Master's Degree Thesis

ISRN: BTH-AMT-EX--2007/D-07--SE

Output only Modal Analysis

Test Structure

Scaled
Mode Shapes

Ejaz Yousaf

Department of Mechanical Engineering

Blekinge Institute of Technology
Karlskrona, Sweden
2007
Supervisor: Kjell Ahlin, Professor Mech. Eng.
 Output only modal Analysis

Ejaz Yousaf
Department of Mechanical Engineering
Blekinge Institute of Technology
Karlskrona, Sweden
2007

Thesis submitted in fulfillment of the requirements for the award of the Master of Science
in Mechanical Engineering by the Department of Mechanical Engineering, Blekinge
Institute of Technology, Karlskrona, Sweden.

ABSTRACT

This thesis work aims at estimating the modal parameters of a system using a technique
amongst different proposed in Output only modal analysis. This technique assumes that
impulse response function can be approximated by cross correlation between responses
and transfer function by cross power spectral density of the responses. A method which
involves added masses on some points of structure under test is used to obtain a scaling
factor. This scaling factor is used for scaling the unscaled modal vectors from Output-
Only modal testing. A simulation on four degrees of freedom system in Matlab is
performed in order to check the validation of theory. Experiments are performed on a
structure with bolt channel joints. Modal identification tool box developed by Saven
Edutech AB in Matlab and a few Matlab built-in functions are used during the analysis of
the thesis work. All the structures under study in this thesis are modeled as linear time
invariant systems.

Keywords
Cross correlation function, Cross power spectral density, FRF, Impulse response
function, Scaling factor, Scaled mode shapes, ambient excitations
Acknowledgements

Some people are more respectful and nice than what they look to us. Their personality
seems far better than we describe in our words. There are a number of personalities, who
have helped and contributed in what I have achieved today and I am grateful to all of
them.

Saying the word “Thanks” to someone means paying back for what they have given to
you or helped you in achieving something, but I cannot pay back to Professor Kjell Ahlin
and my parents just by saying “Thanks” for their guidance, help and supporting me
throughout my career especially my Masters studies. I will remember their contributions,
guidance and help for the rest of my life. It would have been impossible for me to reach
up to this stage without their support and guidance. I am very grateful to all of them.

I would like to mention the names of Mr. Martin Magnevall and Andreas Josefsson (PhD.
Students) for their help, guidance and support throughout the course work and thesis.

I am also grateful to the whole department of Mechanical Engineering at Blekinge

Institute of Technology, Karlskrona and to all those, whose published data and materials
are used as references in this thesis work.

Karlskrona April, 2007

Ejaz Yousaf

2
Contents
1 NOTATIONS .................................................................................................................... 5

1.1 LIST OF SYMBOLS .............................................................................................................. 5

1.2 LIST OF OPERATORS ......................................................................................................... 6
1.3 LIST OF ABBREVIATIONS .................................................................................................. 7

2 INTRODUCTION ............................................................................................................ 8

2.1 BACKGROUND ................................................................................................................... 8

2.2 EXPERIMENTAL MODAL ANALYSIS ................................................................................. 8
2.3 STRUCTURAL MODIFICATION .........................................................................................10
2.4 OUTPUT ONLY MODAL ANALYSIS ..................................................................................11
2.4.1 NEED ..................................................................................................................................11
2.4.2 APPLICATIONS ...................................................................................................................11
2.4.3 PROCEDURE .......................................................................................................................12
2.4.4 THINGS TO REMEMBER ......................................................................................................15

3 OUTPUT ONLY SYSTEM IDENTIFICATION .........................................................16

3.1 SYSTEMS IDENTIFICATION OF CIVIL ENGINEERING STRUCTURES ..............................16

3.1.1 NON PARAMETRIC METHODS ............................................................................................17
3.1.2 PARAMETRIC METHODS .....................................................................................................17

4 PARAMETER IDENTIFICATION METHOD ...........................................................20

4.1 POWER SPECTRAL DENSITY ESTIMATION .....................................................................20

4.1.1 PERIODOGRAM METHOD ...................................................................................................20
4.1.2 CORRELATION METHOD ....................................................................................................22
4.2 MODAL DECOMPOSITION OF SPECTRA DENSITIES .......................................................22
4.3 TRUNCATION OF THE CROSS-SPECTRAL DENSITY MATRIX .........................................25
4.4 POLES AND RESIDUES ESTIMATION ................................................................................25
4.5 SELECTION OF REFERENCE RESPONSES ........................................................................26

