COMPARATIVE STUDY BETWEEN VARIOUS
REDUCED MODEL AND MODAL ANALYSIS OF
VISCOELASTIC ROTORS
A thesis submitted to National Institute of Technology, Rourkela in partial fulfillment
for the degree of
Master of Technology
in
Mechanical Engineering
by
Yogesh Verma
212ME1286
Department of Mechanical Engineering
National Institute of Technology, Rourkela
Rourkela - 769008, Odisha, India.
June – 2014
COMPARATIVE STUDY BETWEEN VARIOUS
REDUCED MODEL AND MODAL ANALYSIS OF
VISCOELASTIC ROTORS
A thesis submitted to National Institute of Technology, Rourkela in partial fulfillment
for the degree of
Master of Technology
in
Mechanical Engineering
by
Yogesh Verma
212ME1286
Under the guidance of
Dr. H. Roy
Assistant Professor
Department of Mechanical Engineering
National Institute of Technology, Rourkela
Rourkela - 769008, Odisha, India.
June - 2014
National Institute of Technology, Rourkela
CERTIFICATE
This is to certify that the thesis entitled, “Comparative study between
various reduced model and modal analysis of viscoelastic rotors”,
which is submitted by Mr. Yogesh Verma in partial fulfillment of the requirement
for the award of degree of M.Tech in Mechanical Engineering to National Institute
of Technology, Rourkela is a record of candidate‟s own work carried out by him
under my supervision. The matter embodied in this thesis is original and has not been
used for the award of any other degree.
Date: 2/06/2014
Dr. H. Roy
Assistant Professor
Mechanical Engineering Department
Rourkela
i
ACKNOWLEDGEMENT
I am grateful to my supervisor Dr. Haraprasad Roy, whose valuable advice, interest
and patience made this work a truly rewarding experience on so many levels. I am
also thankful to my friends and colleagues for standing by me during the past difficult
times. Particularly, I am indebted to Mr. Saurabh Chandraker for his utterly selfless
help.
As for Rohit Kumar singh, Md.Abdul Hussain, Abhinav Khare, Sachin Sahu,
Dilshad Ahmad and Ranjan Kumar Behera: You were there for me when really
needed and I am yours forever.
Yogesh Verma
ii
ABSTRACT
The present work includes the study of dynamic characteristics on a flexible
rotor shaft system. This arises due to the internal material damping of rotor bearing
system, which produces a tangential force on a rotor and increasing with the rotor
spin speed. Due to these dynamic characteristics of rotor is influence which
destabilizes the rotor shaft system. Under this dynamic behaviour of rotor shaft
system is studied to get the dynamic nature of rotor shaft system. This can be
estimated in terms of Campbell diagram, modal damping factor, mode shape and
directional frequency response function. These plots are obtained by using the matlab
software by solving a eigenvalues problem.
finite element approach plays a significant role in modelling continuum
system after discretizing it into some finite number of element. More number of
elements or enhancing the mesh size give better accuracy in results. But discretizing
the system into infinite elements inherits the swelling size of system matrices.
Substantial increment of the system matrices sometime causes very high
mathematical complications and takes an unwanted computational time. Model
reduction is techniques for reducing the degree of freedom from the full system
model to produce a reduced model but its dynamic characteristics is maintain. System
equivalent reduction process, Improved reduced system, Guyan reduction are used to
reduce the large system of equation of motion to fewer degree of freedom. The full
system model also includes internal and external damping and gyroscopic effect.
Since it is not practical to measure all degree of freedom, so the model is reduced
using model reduction techniques. The reduced model is used to plot Campbell
diagram, unbalance response using matlab software and comparison is done with
original system to show its effectiveness.
iii
NOMENCLATURE
M
M
Translational mass matrix
[G]
Gyroscopic matrix
Bending stiffness matrix
T
Rotary inertia matrix
R
K
B
K C
Skew symmetric circulatory matrix
Bending stress
x
Strain in x-direction
x
Strain rate
Position vector of displaced centre
rotation
Bending moment in y-direction
R
M
yy
M
zz
Bending moment in z-direction
M
C
K
F
Q
X
u and v
E
q , xn
Global Mass matrix
Global Damping matrix
Global Stiffness matrix
Force vector
Nodal displacement vector
State vector
Displaced position of cross-section
Young modulus
Full degree of freedom
T
Transformation matrix
Co-ordinate of full and master degree of
freedom
Stiffness matrices at state1 and state2.
n
q ,q
1
2
K ,K
M , K ,C
1
2
n
n
Mass, stiffness and damping matrix with
full degree of freedom
Master and slave degrees of freedom
Master- master degree of freedom
Master-slave degree of freedom
slave-master degree of freedom
slave- slave degree of freedom
Co-ordinate of master degree of freedom
n
m ,s
mm
ms
sm
ss
x
x
T
m
Co-ordinate of slave degree of freedom
s
Transformation matrix of Guyan
reduction
Transformation matrix of IRS
s
T
T
X
m
X
s
i
Transformation matrix of Serep
u
Selected master degree of freedom in
Eigen vector
Selected slave degree of freedom in
Eigen vector
iv
P
A
f
,
y
f
Modal participation vector
Mode of interest
Forces in y and z direction
z
u ,v
,l
Eigen values
Right hand and left hand Eigen vector
Right hand and left hand Eigen vector
Kronecker delta
ir
r
Modal force vector
Coefficient of viscous damping
v
M , K ,C
r
r
Rotor spin speed
Mass, stiffness and damping matrix
r
v
Content
Description
Page no
Certificate
i
Acknowledgement
ii
Abstract
iii
Nomenclature
iv-v
List of figures
vii
List of tables
viii
Chapter 1
1-8
1.1 Background and importance
1-2
1.2 Various terminology related to this study
3
1.2.1 Material damping
3
1.2.2 Internal damping
3
1.2.3 Modal damping factor
3
1.2.4 Resonant frequency
4
1.2.5 Frequency Response Function(FRF)
4
1.2.6 Directional frequency response Function (dFRF)
4
1.3 Modal Analysis
4-5
1.3.1 Modal Analysis approaches
5
1.3.2 Literature survey on modal analysis
5-7
1.4 Model Reduction
7
1.5 objective of the thesis
7-8
1.6 outline of the present thesis
8
Chapter 2
9-21
2.1 Finite Element Formulation
9-11
2.2 General Reduction Procedure
11-13
2.2.1 Guyan Reduction
13-14
2.2.2 Improved Reduced System
14-15
2.2.3 System equivalent reduction expansion process
15-16
2.2.3.1 Main point to be remember using Serep
17
2.3 Numerical Problem
17
2.3.1 Mass unbalance response
18
2.3.1.1 Comparison between full system, IRS and Guyan Reduction
18
2.3.1.2 Comparison between full system, Serep and Guyan Reduction
19
2.3.2 Campbell Diagram
19-21
Chapter 3
22-29
3.1 Modal Analysis in Rotor
22-24
3.2 Numerical Problem
24
3.2.1 Campbell Diagram
24-25
3.2.2 Modal damping factor
25-26
3.2.3 Three Dimensional mode shape for simply supported rotor system 26-27
3.2.4 Frequency response function (FRF) and Directional frequency
27-29
response function (dFRF)
Chapter 4
30-36
4.1 Conclusions
30
4.2 Future scope
31
4.3 Reference
32-34
4.4 Appendix
35-36
vi
List of Figures
Figure
Title
Page no
Figure 2.1
Displaced position of the shaft cross-section
