors without trial weights
Obviously, we took care to verify that the added weight not
based on finite element modal analysis, Journal of
increase the pump axial vibration and not cause any problem to
Vibration and Control, Volume: 19 issue: 3, pp. 461-
the gearbox. Indeed, we have observed with the eddy probe
470, 2012,
sensors that the level of these vibrations have remained normal
https://doi.org/10.1177/1077546311433916.
even after the balancing.
[8] Cammalleri M., Costanza A., A Closed-Form
Solution for Natural Frequencies of Thin Walled
IV. CONCLUSION
Cylinders with Clamped Edges, International Journal
The main result of this study is having proved that it is of Mechanical Sciences, Volume 110, May 2016,
conceivable to perform the on-site balancing of a multi-stage Pages 116–126, Elsevier, Doi:
centrifugal pump, which ordinarily cannot be balanced on-site, 10.1016/j.ijmecsci.2016.03.005
because the rotor is accessible only with the disassembly of the [9] Ye, R., Wang, L., Hou, X. et al., Balancing method
pump. The set out approach allows to solve the unbalance in without trial weights for rotor systems based on
the shortest possible time, reducing the time required to an hour similitude scale model, Front. Mech. Eng. (2018) 13:
on average. It allows to bring the machine in acceptable 571. https://doi.org/10.1007/s11465-018-0478-x
reliability conditions, but it not replaces definitely the normal [10] P. Gnilka; Modal balancing of flexible rotors without
balancing, which is necessary in other unbalance cases, as the test runs: an experimental investigation, 1983, Journal
ones due to wear, deformations, shifts to the assembled of Vibration 90, pp. 157-172.
elements, or in the context of a general planned review. [11] D. Rotella, M. Cammalleri, Direct analysis of power-
The proposed balancing technique is performed using the split CVTs: A unified method, Mech. Mach. Theory
coupling hub fitted to the pump as a single balancing plane, so 121 (2018) 116–127.
overcoming the constrain of the inaccessibility of the pump [12] S. Edwards, A.W. Lees and M.I., Experimental
impellers. However, this is possible only when the initial Identification of Excitation and Support Parameters of
vibration on the driver end support is very higher than the non- a Flexible Rotor-Bearings-Foundation System from a
drive end support one. We have experimental verified this Single Run-Down, Friswell, 2000, Journal of Sound
approach on an industrial pump used in the process of heavy and Vibration, 232(5), pp. 963-992.
petrochemical products and we have obtained a good success. [13] Y. A. Khulief, M. A. Mohiuddin, and M. El-Gebeily,
We used the Influence Coefficients Method, both with two A New Method for Field-Balancing of High-Speed
measurement planes and with only one. The two approaches Flexible Rotors without Trial Weights, International
have provided nearly the same result. The second approach, Journal of Rotating Machinery, Volume 2014,
carried out by means of a removable accelerometer, has the http://dx.doi.org/10.1155/2014/603241.
advantage to be relatively more simple and more easily [14] Cammalleri. M., Sorge F., Approximate Closed-Form
applicable whereas many machines, for cost reason, are not Solutions for the Shift Mechanics of Rubber Belt
equipped with fixed sensors. Variators, Proceedings of the ASME International

934
International Journal of Engineering Research and Technology. ISSN 0974-3154, Volume 12, Number 7 (2019), pp. 929-935
© International Research Publication House. http://www.irphouse.com