5 NORMALIZATION OF MODE SHAPES ...................................................................27

5.1 DERIVATION .................................................................................................................27

5.1.1 SCALING FACTOR ...............................................................................................................30
5.2 THINGS TO REMEMBER ...................................................................................................34

6 SIMULATION.................................................................................................................35

6.1 DESCRIPTION OF SYSTEM ................................................................................................35

6.2 RESULTS AND COMPARISONS .........................................................................................37

3
7 APPLICATION TO A BOLT CHANNEL JOINT STRUCTURE.............................41

7.1 INTRODUCTION ................................................................................................................41

7.2 EQUIPMENT SETUP...........................................................................................................41
7.3 EXPERIMENTAL SETUP ....................................................................................................43
7.3.1 PROCEDURE .......................................................................................................................44
7.3.2 ESTIMATION OF MODE SHAPES .........................................................................................47
7.4 COMPARISON OF RESULTS ..............................................................................................48
7.5 DISCUSSIONS.....................................................................................................................53

CLOSING REMARKS ................................................................................................................56

REFERENCES .............................................................................................................................57

4
1 Notations

1.1 List of symbols

F(t) excitation force at time t

f Frequency [Hz]
fs Sampling frequency [Hz]

H (ω ) Frequency response function

I Identity matrix
j Imaginary unit j 2 = −1
k Discrete time instant t = k Δt
M, C, K Mass, damping and stiffness matrix
Ni Number of inputs

No Number of outputs

Nm Number of DOF

N ref Number of references

N it Number of iterations

Na Number of averages

Q Modal participation factor

R Residue matrix
Ri Output covariance matrix at time lags i

R xy Covariance between variables x and y

r Number of references
S (ω ) Power spectrum continuous time variable
u(t) Input at time t
uk Input at time k

V Mode shape matrix V = CΨ

5
x(t) State at time t
y(t) Output at time t
yk Output at time instant k

y kref Reference output at time instant k

ξ Damping ratio

ψ Operational mode shape

φ Normalized mode shape

α Operational scaling factor

Δm Local change in mass
Δt Sampling interval
δ (t ) Dirac delta

δk Kronecker delta

λi Continuous-time Eigen value

μi Discrete-time Eigen value

ω Angular frequency [rad/s]

1.2 List of operators

(⋅) T Matrix transpose

−1
(⋅) Matrix inverse
*
(⋅) Complex conjugate

(⋅) H Hermitian transpose: complex conjugate transpose of matrix

Re (⋅) Real part of

Im (⋅) Imaginary part of

6
1.3 List of abbreviations

AR Autoregressive
ARMA Autoregressive Moving Average
CMIF Complex Mode Indicator Function
DFT Discrete Fourier Transform
DOF Degree of Freedom
EMA Experimental Modal Analysis
EMT Experimental Modal Testing
FE Finite Element
FEM Finite Element Method
FFT Fast Fourier Transform
FRF Frequency Response Function
IFFT Inverse Fast Fourier Transform
LS Least Squares
LSCE Least-Square Complex Exponential
MAC Modal Assurance Criterion
ODS Operational Deflection Shape
OMA Operational Modal Analysis
OMT Output-only Modal Testing
PP Peak Picking
RMS Root Mean Square
SNR Signal to Noise Ratio
SSI-COV Covariance-driven Stochastic Subspace Identification
SSI-DATA DATA-driven Stochastic Subspace Identification
SVD Singular Value Decomposition
RD Random Decrement
RMITD Reference Ibrahim Time Domain
ITD Ibrahim Time Domain

7
2 Introduction

2.1 Background

Main target of Engineers has been to make the end product qualitative and cost effective
with short lead time. This phenomenon lead to improvement in the technological methods
involved in the testing and design phases of a product.