10
Figure 2.2
Schematic Diagram of the Rotor with three discs
17
Figure 2.3
Comparison of Unbalance Response between full
system, IRS and Guyan Reduction
18
Figure 2.4
Comparison of Unbalance Response between full
system, SEREP and Guyan Reduction
19
Figure 2.5
Campbell diagram with full system
20
Figure 2.6(a)
Campbell diagram showing Full system and Guyan
reduction model
21
Figure 2.6(b)
Campbell diagram showing Full system and Serep model
21
Figure 3.1
Schematic Diagram of the Rotor with one discs
24
Figure 3.2
Natural frequency Vs. Excitation frequency
25
Figure 3.3(a)
Modal damping factor Vs. rotor spin speed (without
26
support damping)
Figure 3.3(b)
Modal damping factor Vs. rotor spin speed (with support
damping)
26
Figure 3.4(a)
First Backward whirl
27
Figure 3.4(b)
First Forward whirl
27
Figure 3.4(c)
Second Backward whirl
27
Figure 3.4(d)
Second Forward whirl
27
Figure 3.5(a)
Frequency Response Function (Hyy)
28
Figure 3.5(b)
Frequency Response Function (Hyz)
28
Figure 3.6
Directional Frequency Response Function (Hpg)
28
Figure 3.7
Directional frequency response function (Hpg) at node 6
29
Figure 3.8
Directional frequency response function (Hpg) at node 4
29
vii
List of Tables
Table
Title
Page no
Table 2.1
Material data
17
Table 2.2
Bearing data
17
Table 2.3
Disc data
17
Table 2.4
Comparison of Eigen values between full 20-21
system, Guyan Reduction, IRS, Serep
reduction techniques
Table 3.1
Material Properties
24
Table 3.2
Disc Parameter
24
Table 3.3
Bearing data
24
viii
Chapter 1
INTRODUCTION
1.1 Background and importance
Rotor dynamics is a branch of system dynamics deals with mechanical devices in
which at least one part is defined as rotor, which rotates with some angular
momentum. Following the ISO definition a rotor is a body suspended through a set of
cylindrical hinges or bearings that allows it to rotate freely about an axis fixed in
space. It deals with behaviour of high speed rotary machines which extending from
very large systems like steam power plant rotors for example, a turbo generator, to
very minor systems like enigma machine, with a variety of rotors is used in
centrifuge, steam turbine, motors etc. Genta [1]. The principal component of rotordynamic system is the shaft or rotor with bearing and disk. The shaft or rotor is the
rotating component of the system rotating at higher speed in order in order to
maximize the power output. Rotors with bearing support to restrain their spin axis in
one or other rigid way to a fixed position in space which are usually known as fixed
rotor, whereas the rotor which is not constrain is defined as free rotor. The parts of
machine which do not rotate are usually defined as stator. Rotors are the main causes
of vibration in most of the rotating machineries. At higher speeds of rotor, vibrations
caused by the mass unbalance results in some severe problems Rao [2]. So it is
necessary to decrease or to minimize these vibrations for operational protection and
stability. It can be ended by suitable assessment of dynamics of system. The dynamic
behaviour of a mechanical system must be examined in its design phase so that you
can determine whether it will present a satisfactory
Performance or not in its condition of the planned operation. The natural frequencies,
damping factors and vibration modes of these systems can be determined
analytically, numerically or experimentally.
The simplest model of rotor is used to study the dynamics behaviour of rotor consists
of a point mass connected to a massless shaft. It dynamics behaviour was carefully
studied by Jeffcott [3], it is often referred to as Jeffcott rotor.
Unlike the viscoelastic structures (which do not spin) viscoelastic rotors are acted
upon by rotating damping force generated by the internal material damping, which
tends to disrupt the rotor shaft bearing system by generating a tangential force which
1
is proportional to the rotor spin speed. Thus a reliable model is required which
consider constitutive relationship of a rotor material by taking into account the
internal material damping for understanding a dynamic behaviour of viscoelastic
rotor. Gunter‟s [4, 5] work on internal material damping. It‟s work give the idea of
destabilize of rotor due to internal material damping (viscous) as the rotor spin speed
increases. Dimentberg [6] work included both viscous and hysteretic internal
damping, hysteretic internal damping destabilize the rotor at all speed.
Modal analysis is based on the severe mathematical treatment to convert multidegree
of freedom into single degree of freedom. Therefore understanding of modal
behaviour of rotor modal analysis is necessary by which we came to know about the
importance of directivity and mode shapes with the help of modal matrices like
modal mass, modal stiffness and modal damping.
Due to the higher demand of improving the performance of high speed rotating
machinery the influence of rotor dynamics is increased. Increased power output
through the use of high speed, more flexible rotor has been increased the need, at all
stage of design. The acceptable performance of a turbo machine depends on the
adequate design and operation of the rotor supporting a bearing. Turbo machine also
include other mechanical element which provide stiffness and damping characteristic
affect the dynamics of the rotor and shaft system.
The rotor dynamics of turbo machine comprises of structural analysis of rotor (shaft
and disks) and design of bearing that determine the best dynamic performance under
given operating condition. Rotor dynamic instabilities have become common as the
speed and power of high speed turbo machinery is increased. Sometime, these
instabilities resulting in increasing vibration amplitude. So it is essential to minimize
the vibration for the operational protection and stability. The purpose of modal
analysis is to get an idea about the dynamic behaviour of the system. By the use of
modal analysis we find out critical and stability limit speed from this we can limit
this vibration or minimize it.
Model reduction is necessary for the higher order system having large degree of
freedom by this techniques we reduce the original system in to a reduce model by
different techniques, with a reduce model we deal with a fewer degree of freedom as
compared to the original system by preserving all its dynamic characteristics in a
reduced model.
2
1.2 Various terminology related to this study
1.2.1 Material damping
Existence of damping in the linear system makes it viscoelastic. Viscoelasticity is a
property of material that combines both elasticity and viscosity. These materials store
the energy as well as dissipate it under dynamic deformation. Thus, the stress in such
materials is not in phase with the strain. Due to these properties, it is extensively used
in various high speed machinery applications for controlling the amplitude of
resonance vibration in the system and adjusting wave attenuation and increasing
mechanical life through reduction in mechanical fatigue Dutt and Nakra [19].