Design Engineering Technical Conferences and rotor balancing through multiple critical speeds,
Computers and Information in Engineering ASME Journal of Engineering for Industry, 1976, pp.
Conference 2009, DETC2009. 6, pp. 187-196. 988 998.
[15] R. Tiwari, Conditioning of Regression Matrices for [28] Rotella D., Cammalleri M., Qin D., Zhou X. (2019) A
Simultaneous Estimation of the Residual Unbalance Simple Method for the Design of Hybrid Electric
and Bearing Dynamic Parameters; 2005, Mechanical Power-Split CVTs: A Case Study. In: Advances in
System and Signal Processing. Italian Mechanism Science. IFToMM ITALY 2018.
[16] D.J. Rodrigues a , A.R. Champneys a, M.I. Friswell , Mechanisms and Machine Science, vol 68. Springer.
R.E. Wilson; Experimental investigation of a single- [29] Sorge F., Cammalleri. M., Helical Shift Mechanics of
plane automatic balancing mechanism for a rigid rotor, Rubber V-Belt Variators, Journal of Mechanical
Journal of Sound and Vibration 330 (2011), pp. 385– Design, Volume 133, Issue 4, (2011). ISSN 1050-
403. 0472, doi: 10.1115/1.4003803.
[17] R.E.D. Bishop, G.M.L. Gladwell, The vibration and [30] Van De Vegte J., Balancing of Flexible Rotors during
balancing of an unbalanced rotor, Journal of Operation, Journal of Mechanical Engineering
Mechanical Engineering Science, 1959, pp. 66–67. Science, 1981, Vol. 23, N° 5, pp. 257-261.
[18] Bishop R. E. D., On the Possibility of Balancing [31] Cammalleri M., Efficiency of Split-Way CVT's. A
Rotating Flexible Shafts, Journal of Mechanical simplified model, SAE Technical Paper, vol. 2007-24-
Engineering Science, 1982, Vol. 24, N° 4, pp. 215-220. 0133, ISSN: 0148-7191, doi:10.4271/2007-24-0133.
[19] Rotella D., Cammalleri M., Power losses in power- [32] Lee J., Van Moorhem W. K., Analytical and
split CVTs: A fast black-box approximate method, Experimental Analysis of a Self-compensating
Mechanism and Machine Theory 128 (2018) 528–543, Dynamic Balancer in a Rotating Mechanism, ASME
https://doi.org/10.1016/j.mechmachtheory.2018.06.0 Journal of Dynamic Systems Measurement and
11. Control, 1996, Vol. 118, pp. 468-475.
[20] S.G. Tan, X.X Wang, A theoretical introduction to [33] Goodman T. P., A Least-Squares Method for
low speed balancing of flexible rotors: unification and Computing Balance Corrections, ASME Journal of
development of the modal balancing and influence Engineering for Industry, 1964, pp.273-279.
coefficients techniques, Journal of Sound and [34] Tessarzik, J. M. and Badgley, R. H., Experimental
Vibration 168, 1993, pp. 385–394. evaluation of the exact point-speed and least-squares
[21] M. Cammalleri, E. Pipitone, S. Beccari, G. Genchi, A procedures for flexible rotor balancing by the
Mathematical Model for the Prediction of the Injected influence coefficient method, ASME Journal of
Mass Diagram of a S.I. Engine Gas Injector, Journal Engineering for Industry, Series B, 1974, 96(2), pp.
of Mechanical Science and Technology, Springer vol. 633- 643.
11 (2013) pagg. 3253-3265, doi: 10.1007/s12206- [35] F. Sorge, M. Cammalleri, A Theoretical Approach to
013-0848-6. Pneumatic Muscle Mechanics IEEE/ASME
[22] Markert R., Modal Balancing of Flexible Rotors with International Conference on Advanced Intelligent
Data Acquisition from Non-stationary Run-up and Mechatronics (AIM) July 9-12, 2013 Wollongong,
Run-down, International Modal Analysis Conference, Australia, ISBN 978-1-4673-5318-2, doi:
6 th, Kissimmee (FL), feb. 1-4 1988, Vol. 1, pp. 210- 10.1109/AIM.2013.6584228
216. [36] Tanguy Messager, Mariusz Pyrz, Discrete
[23] Sorge F., Cammalleri. M. Control of hysteretic optimization of rigid rotor balancing, Journal of
instability in rotating machinery by elastic suspension Mechanical Science and Technology, 2013, Volume
systems subject to dry and viscous friction., Journal of 27, Issue 8, pp 2231–2236,
Sound and Vibration Elsevier, Vol. 329, Issue 10, https://doi.org/10.1007/s12206-013-0605-x.
Pages 1686-1701, (2010) [37] S. Beccari, E. Pipitone, M. Cammalleri, G. Genchi,
doi:10.1016/j.jsv.2009.12.007. Model-based optimization of injection strategies for
[24] Rotella D., Cammalleri M., Qin D., Zhou X. (2019) A SI engine gas injectors, Journal of Mechanical Science
Simple Method for the Design of Hybrid Electric and Technology, Springer, Volume 28, Issue 8,
Power-Split CVTs: A Case Study. Mechanisms and August 2014, Pages 3311-3323, doi: 10.1007/s12206-
Machine Science,Volume 68, 2019, Pages 70-79. 014-0742-x.
[25] J.M. Tessarzik, R.H. Badgley, W.J. Anderson, [38] John Harrell, Precision Balancing with the CMVA 55
Flexible rotor balancing by the exact point–speed Microlog, SKF Reliability Systems 4141 Ruffin Road
influence coefficient method, Journal of Engineering San Diego, California 92123 USA, CM3028, 199.
for Industry Series B ASME 94, 1974, pp. 148–158. [39] Berry, J.E. Illustrated Vibration Diagnostics Chart.
[26] E. Pipitone, S. Beccari, M. Cammalleri, G. Genchi, Technical Associates of Charlotte, P.C., 2007.
Experimental Model Based Linearization of a S.I. [40] ISO 13373-2: Condition monitoring and diagnostics
Engine Gas Injector Flow Chart, Journal of of machines - Vibration condition monitoring - Part 2:
Mechanical Engineering, Vol 60, No 11 (2014), doi: processing, presentation and analysis of vibration data.
10.5545/sv-jme.2013.1321.
[27] Tessarzik, J. M., Badgley, R. H. and Fleming, D. P.,
Experimental evaluation of multiplane-multispeed

935

You might also like