The dynamical systems vibrate during operation as the body does not remain in
equilibrium. If the frequency of excitation force matches with any resonance frequency of
the structure, it could vibrate with unexpectedly high amplitudes causing damage or
failure of the structure.
Modal Analysis is an important tool in vibration analysis of a structure. Objective of
modal analysis is the evaluation of modal parameters of a structure under dynamic loads.
Modal parameters provide information of protecting the structure from failure.
Resonance frequencies, damping ratios and mode shapes are the required modal
parameters of a dynamical system. Such frequencies at which structure vibrates violently
even at small amplitude excitation are known as resonance frequencies. Resonance
frequencies are the Eigen values of the equation of motion. Damping ratios decide how
quickly or slowly the structure can dissipate the vibration energy. The relative position of
points on a structure at a given natural frequency is known as mode shape. Mode shapes
are termed as Eigen vectors.

2.2 Experimental Modal Analysis

Finite element analysis FEA uses the Eigen value solvers to estimate the modal
parameters. When experiment is performed on real life structure, we divide the structure
into discrete points. But true information regarding mass, stiffness and damping matrices
is missing for the chosen points. In Experimental modal analysis, we excite the structure
(location depend on the selection of excitation) and measure the response at all the points.
Different methods are available in both time and frequency domain to evaluate the modal
parameters. Measurement data is usually used to calculate the impulse response and
frequency response functions. Modal parameters are extracted from these two functions.
A frequency response function is estimated by the equation (2.1)

GYX ( f )
H( f ) = (2.1)
G XX ( f )

8
Where,

G XX ( f ) is the auto power spectral density of the input signal

GYX ( f ) is the cross power spectral density of the input and output signal
H ( f ) is the frequency response function

Steps used in traditional experimental modal analysis are presented in the following block
diagram.

Figure 2.1 Block Diagram of Experimental Modal Analysis

9
2.3 Structural Modification

After agreement between experimental modal analysis and finite element analysis, our
modal model is developed. Model is simulated in the operating conditions. If model
operates safely under the operating conditions, it marks the end of the modal analysis
design phase. If model is not operating safely in the operating conditions, the structural
dynamic modification (SDM) is carried out. Modification is done until required
conditions are satisfied.

Experimental Modal Finite Element

Testing Modeling

Modal Parameter Eigen Solution

Estimation

Development of Modal
Model

Structural Changes No
required?

Yes
DONE

Use SDM

Figure 2.2 Design Phases of Modal Analysis and Structural Dynamic Modification

10
2.4 Output Only Modal Analysis

In traditional modal analysis, excitations and responses of the structure are measured to
extract the modal parameters. If designer could extract the modal parameters with as high
accuracy as with traditional modal analysis without measuring the input force, it will
decrease the project cost as well as analysis time. Output only modal analysis is a
technique in which we do not need to measure the input excitations and just responses are
measured. Assumption in output only modal analysis is that the excitation force is white
noise. Output-Only modal analysis is also known as Operational Modal analysis because
it deals with testing of structure during normal operation.

2.4.1 Need

For large engineering structures like bridges, dams and high rise buildings, it is difficult
to excite them artificially and to measure the excitations. Random forces act on them
such as on bridge different forces such as vehicle trafficking, wind excitation, ocean
waves etc excite the structure together so it’s almost impossible to measure all these
forces simultaneously. If the forces are not measured correctly, then modal analysis
cannot give accurate estimates of the modal parameters. On the other hand, when we
excite such structure, large amplitude force is required to vibrate the structure at all points
under observation. Such force can cause local damages to the structure. Another solution
to the problem was advised, which state that model of structure should be tested in
laboratory to estimate the modal parameters. Natural conditions under which structure
operates are very difficult to be produced in the laboratory. So, output only modal
analysis is the best option available in these conditions as it depends on the responses
only and the responses could be measured with high accuracy.

Output only modal analysis is good in a way that we do not need to make laboratory
models for it. We can perform the experiments on the site of the structure. It means that
structure is not disturbed of its usage and modal analysis could be performed during the
normal operating conditions of the structure. To obtain a very accurate modal model of a
structure, huge data is required which means long testing time is required i.e. for days or
weeks. Traditional modal analysis needs to stop the operations of the structure in case of
monitoring. On the other hand, output only modal analysis considers the natural
excitations as input so all we need are the installation of sensors.

2.4.2 Applications

Output only modal testing is used in many engineering structures such as automotives,
bridges, high rise buildings, space shuttles, dams etc.

11
Like traditional modal analysis, finite element model updating is done with output only
modal analysis. Updated finite element model can predict the dynamic behavior of the
structure more reliably.