Some characteristics of viscoelastic materials are:
[1] Creep.
[2] Relaxation.
[3] The actual stiffness depends on the amount of application of loading.
[4] If repeated load is applied, hysteresis occurs.
There are two type of internal damping hysteretic and viscous form of internal
damping, only viscous internal damping is considered and used to drive the equation
of motion. This type of damping produces a tangential force on rotor and destabilizes
it as the rotor spin speed increases. Under this condition dynamic performance of
rotor system is studied to get the dynamics characteristics.
1.2.2 Internal Damping:
Damping is due to the rotating and non-rotating parts of the structure. Damping
associated with non-rotating parts, called as external damping, has stabilizing effect
on the system. And damping associated with rotating part result in instability in
supercritical range. Due to rotation of rotors rotary damping arise, which increase as
the spin speed is increased and act tangential to the rotor orbit. Due to this instability
occur in the system. Therefore, a reliable model is required to represent the rotor
internal material damping for exact estimate of stability limit of spin speed (SLS) of a
rotor shaft system.
1.2.3 Modal Damping Factor:
By the modal analysis we find out the modal damping factor. It is the plot between
Modal damping factor and rotor spin speed. It has the incremental value for backward
whirl with respect to spin speed and has decrementing nature for forward whirl and
3
after certain spin speed becomes negative. Positive modal damping describe stability
of the system as the vibrational energy is dissipates and negative modal damping
indicates the instability as rotating energy support rotor spin on addition of energy.
1.2.4 Resonant frequency:
Resonance is simply the natural frequency of a component. All the structure has a
resonance frequency. Resonance problem occur in two primary ways. Critical speed
occurs when a component rotates at its own natural frequency. Structural resonance
occurs when some forcing frequency comes close to the resonant frequency of a
structure. Structural vibration problem, it‟s necessary to identify the resonant
frequencies of a structure. Nowadays, modal analysis has become a common means
of finding out modes of vibration of machines and structures.
1.2.5 Frequency response function (FRF):
It is a region in frequency domain where negative region is same as the positive
region so the positive region is consider for the physical significance when we plot
FRF.
1.2.6 Directional frequency response function (dFRF):
The traditional modal analysis for stationary structure is also applied for rotating
machines. But this analysis requires a theoretical concept. Due to the rotation,
gyroscopic effect appears in the system which result in non- symmetric matrices in
equation of motion and, as a affect the frequency response function does not obey the
Maxwell‟s reciprocity theorem. In any FRF plot, the negative frequency region is
identical of the positive frequency region. Therefore, it is necessary to deal with only
one region of FRF merely a positive one which has some physical meaning. Thus the
directivity of modes, backward and forward mode is not distinguishable in frequency
domain. Therefore, a new complex modal testing is suggested by Lee [26], for modal
parameter identification of rotary machines because by use of traditional method
forward and backward mode is not identified. The new method uses the complex
variable as an input and output source. But by use of complex modal testing separates
forward and backward mode in frequency domain by plotting a directional frequency
response function (dFRF) and no other testing is required.
4
1.3 Modal Analysis:
Modal analysis is the study of the dynamic characteristics of structure under vibration
excitation. Modal analysis of rotor-shaft systems has become very popular know a
days because this shows both spatial and temporal behaviours of the system in
dynamic condition. Modal analysis include rotating forces in the model to achieve a
more accurate modal behaviour of rotor shaft system and also used to see the
influence of damping forces on the mode shape and frequency response function.
Modal analysis involve both theoretical and experimental approaches He and Fu [20].
1.3.1 Modal Analysis approaches:
The modes of a system can be obtained from two very different approaches.
Theoretical modal Analysis
Theoretical Modal analysis is also known as mathematical models means “discretize”
a structure by breaking it up into different parts. This process can be done by using
the finite element (FE) approach. The analytical program then solves for an
eigenvalue problem to get the frequency, mode shape of each mode. The solutions of
the eigenvalue problems provide the modal data for the system.
Experimental modal Analysis:
This type of modal analysis, extracts the modes of vibration directly from FRF‟s
without having to make any assumption about the mass and stiffness distribution and
without solving any eigenvalue problem. The stability and the response level of
machines like aero-planes and steam turbine, cars are predicted by analytical model,
must be validated experimentally.
This type of modal analysis comprises three steps: Test planning, Frequency response
measurement and Modal parameter identification.
1. The first step involves the selection of structural support, type of excitation force
and location of excitation force, data acquisition system to measure force. Structure is
break up into different part. Accelerometer is connected at the selected nodes.
2. The impact force is applied at that location where we want to obtain FRF matrix
using the exciter and corresponding response are noted using the data acquisition
system. After that FRF matrix is analyze to identify modal parameters of the tested
structure.
5
1.3.2 Literature survey on Modal Analysis:
Rotor shaft system is subjected to circulatory force and rotating force originates from
several sources. Damping force of the shaft material, riveted joint in built up rotor,
force generated by shrink fitted rotor assemblies and fluid film forces are crucial to
be considering while dealing with rotor shaft bearing system. Tondl [21], as all these
forces are tend to destabilizing the rotor shaft system above the spin speed limit.
Therefore, the damping has been analysed by examine two types of model viz.
viscous damping model and hysteretic damping model. Genta [1], has been reported
modelling techniques for hysteretic form of material damping. Dutt [18] has reported
equation of motion for a rotor-shaft system by considering linear viscoelastic model
to represent the shaft material damping. After that effect of both form of damping
model and the hysteretic damping model has been studied by many researchers on
modal frequencies and modal damping Zorzi and Nelson [22], and Ku [23]
considered the combined effect of internal viscous damping, hysteretic damping and
shear deformation in the analysis. And result of forward and backward whirl speed
are presented and compared it with previous papers. Better convergence of the result
and high accuracy of the finite element model is presented with numerical example.
Modal analysis is based upon severe mathematical treatment, the significance of the
mode shapes and directivity of the modes. There are two methods for modal testing
classical and complex modal testing. The classical method is widely used for modal
parameter identification of structure of all kinds, except rotary machines, limited
attempt [24, 25] have been prepared to develop the modal testing method for rotary
machines. The complex method is proposed for modal parameter identification of
rotating machines. This method uses the complex notation which is a dominant
mathematical tool used in this analysis which not only allow perfect physical
understanding of forward and backward modes, but also help to separate these modes
in the frequency domain. The complex method is first developed by the Lee [ 26 ].
Kim and Kessler [27], proposed complex variables to describe planar motions, which
is directly relates physical motions to mathematical expression, is fully utilized in the
proposed procedure for complex variable based rotor analysis. In which the equation
of motion is formulated for the free vibration solution of equation of motion is
defined the directional natural modes because it not only describe the frequency and
the shapes but also direction of free vibrations response. For the forced vibration
6
solution directional frequency response function is obtained which clearly allows the
understanding of unique characteristics of rotor vibration.