Obtained modal model of different components of complex structures the dynamical

behavior of the complete structure can be computed by using sub-structuring techniques
[1].

A structure suffering from vibration problem could be modified by structural dynamics

modification. So modal model obtained can be modified by making changes in mass,
stiffness and damping at suitable locations in order to get rid of unwanted vibrations.

In the stage of product design and development, when designers agree with a model, it is
considered as a reference modal with known modal parameters. During operation,
structural health monitoring is done in order to check faults, performance degradation,
component deterioration, impending failure etc. [2]. So, the standard reference model is
compared with in-operation model to predict the damages. Non destructive testing and
fatigue analysis are the concerned fields in structural health monitoring [3].

If the vibration level needs to be documented in locations where no measurements can be

made, Output only modal analysis can do it if we have a Finite Element Model available.
From the modal test we will obtain the modal coordinates or modal response at some
measurable locations. These modal responses will then be extrapolated to other
unmeasured locations through the mode shapes of the FE model, and by superposition the
actual responses at the location is estimated. Even though the FE models only return
normal modes this extrapolation will be good most structures.

2.4.3 Procedure

In Output-Only modal analysis, there are different techniques followed to estimate the
modal parameters. A discussion on different methods is detailed in chapter 2 followed by
the theoretical treatment and aspects of the techniques in chapter 3.

Method used in the thesis work approximates FRFs by the cross PSD between the
responses and Impulse response functions by the cross correlations between responses.
Details of the relevant theory are discussed in chapter 3 and 4.

To calculate the modal participation factor, force exciting the structure should be
measured. This is the requirement to obtain scaled mode shapes. In output only modal
analysis, only responses are measured which means that scaled mode shapes could not be
obtained like in traditional modal analysis. During past few years, many techniques have
been developed to obtain scaled mode shapes from output only modal analysis.

12
One simple method amongst devised methods is mass changes in different locations of
the structure. A scaling factor which is equivalent to the modal participation factor is
obtained by this technique formulated in chapter 5. Using this scaling factor, unscaled
mode shapes obtained from response data are scaled.

Experiment is held in two stages. In first stage, modal parameters are extracted on the
reference structure which is not loaded by masses. Same procedure is repeated on the
structure loaded by masses at certain locations. Modal parameters obtained from the two
stages experiment are used to derive a scaling factor. Details of derivation of scaling
factor are given in Chapter 5. Mode shapes are scaled using this scaling factor.

Outlines of the procedure employed in the thesis work are shown in the block diagram,
figure 2.3. One location is considered as reference. Cross correlations and cross PSD’s
are calculated between the reference and rest of locations. Methods used to estimate the
modal parameters are discussed in 3 and 4.

13
 Test Structure

Response from Response from

Reference location Rest of locations

Cross Correlations
b/w responses

Cross PSDs b/w

responses

MODAL PARAMETERS
• Resonance Frequencies
• Damping
• Modal Vector (unscaled)

Repeat experiment with added

masses on structure

MODAL PARAMETERS
• Resonance Frequencies
• Damping
• Modal Vector (unscaled)

Scaling Factor

Scaled
Mode Shapes

Figure 2.3 Procedure of Estimation of Modal Parameters employed in thesis work

14
Initially, theory is applied in a simulation on a four DOF system detailed in chapter 6.
Proposed method is applied by performing required experiments to a bolt channel joint
structure detailed in chapter 7.

2.4.4 Things to Remember

In Output only modal testing, from experimental view we have to take care of a few
general things.

1 Structure is divided in to discrete locations where responses are to be measured.

There are certain locations on a structure where structure does not vibrate. Such locations
are known as nodes. These locations should be avoided. In order to be sure that enough
and correct locations are selected, a finite element model is useful in the initial stages.

2 We should get good data. Signal to noise ratio must be good and we should
remove spikes from data. Some time domain techniques are applied to the collected data
which give smoothness to the data.

3 We should collect enough data in order to get nice frequency resolution and to get
more averaging with large block sizes. Bias errors are removed by selecting large block
sizes, while increasing the averages reduce the random error in the spectrums.

4 Excitation signal should have a flat spectrum in the vicinity of resonances of the
structure. Unevenness in the excitation spectrum produces skewness in the frequency
response function.

15
3 Output Only System Identification

Mathematical models are a convenient way of describing a dynamic system.

Mathematical models are described in two ways [4]

● Physical modeling
● System Identification

In physical modeling, dynamic model is constructed on physical knowledge and

fundamental laws of motion i.e. M, C and K matrices should be known. If physical
knowledge is limited, a model based on input/output behavior of the system can be
obtained through system identification based on calibration of a model using
experimental data. System identification is not a complete substitute of physical
modeling. System Identification basically means modeling of a dynamic system from
experimental data. Our aim in system identification is to produce modal model.

3.1 Systems Identification of Civil Engineering Structures

System is excited by Impulse hammers or electro dynamic shakers while responses are
recorded usually by accelerometers. According to traditional modal analysis, inputs to the
system i.e. excitation forces must be measured to estimate a perfect modal model.
Measurement of excitation is very easy when the excitation force is the dominant and
noise is negligible. In most of the cases, noise is of negligible level as compared to the
input excitation. In such cases, we can predict modal model which is reliable as changes
in the modal model due to disturbance are negligible.

Excitations on large structures are difficult to measure. Also, exciting these structures
with artificial exciters is a tough job because of many practical difficulties. Like on a
bridge, wind excitation, vehicle trafficking, water waves are the three primary
excitations. It’s difficult to measure all of them as the nature of excitation is random at
different locations. During such a case, we cannot measure the excitations accurately and
it results in an erroneous modal model.

The only parameter which could be measured accurately is the response data. Output
only modal analysis is system identification method which estimates a modal model by
using responses data only. Input excitation is considered to be white noise.

16
There are two main groups of output-only modal identification methods: nonparametric
methods essentially developed in frequency domain and parametric methods in time
domain.

3.1.1 Non Parametric methods

Following non parametric methods are used in practice.

3.1.1.1 Peak Picking Method

In this method, average normalized power spectral densities (ANPSDs) and ambient
response functions between all the measurement points of the structure are evaluated.
Peaks in the spectra give the Eigen frequencies and modal damping is obtained by half
power band width method. For more details about half power band width, please see [5].
This method has been used in the modal identification of buildings [6], [7] and bridges
[8], [9]. Felber [10] systematized this method later.

3.1.1.2 Frequency Domain Decomposition

Power spectral densities of a set of SDOF systems are evaluated by performing single
value decomposition of the response spectra matrix. This method was systematized and
detailed by Brincker [11]. In extracting modal damping factors estimate, more
enhancements were done by [12].

3.1.1.3 Enhanced Frequency Domain Decomposition

In this method, first we obtain the power spectral densities of a set of SDOF systems.
Then the auto correlation functions are evaluated by taking inverse Fourier transform of
power spectral densities. Inspections of modal parameters are done from the inspection of
the decay of auto correlation functions.

3.1.2 Parametric methods

Parametric methods for modal identification are as follows.

3.1.2.1 Stochastic Subspace and Auto regressive moving averages (ARMA) models

An idealized mathematical model for the structural behavior of the system is devised in
these methods. These methods can be directly applied to response time data or to
response correlation functions. These functions are either evaluated using FFT algorithm
[13] or by using Random Decrement method (RD) [14]. Modal parameters are estimated
and the model is fitted to experimental data as much as possible. For more details about

17
Stochastic subspace method and ARMA models, interested reader is referred to [15], [16]
and [4].

3.1.2.2 Least Squares Complex Exponential, Poly Reference Complex Exponential

and Covariance- Driven Stochastic Subspace Identification methods

These methods use response correlation functions to estimate the modal parameters. See
[17] for Least squares complex exponential (LSCE), [18] for Poly reference complex
exponential (PRCE), [15] for covariance driven stochastic subspace identification
methods for more details.

3.1.2.3 Data driven Stochastic Subspace Identification Method (SSI –

DATA)

This method makes use of the direct application to the response time series. For more
details, reader is referred to [19].

Random Decrement technique is associated with time domain methods but it can also be
applied to frequency domain methods. It can be associated with PP, FDD and EFDD to
obtain power spectral densities by using FFT algorithm [20]. These methods which are
Random Decrement Peak Picking method (RD-PP), Random Decrement Frequency
Domain Decomposition (RD-FDD) and Random Decrement Enhanced Frequency
Domain Decomposition (RD-EFDD) reduce the noise effect.