Mesquita [28], in any frequency response function plot the negative frequency region
of FRF is same as the positive frequency region so it is necessary to treat only one
region merely a positive one because it have some physical meaning. Thus the
directivity of modes, backward or forward cannot be distinguished with the use of
traditional modal analysis. The complex modal analysis is used to distinguish the
directivity of modes in frequency domain.
Chouksey [29], has reported the equation of motion with material damping in shaft is
consider because both stationary and rotary damping forces in shaft system play an
important roles in deciding the dynamic behaviour. Therefore, rotating damping
forces originating due to material damping in shaft is consider and aims is to obtain a
more complete model and a more accurate modal analysis is obtained for rotor shaft
system. The effect of damping forces on the directional frequency response has been
obtained.
1.4 Model Reduction:
Model reduction is a technique to reduce a large finite element system to one with
fewer degrees of freedom while maintaining its vibrant features of the system.
Methods such as Guyan reduction, Improved Reduced System approach and the
System Equivalent Reduction Expansion process may be used for undamped and
non-rotating structure.
O‟Callahan et. al [10] suggested a improved method which is known as the Improved
Reduced System (IRS) method. In this method an additional term is added to the
Guyan reduction transformation to take some effect of the inertia terms. But it
depends on the Guyan reduced model.
O‟Callahan et. al [15] suggested other model reduction techniques known as system
equivalent reduction expansion process for undamped system depend on the arbitrary
selection of mode of interest.
Friswell et. al[12,13], used these reduction techniques for damped and rotating
structures, compare these techniques for damped and undamped structures and
discusses the errors introducing by using approaches based on the undamped model.
Das and Dutt [18], uses an improved System Equivalent Reduction Expansion
Process (SEREP) are used to reduce huge linear system of equation of motion. In this
7
equation of motion gyroscopic effect, internal material damping is included. This
techniques is applied on the state space model. And nice plot of Campbell diagram,
and unbalance response is plotted for reduced and original system show effectiveness
of the reduced model.
1.5 Objective of the thesis:
Based on that work, the aim and scope proposed in this work are as follow:
1. The equation of motion and finite element formulation of the viscoelastic rotor are
derived. Euler Bernouli beam theory is used for Finite element formulation to
discritize the rotor continuum. Two element voigt model is used to incorporate the
internal damping of the rotor shaft.
2. Effect of modal damping factor on rotor spin speed is analyzed. The mode shape
are found using eigenvector. Further dFRF plot is obtained to explore the direction of
whirl.
3. By model reduction techniques the full system model is reduced to fewer degrees
of freedom. A MATLAB code is generated for the model reduction techniques and
after that Campbell diagram and unbalance response is plotted and compared with the
original full system model.
4. Comparison between Guyan Reduction, Improved Reduced System, and System
Equivalent Reduction Expansion process is done.
1.6 Outline of the present thesis:
In present thesis chapter 2, equation of motion for damped rotor is written after
discretizing it into beam finite element. Model reduction techniques are explained in
detail like Guyan reduction, Improved reduced system and System equivalent
reduction expansion system. A comparative study of that technique is done from
Unbalance response and Campbell diagram with the help of Matlab software.
In chapter 3, Modal analysis in rotor is done to get the idea of dynamic characteristics
of rotor. An example is taken to identify various modal parameters like Modal
damping factor, 3-D mode shape, frequency response function and directional
frequency response function.
In chapter 4, some salient conclusions and related future scopes for model reduction
and modal analysis are given.
8
Chapter 2
A reduced model for rotor shaft system
In this chapter the mathematical modelling of viscoelastic rotor shaft is presented,
where external damping from viscoelastic support is considered. The system matrix
like mass matrix, stiffness matrix, gyroscopic matrix are obtain through finite
element formulation. Euler Bernouli beam theory is used in this purpose. The finite
element model is further used to develop reduced model. The transformation matrix
for various reduced model is derived. Finally an example is taken for comparison of
numerical results for different reduced model.
2.1 Finite element formulation
Finite element approach is used to model the rotor shaft system. First equation of
motion is derived from the constitutive relation. The dynamic longitudinal stress and
strain induced in the infinite area are
and
x
x
and
x respectively.
The formulation of
x
at an instant of time are given by Zorzi and Nelson [22].
x E v ;
x r cos t
R ( x, t )
2
x
2
(2.1)
Figure 2.1 display the moved position of the shaft cross section (u and v )describe the
displacement of the shaft centre along Y and Z direction and an element is consider
of differential radial thickness of
dr
at a distance r (where r varies from 0 to ro ) and
subtend an angle of d (t ) where is the spin speed in rad/sec and t lies from 0 to
2 at any instant of time ' t ' . Due to transverse vibration of the shaft is under two
types of rotation at the same time, i.e., spin and whirl.
shaft centre when it is at rest.
9
is the whirl speed. O is the
z
d t
dr
t
r
u
r0
R
v
t
y
Figure 2.1 Displaced position of the shaft cross- section
Following Zorzi and Nelson [22] the bending moments at any instant about the y and
z-axes are given as:
2 ro
M zz u r cos t x rdrd (t )
0 0
2 ro
M yy v r sin t x rdrd (t )
0 0
(2.2)
Put equation 2.1 into equation 2.2 and following zorzi and Nelson[22], the governing
differential equation for one shaft element is given as:
M M q K G q
K K q f
T
R
v
B
v
B
c
(2.3)
In the preceding equation M T 88 , M R 88 , G 88 , K B 88 and KC 88 are the
translational mass matrix, rotary inertia matrix, gyroscopic matrix, bending stiffness
matrix and skew symmetric circulatory matrix, respectively. The expressions for these
matrices are given below. The full matrices are given in the Appendix.
where subscripts in the element describe the respective planes.
The equation of motion for full system is obtained by assembled the elements matrix
into global matrix and it is written in the form as:
M q C q K q f
(2.4)
10
Where M , C and K are the global mass, damping and stiffness matrices,
respectively and
f
is the external force applied. Their expressions are written as:
M M M ; C K G ; K K K
T
v
R
B
B
v
c
The disc mass is added to the global mass matrix at a respective node. The global
damping matrix contains the gyroscopic effects of shaft and disc, and effects of
rotating and non-rotating damping.
Equation (2.4) once again is added by an identity equation to get the states space
equation.
A X B X Q
(2.5)
Where,
C
A M
M
K
, B
0
0
0
M
q
0
X , Q
q
f
Free vibration Equation of motion (2.4) for an eigenvalue problem can be written as
by assuming, u e
t
y
A X B X 0
(2.6)
is a complex Eigen values which represent the imaginary part and the real part
indicates the natural frequency.