A road map to estimate the modal parameters with the help of above mentioned methods
is shown in the figure

18
 Figure 3.1 Block diagram for different methods to estimate modal parameters

Procedure employed in the thesis work is shown in the block diagram in figure 2.3. Road
map to estimate the modal parameters is highlighted by spheres in the figure 3.1. Cross
correlations between response data are obtained by taking the inverse Fourier transform
of the cross PSDs between responses. Cross PSDs are obtained using Welch method
(FFT based method). Modal parameters are then obtained using LSCE method.

19
4 Parameter identification method

Modal parameter identification method employed in this thesis work is shown as a block
diagram in chapter 1. If compared with traditional modal analysis, frequency response
functions between input and output are approximated as cross power spectral densities
between responses while the impulse response functions are approximated by cross
correlations between responses. Prime objective in this chapter is to study the underlying
theory related to the estimation of natural frequencies, damping ratios and unscaled mode
shapes in the thesis work. Topics under study include power spectra estimation, modal
decomposition of PSD, truncation of PSD in OMA, Poles and Residues estimation.

4.1 Power Spectral Density Estimation

Power spectral density is the expected (mean or average) energy density of the signal as a
function of frequency.

For zero mean signals, this equals the variance of each frequency sample. In estimating
power spectral density (PSD) of a signal, there are two tradeoffs. One is frequency
resolution and the other is noise in the signal. During measurements, we have finite
length of signal. Two types of errors could occur during PSD estimation. Bias error is
concerned with block size of signal while random error is concerned with averages of
signal. To obtain a good estimate of PSD, we should have large length of the signal. Only
this fact can provide good estimation of PSD. If we take fewer averages in the signal then
spectrum will not be smooth. On the other side, if we take small block size, bad
frequency resolution could introduce leakage in the spectrum and the necessary
information required from the spectrum will be erroneous.

There are two basic methods to estimate the power spectral density explained in sections
(4.1.1) and (4.1.2).

4.1.1 Periodogram Method

Simple periodogram method is nothing but quotient of the squared magnitude of the
Fourier transform of the signal and length of the signal. Periodogram method is improved
in Modified Periodogram approach where certain window other than rectangular window
is applied to the signal before taking the Fourier transform. Windows solve the leakage
problem. Bartlett Periodogram averaging introduces the concept of averaging of the
different blocks of the signal. It will decrease variance of the signal at the expense of

20
resolution. Welch’s Periodogram averaging with overlapping is a method which
introduces all the properties of above mentioned methods. This method divides the signal
in to blocks and then increases the averaging by taking overlaps of the blocks.

If we have two discrete signals x(n) and y(n) where, (x(n) is considered to be the
reference signal) with L samples. We divide it into M blocks with N samples in each
block. Consider overlapping of the blocks as well and also let the signals are windowed
i.e. multiplied with a window in time domain. The cross power spectra between these two
signals is expressed as

2 M Yw,m (k ) X ∗ w,m
S yx (k ) =
MBe
∑
m =1 CAN
k = 1, 2…….. N/2 +1 (4.1)

X ∗ w,m is the complex conjugate of the windowed Fourier transform of a block from signal
x(n).
C A is the window compensation factor given by following relationship,

N −1

∑ w(n)
CA = n =0
(4.2)
N

Be is the equivalent noise band width for the window. It is given by following
relationship,

N −1
N ∑ w 2 ( n)
fs
Be = n =0
2
. (4.3)
⎛N −1
⎞ N
⎜ ∑ w(n) ⎟
⎝ n =0 ⎠

fs is the sampling frequency of the signal.

w(n) is the discrete window signal
N is the block size.

Many windows are available, each one having specific application in signal processing.
Hanning window is used to treat the random signals. So in PSD estimation, we use
hanning window to reduce the leakage effect if the excitation signal is random.