2.2 General Reduction Procedure
Model reduction is usually used to reduce large analytical model to develop a more
effective model for further analytical studies, which is done by discarding few
coordinates. Generally total degree of freedom or number of coordinate is classified
in two categories, i) master coordinate and ii) slave coordinate. Master coordinate is
also known as active or measured or retained coordinate and slave coordinate is also
known as deleted or discarded or omitted coordinate. There are few methods for
selecting those coordinates.
11
a) Slave coordinates, whose inertia force are insignificant compared to the elastic
force. Thus it should be selected where inertia is small and stiffness is high.
b) Master coordinates, where inertia is high and stiffness are small.
c) Diagonal term in the ratio of stiffness and mass matrices,
K
M
coordinates. If
jj
K
M
jj
, for the
j
th
jj
is very small then there exist major inertia effects and associated
jj
coordinate is master coordinate.
d) If the diagonal term in the ratio
K
M
jj
is large then the
j
th
coordinates should be
jj
selected as a slave coordinate.
e) Another method to choose master and slave is that all the translational degree of
freedom is chosen as master co-ordinates and all rotational co-ordinates are chosen as
slave coordinate.
Mn, Kn,Cn are the full set of degrees of freedom and written in the form after the
selection of master and slave co-ordinate.
M
K
n
C
n
n
M mm
M sm
K mm
K sm
C mm
C sm
M
M
K
K
C
C
ss
ms
ss
ms
ss
ms
In general the relationship between the full set of analytical model and the reduced
set of master degree of freedom as
q x T x
n
m
xs
m
(2.7)
Subscript 'n' represent the full set of analytical degree of freedom, 'm' represent the
master degree of freedom and‟s‟ denotes the slave degree of freedom. „T‟ represents
12
transformation matrix between these two set of degree of freedom, which depends on
the different reduction techniques used and will discuss in the following section.
Therefore, the reduced mass, stiffness and damping matrices are obtained by pre- and
post multiplying the transformation matrix „T‟ to the full set of degree of freedom
matrices of ( M n , K n , C n ).
M r
K r
C r
T M T
T
n
T K T
T
n
T C T
T
n
Where the size of the reduced matrices is ( r r ).
2.2.1 Guyan Reduction
In Guyan reduction [7], the stiffness and mass matrices, are divided into separate
quantities which relates the master and slave degrees of freedom. Assuming that no
force is applied to the slave degrees of freedom and the damping is not considered,
then the equation of motion becomes from equation (2.4)
M mm
M sm
M
M
x m
K mm
xs
K sm
ss
ms
K
K
ms
ss
f
x m
xs
0
(2.8)
Neglect the inertia term
K mm
K sm
K
K
ms
ss
f
xm
xs
0
K x K x
sm
xs
m
ss
s
K sm x m
xm
K ss
(2.9)
0
(2.10)
K ss
1
K sm x m
f
1
K mm K ms K K sm
ss
13
f
1
K K sm
K
mm K ms
ss
x m
1
xn
xs
K ss K sm f
1
K K sm
K
mm K ms
ss
I
1
x n
K ss K sm
x = T x
m
(2.11)
m
s
T s is the transformation matrix for Guyan reduction.
2.2.2 Improved Reduced System
O‟Callahan [10] improves the Guyan reduction method by developing a new
technique known as the Improved Reduced System (IRS) method. It is an extension
of the Guyan reduction in which some additional effect of slave degree of freedom
and inertia term is considered which causes distortion in the Guyan reduction
techniques. The development is based on the circumstance that the static mechanical
model containing distributed forces can be reduced.
For sinusoidal excitation, equation (2.8 is written as
K ss
M x
2
ss
s
K sm
2
M sm x m
(2.12)
1
2
x s K ss 2M ss K sm M sm x m
By using the binomial theorem
x
K ss K sm
1
s
Where, O
M K K
1
2
ss
ss
sm
x
4
M sm O
m
is an error of order . The main aim is to improve the natural
4
4
frequency from the reduced model is based on the Guyan reduction, to first order in
2
.
M x K x
2
R
R
m
m
or
x M K x
2
m
1
R
R
m
x s K ss K sm K ss M sm M ss K ss K sm M R K R x m
1
1
1
1
T i T s SM T s M R K R
1
(2.13)
(2.14)
n
14
Where
1
t s K ss K sm from Guyan Reduction
I
T s
t s
0
S
0
1
K ss
0
M R and K R are reduced mass and stiffness matrix taken from Guyan reduction.
Reduced mass and stiffness matrices by Improved Reduction system are
T
M R T i M n T i
T
K R T i K n T i
T
C R T C n T i
i
T i is the transformation matrix for IRS.
2.2.3 System equivalent reduction expansion process
In Serep [15] reduction process there is a relationship between the master degree of
freedom and the slave degree of freedom which can be written in general form as
T
xs
x x
n
m
x
m
(2.15)
The modal transformation can be written as:
X m
p
x s X s
x x
n
m
(2.16)
The modal matrix is obtained from Eigen vector and is partitioned into the 'm' active
and‟s‟ slave or deleted set of degrees of freedom. The relationship for the active or
master set of degrees of freedom is.
x X p
m
m
(2.17)
p is the modal participation vector obtained from least square solution.
15
The inverse specification of above equation contains a generalized inverse then the
number of unknowns is not equal to the number of equations to be solved. There are
two probable Solution.
1. When the number of equation „m‟ is greater than or „a‟ equal to the mode of
interest.
2. When the number of equation „m‟ is fewer than the number of solution variables
„a‟ means mode of interest.
Least Squares Solution – m a (mode of interest).
X m x X m X P
T
T
m
m
X m x X m X m X m X
1
T
X m X m
p
X
T
m
X m
T
m
1
T
T
1
m
p
X m x X m x
T
g
m
m
(2.18)
X m is also called pseudo-inverse .
g
Average solution- when m a .
p X m X m X m
T
x X
n
n
T
1
x X m x
g
m
m
(2.19)
X m x T x
g
m
u
m
(2.20)
The Serep transformation matrix T u is used for the reduction of full original
system. Serep is heavily relies on a “well developed” finite element model from
which an „n‟ dimensional Eigen solution of the problem are obtained for developing
the mapping between the full set of n DOF and the reduced model of m DOF. The
quality of the result obtained from most reduction techniques depend on the chosen
of active degree of freedom, however it is not a concern when we use Serep
techniques. In Serep techniques an arbitrary selection of active or master degree of
freedom as well as an randomly selection of modes does not affect the natural
frequencies which are conserved in the reduced system when using Serep technique.
16
2.2.3.1 Main point to be remember using System Equivalent
reduction expansion process.
1. When the number of mode of interest used is less then active degree of freedom in
reduced system model (m>a). The size of the reduced matrices are „m‟ by „m‟, then
rank of the reduced system model is only ‟a‟. Hence, the reduced stiffness and mass
matrices are rank deficient; therefore precaution must be taken for these reduced
matrices like mass, stiffness and damping matrix. Due to this rank deficiency result.
2. The Serep process produce an exact solution when active degree of freedom is
equal to the mode of interest means (m=a).
2.3 Numerical problem
A rotor bearing system with three disk on the shaft and discretised into 13 beam
element having same length and cross-sectional radius of shaft is 0.05m. It is
supported with two orthotropic bearing. The shaft and disk material is steel and
unbalance mass 200gm is situated on disk2.
y
L
C
A
1
kyy
2
D
3
4
5
E
6
7
8
9
10
11
12
13
B
14
kyy
dyy
x
dyy
z
Figure 2.2: Schematic Diagram of the Rotor
Table 2.1: Material property and shaft data
Material
Density
(Kg/m^3)
Young
Modulus
(GPa)
Length
Diameter
Mild steel
7800
200
1.3
0.1
Damping
Coefficient
(N-s/m)
0
Table 2.2: Bearing data
Bearing properties
Stiffness
Damping
Plane xx
5e7
5e2
Plane zz
7e7
7e2
17
Table 2.3: Disc data
Disk
Disk 1
Disk 2
Disk 3
Inner radius (m)
0.05
0.05
0.05
Outer radius (m)
0.12
0.2
0.2
Thickness (m)
0.05
0.05
0.06
2.3.1 Mass unbalance response
The mass unbalance of 200gm is situated at disk 2 at node 6. The response amplitude
is plotted for three different reduced model viz. Guyan Reduction, System Equivalent
Reduction Expansion Process, Improved Reduced System. Before comparing these
techniques global matrices is divided into two parts master and slave coordinate and
then applying different techniques to plot unbalance response and effectiveness of
these techniques is noted by compared it with original plot.
2.3.1.1 Comparison between Full system, IRS and Guyan Reduction
The full system have 56 degree of freedom by applying reduction process system is
reduced to 24 degree of freedom. And unbalance response is plotted from Guyan
reduction and Improved reduced system techniques and compare it with original
system with 56 degree of freedom. We noted from the figure 2.3 Guyan reduction,
IRS and full system plot of unbalance response show the effectiveness of three
techniques. It is seen from figure 2.3 the Guyan reduction is very close to the original
system but its highest peak is not coinciding with the original system. Same nature is
observed for IRS also when compared with the full system. In IRS techniques some
inertia term is included to get the effect of mass.
18
Figure 2.3: Comparison of unbalance response between full system, IRS and Guyan
Reduction
2.3.1.2 Comparison between Full system, Serep and Guyan
Reduction
In figure 2.4 Serep process shows effectiveness than the Guyan reduction because we
see from the plot highest peak is same as the original one and after that it is same as
the Guyan reduction. In Serep process we reduce the state space equation with the
help of eigenvectors, and arbitrary selection of master co-ordinate and arbitrary
selection of mode of interest does not affect its accuracy when compared with the
original system. So, from the comparison between three techniques we see that Serep
is close to the actual or original full system model
19
Figure 2.4: Comparison of unbalance response between full system, SEREP and
Guyan Reduction
2.3.2 Campbell Diagram
Campbell diagram is plot between imaginary part of eigenvalues or whirl line (WL)
vs. spin speed. The Campbell diagram is used to find out the critical speed (Ωcr). A
line having inclination of 45 ° is known as synchronous whirl line (SWL).
Intersection between SWL and WL indicates the critical speed. The Campbell
diagram for full system drawn in figure 2.5 with 6 natural frequencies is considered.
It shows the forward and backward whirl. The full system having 56 degree of
freedom is reduced to 24 degree of freedom by applying different techniques like
Guyan, IRS, and Serep reduction. The Campbell diagram is plotted from the reduced
model and compares it with the full system. Table 2.4 shows the comparison of
eigenvalues for various techniques.
20
Figure 2.5: Campbell diagram with full system
Table 2.4: Comparison of eigenvalues between full systems, Guyan, IRS, Serep
reduction techniques
Mode
Full system
Guyan
IRS
Serep
reduction
1
55
53.9
53.7
53.8
2
68
69.6
69.5
69.5
3
157
151.6
150.3
150.3
4
197
205.1
204.5
204.5
5
238
235.4
225.9
225.9
6
415
435.8
424.8
424.7
7
456
459.6
454.5
454.5
8
616
682.4
598.0
597.8
9
738
801.6
797.0
796.7
10
1130
1196.8
1067.4
1067.2
11
1144
1223.4
1188.5
1187.8
12
1494
1602.2
1454.2
1453.9
From the figure 2.6(a) and figure 2.6(b) we see that the Campbell diagram for
reduced system is close to the full system. The natural frequencies obtained from full
system are close enough as compared with Guyan reduction for lower modes. IRS
21
and Serep techniques have same values of natural frequencies for lower modes as
well as for higher modes also and close to the full system model. But Guyan gives
better result than IRS and SEREP in lower mode.
Figure 2.6(a): Campbell diagram for Full Figure 2.6(b): Campbell diagram for
system and Guyan reduction model
Full system and Serep model
22
Chapter 3
Modal Analysis
In this chapter modal analysis of rotor is done because it is an important
mathematical tool to get the idea of dynamic behavior and modal identification
parameter of the system for example the mode shape, modal damping factor and
Campbell diagram, frequency response function, directional frequency response
function.
3.1 Modal Analysis in Rotor
The equation of motion of rotor from equation (2.4) with internal material damping is
considered and once again equation is written as following Mesquita [28].
y
y
y f y
M
C
K
z
z
z f z
(3.1)
The equation of motion can be written in state space form are
A X B X Q
(3.2)
Where,
C
A M
M
K
, B
0
0
0
,
M
q
X ,
q
y
f
0
Q , f y , q
f
f z
z
The matrices [A] and [B] are real, non symmetrical, and indefinite in general, causing
a non-self-adjoint eigenvalue problem. The eigenvalue problems related with
equation (3.2) are
A B 0 and A B l 0
T
T
(3.3)
Eigenvalues of the above problem are the same. The eigenvalues and eigenvector
appear in complex conjugate pairs. The eigenvectors of the Eigen problems (3.2) are
the vectors known as right and left eigenvectors, and given as
23
u
v
and l
u
v
(3.4)
The vectors {u} and {v} are the eigenvectors of the Eigen problems. The right and
left eigenvectors may be biorthonormalised as
l A r ir
T
li B r r ir
T
(3.5)
To uncouple the equation (3.2), the following coordinate transformation is done
4N
w r r
(3.6)
r 1
lr Q
r
j r
T
r rr lr Q ;
T
(3.7)
Substituting the response in (3.7) into equation (3.6) leads to
l
w
Q
j
T
4N
r
r 1
u v
q
F
j
T
4N
r
or
r
r
r 1
r
( 3.8)
r
Then
T
Y 4 N ur vr Fy
Fy
H
Z
r 1 j r Fz
F
z
(3.9)
Thus we can define the frequency response function matrix as
ur vr 2 N ur vr u r v r
H
r 1 j r
r 1 j r
j r
T
4N
T
T
(3.10)
2N
uir vkr
uir vkr
uir vir
H ik
j r
r 1 j r
r 1 j r
4N
H yy
H
H zy
H yz
H zz
(3.11)
Therefore, Directional Frequency Response Function or Complex Frequency
Response Function (dFRF) by Lee [26]
H
pg
1
2
H
yy
H zz i H yz H zy
24
1
H
2
1
H
2
H pgˆ 2 H yy H zz i H yz H zy
1
H pg
ˆ
H
ˆˆ
pg
yy
yy
H zz i H yz H zy
H zz i H yz H zy
(3.12)
From above equation we can conclude as:
H
ˆ
pg
(i ) H pgˆ (i );
H
ˆˆ
pg
(i ) H pg (i )
3.2 Numerical Problem
A rotor shaft system with flexible supports at its ends and having one offset circular
disc, as shown in figure 3.1. In all the considerations, bearing anisotropy and cross
coupled stiffness is not considered.