21
4.1.2 Correlation Method

[23] This method uses the auto or cross correlation between the signals. The correlation
function is then Fourier transformed to obtain the corresponding auto or cross spectral
densities. Averaging required estimating the PSD is performed in the time domain in this
approach. The cross correlation between two discrete signals x(n) and y(n) is given by
(x(n) is kept as reference) each with finite block length of N samples,

1 N −(1− n )
r ( n) = ∑ ( y(k ) x ∗ (n + k ))
N − n k =0
(4.4)

With increasing lag n, fewer values are averaged, so more noise is introduced in to the
estimated power spectrum. Windowing gives optimum solution to this problem at the
expense of less resolution. To cope with it, cross correlation function is multiplied with
window and then it is Fourier transformed to find the frequency function. So, final
expression for estimation of cross power spectral density is given by following
mathematical equation,

∑ (r (n)w(n)e )
M
− jwn
S yx ( w) = (4.5)
n=− M

4.2 Modal Decomposition of Spectra Densities

As mentioned earlier, we will approximate the cross power spectral densities as the
frequency response functions in output only modal analysis. Any structure vibrated with
either impulse hammer or shaker can produce a row or a column in FRF matrix. If we
measure one row or one column in the FRF matrix, we can synthesize a complete FRF
matrix because FRF matrix for linear time invariant system is considered to be
symmetric. FRF matrix of size (Nm x Nm) is written as

⎡ H 11 ( jω ) L H N M 1 ( jω ) ⎤
⎢ ⎥
H ( jω ) = ⎢ M O M ⎥ (4.6)
⎢ H 1 N ( jω ) L H N N ( jω ) ⎥
⎣ m m m ⎦

Any ‘N’ degrees of freedom system could be expressed as a sum of ‘N’ single degree of
freedom systems as follows.

22
 Nm [Ar ]N [A ] *

[H ( jω )]N m ×Nm
=∑ m ×Nm
+
r Nm ×Nm
(4.7)
r =1 jω − λ r jω − λ*r

[H ( jω )] is the FRF matrix, [Ar ] is residue and λr

is a pole for a particular mode.
Poles are complex conjugate pair given by following equation,

λ r , λ r * = −ζ r ω r ± jω r 1 − ζ r 2 (4.8)

ω r is the undamped resonance frequency of mode r, and ζ r is the relative damping of

mode r. The residue matrix for mode r, [ Ar ] can be shown to consist of a product of the
mode shape vector {ψ }r i.e.

[A]r = Qr {ψ }r {ψ }r
T
(4.9)

Where Qr is a scaling constant that has to be used because the eigenvectors can be
arbitrarily scaled (they only tell the shape of the modes).
If equation (4.9) is expanded it gives

⎡ψ 1ψ 1 ψ 1ψ 2 L ψ 1ψ N ⎤
⎢ψ ψ ψ ψ ψ 2ψ N ⎥⎥
[A]r = Qr ⎢ 2 1 2 2 (4.10)
⎢ ⎥
⎢ ⎥
ψ ψ
⎣ N 1 ψ ψ
N 2 ψ N ψ N ⎦r

Equation 4.6 can be written in terms of modal vectors by using equation (4.10) as follows,

Q {ψ }r {ψ }r Qr {ψ }r {ψ } r
Nm T * * *T
[H ( jω )]N × N =∑ r + (4.11)
r =1 jω − λ r jω − λ*r

The time domain equivalent of the modal model in equation (4.6) is the impulse response
matrix [h (t)], where

23
 [h(t )] = ∑ ([A]r e λ t + [A∗ ]r e λ t )
Nm
∗
(4.12)
r r

r =1

As mentioned earlier, input excitation need to be measured to develop FRF. If the

excitation forces are considered to be white noise then [7] gives the relationship between
the FRF and spectra densities as follows,

[S XX (ω )]= [H (ω )] [S FF (ω )] [H (ω )]H (4.13)

Where,
[S FF (ω )] is the spectra density matrix of the (unknown) input forces.
[S XX (ω )] is the spectra density matrix of the responses.
Since the excitation forces are assumed to be white noise sequences, [S FF (ω )] can be
considered a constant matrix with respect to the frequency. By substituting equation (4.7)
in (4.13) and assuming white noise input, the matrix [S XX (ω )], with cross power spectra
of the responses, evaluated at frequency ω , can be modally decomposed as follows:

N ⎛
{ }{ }T { }*{ }H { }{ }T { }* { }H ⎞
[S XX (ω)] = ∑⎜⎜ ψ r Qr + ψ r Qr + Qr ψ r + Qr ψ r * ⎟⎟
m

(4.14)
r =1 ⎝ jω − λr jω − λr − jω − λr − jω − λr ⎠

Where, {ψ r } and {Qr } are respectively the mode shape and operational reference vector
for mode r. The reference vector is a complex function of the spectral density matrix of
the unknown random input force(s) and the modal parameters of the structure. It should
be noted that the modal participation factors Qr can not be determined. Since these
factors are required for the normalization of the mode shapes, the operational mode
shapes {ψ r } remain unscaled i.e. dependent on the unknown excitation forces acting on
the structure. Equation (4.14) forms the basis for frequency-domain output-only modal
analysis. The order of the model in (4.14) is twice the order of the model in (4.7).