y
L
C
A
1
kyy
2
D
3
4
5
E
6
7
8
9
10
11
12
13
14
B
kyy
dyy
x
dyy
z
Figure 3.1 Schematic diagram of rotor
Material
Mild steel
Table 3.1: Material properties and shaft parameters
Density
Young
Length
Diameter
Damping
(Kg/m^3)
Modulus
Coefficient
(GPa)
(N-s/m)
7800
Disc
1
Bearing properties
200
1.4
0.1
0.0002
Table 3.2: Disc Parameter
Diameter (m)
Thickness (m)
0.40
0.05
Table 3.3: Bearing data
Stiffness
Damping
Plane yy
1.75e7
7e2
Plane zz
1.75e7
7e2
K yy K zz 1.75e7 N / m, Cyy Czz 700 N sec/ m, v 0.0002
25
3.2.1 Campbell Diagram
The Campbell diagram for damped system is plotted in figure 3.2 with four natural
frequencies which give first forward and first backward for first mode and also for
second mode. At first forward and first backward at 8949rpm and 9729rpm at which
the system is at resonant.
Figure 3.2: Campbell diagram
3.2.2 Modal Damping factor
The modal damping factor of two consecutive modes is plotted for undamped and
damped system in figure 3.3(a) and 3.3(b) respectively. In 3.3(a) it is a straight line
because no internal or external damping is considered. In figure 3.3(b), backward
whirl has incremental nature with spin speed and decremental nature for forward
whirl. The positive modal damping factor indicates stability and negative modal
damping factor indicates instability because vibrational energy is dissipates and
rotational energy support rotor whirl due to the addition of energy. Thus the system
may become unstable due to forward whirl.
26
0.05
0.04
modal damping factor (zeta)
0.03
0.02
0.01
0
-0.01
-0.02
-0.03
-0.04
-0.05
0
0.5
1
1.5
rotor spin speed(rpm)
2
2.5
4
x 10
Figure 3.3(a): Modal damping factor
Figure 3.3(b): Modal damping factor Vs.
Vs. rotor spin speed (without support
rotor spin speed (with support damping)
damping)
3.2.3 Three dimensional mode shape for simply supported rotor
system
Mode shape is plotted for simply supported rotor in which eigenvector is used to plot
these modes. The two consecutive modes for forward and backward whirl are
presented here. The clockwise rotation is considered as backward whirl and counter
clockwise is reflected as forward whirl. The stating of the locus is marked with star
and the locus is left incomplete at the end to measure the direction of whirl. The
mode shape for damped rotor is unsymmetric due to the addition of skew symmetric
matrix in the equation of motion.
27
Figure 3.4(a): First Backward whirl
Figure 3.4(b): First Forward whirl
Figure 3.4(c): Second Backward whirl
Figure 3.4(d): Second Forward whirl
3.2.4 Frequency response function (FRF) and Directional Frequency
response function (dFRF).
From the figure 3.5(a) and 3.5(b), Hyy and Hyz we see that there are four modes but
no information of directivity of Frequency means forward and backward modes in
frequency response function (FRF). In figure 3.6, Hpg we see directly from the plot
backward modes, 152Hz and 55.2Hz, appears in the negative frequency zone and
forward modes, 55.6Hz and 157.6Hz, appears in the positive frequency zone. So, we
clearly noted that directional frequency response function (dFRF) has the capability
to separates backward and forward modes, which is mixed in FRF plot.
28
Figure3.5(a): Frequency Response
Figure3.5(b): Frequency Response
Function (Hyy)
Function (Hyz)
Figure 3.6: Directional Frequency Response Function (Hpg)
From the plot we see clearly that forward and backward are separated in their
frequency zone in figure 3.7. The backward modes are 154.4Hz and 55.2Hz appear in
the negative frequency zone and forward modes are 55.6Hz and 158.4Hz appear in
the positive frequency zone at node 6 with internal material damping is considered in
figure 3.8 backward modes, 144.8Hz and 55.2Hz appear in the negative frequency
zone and forward modes are 56.0Hz and 162.8Hz appear in the positive frequency
zone at node 4 with no internal material damping.
29
dFRF
with
internal
and
external dFRF with support damping at node 4
damping at node 6
Figure
3.7:
Directional
frequency Figure 3.8: Directional frequency response
response function (Hpg) at node 6
function (Hpg) at node 4
30
Chapter 4
Conclusions and Future scope
4.1 Conclusions
This work includes the model reduction and modal analysis of a rotor system with
simply supported at its end by considering the external damping and internal material
damping. For this Finite element formulation for the rotor shaft system is first
obtained. From that following conclusion can be made.
1. From the reduced system the nature of unbalance response is close agreement
as compared to the full system.
2. The reduction techniques like Guyan reduction produce the natural
frequencies close to the natural frequencies of the full system for lower
modes. At higher modes IRS and Serep produces the natural frequencies close
to the natural frequencies of the full system.
3. The reduction process like Serep is effective in reproducing that natural
frequencies of the full system whose mode is include in transformation
matrix.
4. From dFRF we separate forward and backward modes in a frequency zone
which is mixed in FRF plot.
5.
From the mode shape we can easily visualized the forward and backward
modes.
6. From the modal damping factor we can conclude that the positive modal
damping factor indicates the stable zone and negative modal damping factor
indicate the unstable zone.
31
4.2 Future Scope
Following are the point for the future research in this area is
1. No one model reduction techniques is fully exact to the full system model so
search is continue for suitable second order model as well as for higher order
model.
2. Modal analysis for higher order model of viscoelastic rotor and also the work
is extended to nonlinear problem.
3. The modal analysis is also done for unsymmetric rotor for symmetric rotor it
is straight forward but unsymmetric rotor required complex modal analysis of
rotor shaft system.
32
4.3 REFERENCES:
1) Genta G.,”Dynamics of Rotating Systems”, Mechanical Engineering Series,
Frederick F. LingSeries Editor, (2006).
2) Rao J. S., “Rotor Dynamics”, New Age International Publishers, (1996).
3) Jeffcott H., “The Lateral Vibration of Loaded Shaft in Neighbourhood of a
Whirling Speed - The Effect of Want of balance”, Phil. Mag., vol. 37(6), pp.
304-314, (1919).
4) Gunter E.J.,” Dynamic stability of Rotor Bearing Systems,” NASA SP-113,
(1996).
5) Gunter
E.J.,”
Rotor-Bearing
Stability,”
Proceeding
of
the
First
Turbomachinery Symposium, pp.119-141 (1972).
6) Dimentberg F.M., “Flexural Vibrations of Rotating Shafts”, Butterworths,
London, (1961).
7) Guyan R.J., “Reduction of stiffness and mass matrices”. AIAA Journal, 3(2),
380, (1965).
8) Subbiah. R., Bhat R.B and Sankar T.S., “Dynamic response of rotors using
modal reduction techniques”, Journal of vibration, Acoustics, stress, and
Reliability in Design, Vol.111, pp.360-365, (1989).
9) Genta G and Delprete C., “Acceleration through critical speeds of an
anisotropic, non-linear, torsionally stiff rotor with many degree of freedom”,
Journal of Sound and Vibration, Vol.180, No.3, pp.369-386, (1995).
10) Callahan J.C.O., “A procedure for an improved reduced system (IRS)
model”, Proceedings of the Seventh International Modal Analysis
Conference, Lasvegas, pp.17-21, (1989).
11) Gordis J.H., “An analysis of the improved reduced system (IRS) Model
reduction procedure. Modal Analysis”: The international Journal of
Analytical and Experimental Modal Analysis 9(4), 269-285, (1994).
12) Friswell M.I., Garvey, Penny S.D, J.E.T., “Model reduction using dynamic
and iterated IRS techniques. Journal of Sound and Vibration”, 186(2), 311323, (1995).
13) Friswell M.I., Garvey, Penny S.D, J.E.T, “The convergence of the iterated
IRS method, Journal of Sound and Vibration” 211(1), 123-132, (1998)
33
14) Kammer D.C., “Test-analysis-model development using exact model
reduction. The International Journal of Analytical and Experimental Modal
Analysis” 2(4), 174-179, (1987).
15) Callahan J.C.O, Avitabile P, and Riemer R., “System Equivalent Reduction
Expansion Process (SEREP)”, Proceedings of the seventh International Modal
Analysis Conference, Las Vegas, pp.29-37, (1989).
16) Friswell M.I., Inman, D.J., “Reduced-order models of structures with
Viscoelastic components”, AIAA Journal 37(10), 1328-1325, (1999).
17) Friswell M.I., Penny, J.E.T., Garvey, S.D, “Model reduction for structures
with damping and gyroscopic effects”, In: Proceedings of ISMA-25 Leuven,
Belgium, September 2000, pp.1151-1158, (2000).
18) Das and Dutt J.K.,”Reduced model rotor-shaft system using modified
SEREP”, Journal of Mechanics Research Communications, 35, 398-407,
(2008).
19) Dutt J.K. and Nakra B.C., “Stability of Rotor Systems with Viscoelastic
Supports,” Journal of Sound and Vibration Vol.153 (1), pp.89-96, (1992).
20) He and Fu., “Modal Analysis”, A member of the Reed Elsevier group,
( 2001).
21) Tondl A., “Some Problems of Rotor Dynamics”, Chapman and Hall Ltd.,
England, (1965).
22) Zorzi E.S. and Nelson H.D., “Finite element simulation of Rotor-Bearing
systems with Internal Damping”, Journal of Engineering for Power,
Transactions of the ASME, vol.99, pp.71-76, (1977).
23) Ku D.M., “Finite element analysis of whirl speed for Rotor-Bearing systems
with Internal Damping”, Mechanical systems and signal Processing, Vol.12,
pp.599-610, (1998).
24) Lee C.W., Katz R, Ulsoy A.G and Scott R.A, “Modal analysis of a
distributed parameter rotating shaft” Journal of Sound and Vibration 122,
119-130, (1998).
25) Jei Y.G and Lee C.W., “Vibrations of anisotropic rotor-bearing systems”,
Twelfth Biennial ASME Conference on Mechanical Vibration and Noise,
Montreal, 89-96, (1989).
26) Lee C.W., “A complex modal testing theory for rotating machinery”,
Mechanical Systems and Signal Processing 5(2), 119-137, (1991).
34
27) Kim J and Kessler C.,” Vibration Analysis of Rotors Utilizing Implicit
Directional Information of Complex Variable Descriptions”, Transaction of
the ASME, Vol.124, pp.340-349, ( 2002).
28) Alexandre L. A. Mesquita, Milton Dias Jr., and Ubatan A. Miranda.,” A
Comparison between the traditional frequency response function (FRF) and
the directional frequency response function (dFRF) in Rotor dynamics
analysis”,Mecania Computational Vol.21, pp.2227-2246, Argentina, (2002).
29) Chouksey M., Dutt J. K and Modak S.V.,” Modal analysis of rotor-shaft
system under the influence of rotor-shaft material damping and fluid film
forces”, Mechanism and Machine Theory, vol.48, pp. 81–93, (2012).
35
4.4 Appendix
22l 0
0
54 13l
0
0
156
2
2
4l
0
0
13l 3l
0
0
156 22l 0
0
54
13l
2
2
4
0
0
13
l
3
Al
l
l
M T
156 22l
0
0
420
2
4l
0
0
Symmetric
156 22l
2
4
l
M
36
3l
0
0 3l 3l
0
0
2
2
4l
0
0 3l l
0
0
36 3l 0
0 36 3l
2
2
4l
0
0
3l l
I D
R
36 3l
0
0
36l
2
4l
0
0
Symmetric
36
3l
2
4
l
0
0 36 3l
0
2
0 3l 4l
0
0
0
36
2 I D
0
3l
G
36l
0
Skew symmetric
36
0
0
3l
4l
0
0
2
36 3l
2
3l l
0
0
0
0
36 3l
2
3l 4l
0
0
0
12
6l 0
0 12 6l
0
0
2
2
4l
0
0 6l 2l
0
0
12 6l
0
0 12 6l
2
2
4l
0
0
6l 2l
EI
K B 3
12 6l
0
0
l
2
4l
0
0
Symmetric
12
6l
2
4
l
0
0 12 6l
0
2
0 6l 4l
0
0
0
12
EI
0
6l
K c 3
0
l
Skew symmetric
37
0
12
0
6l
6l
0
2l
2
0
0
12
0
6l
0
6l
2
2l
0
0
6l
2
4l
0
0
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
advertisement
© 2021