Taking the discrete-time inverse Fourier transform of (4.14) gives the correlation function
matrix R(k) for positive and negative time lags k [24]

⎧ Nm
⎪ ∑ {ψ r }{Qr } e + {ψ r }{Qr } e λr kTs
*
T λr kTs
for k ≥ 0
H

⎪ r =1
R(k ) = ⎨ (4.15)
⎪ Nm
⎪∑ {Qr }{ψ r } e + {Qr } {ψ r } e λr kTs
T λr k Ts * *
for k < 0
H

⎩ r =1

24
It is interesting to note that the casual part of the correlation functions (positive lags) give
rise to the first two terms in (4.14) while the non-casual part (negative lags) give rise to
the last two terms in (4.14) i.e. (Hermitian conjugate terms). Equation (4.15) is based on
the principle of the cross-correlation technique which states that the correlation functions
can be expressed as a sum of decaying sinusoids [25]. Each sinusoid has a damped
natural frequency and damping ratio that is equal to one of the corresponding structural
mode.

4.3 Truncation of the Cross-Spectral Density Matrix

The order of the model in the cross-spectral density matrix (4.14) is twice the order of the
model in the FRF matrix (4.7). As a result of this the cross-spectral needs to be clipped,
which means that one-half of the length of the cross spectra will be truncated.

For this purpose, cross correlations were computed from the responses. Each cross
correlation between responses is truncated i.e. one half of the samples are taken while the
remaining part is thrown in accordance to the order of the frequency response function
matrix. Then, the truncated Cross correlations are Fourier transformed to obtain power
spectral densities whose orders are same as that of frequency response function.

4.4 Poles and Residues Estimation

After acquiring response data, cross correlations and cross PSDs are estimated. Modal
parameters are obtained from the poles and residues of the system given in equations
(4.8) and (4.9). Curve fitting (Modal parameter estimation) algorithms are used to
estimate the modal parameters for each mode of interest in a particular frequency range.

Least squares complex exponential method (LSCE) [17] is used for the estimation of
modal parameters in the thesis. This method uses the exponentially decaying impulse
responses given by equation (4.12). This method is coded in Matlab toolbox used in the
thesis work..

In this method, user is asked to select the frequency range of interest which is used for
curve fitting. Next step in this method is to select the number of poles which are to be
calculated by the estimation algorithm. This assumption is the most decisive in the
algorithm. Usually, the number of poles that give the best result are not exactly the
number of actual poles, but the number of true poles plus a small number of ‘extra’ poles
that take account of errors in the measure FRF’s , sometimes called ‘computational
poles’. In order to select proper number of modes, the algorithm use Stability diagram
and it is obtained using MIF [26]. Examples of stability diagram are shown in the

25
parameter estimation of a simulation on 4DOF system in figure (6.3) and a bolt channel
joint structure in figure (7.5).
After the selection of poles, the driving point FRF is normally used to calculate the
residue of that DOF for each mode. This number is then used in the next step, to obtain
the complete mode shapes, by applying equation (4.10) [26]. Residues are obtained using
least squares frequency domain method in the thesis work.

4.5 Selection of Reference Responses

In output only modal analysis, we are concerned with measuring the responses of the
structure. A structure is divided in different points where measurements are to be
acquired. Usually, finite element model is used initially so that information regarding
nodes is obtained. In traditional modal analysis, either responses are measured with fixed
accelerometers (Impulse hammer excitation case) or roving accelerometers (shaker
excitation case). In output only modal analysis, measurements are done by keeping one or
more locations as reference. This condition depends upon the availability of sensors. If
available sensors are enough for the measurement locations then all the experiment can
finish in one measurement set. In this case, we can select any location as our reference
location and the corresponding cross correlation and cross power spectral density
matrices are computed. In case of limited available sensors, references are chosen and
other locations are measured in by roving sensors in patches.

Two conditions could occur here. Either the patch of reference responses is moved after
two measurements to a new location or they remain fixed at particular location. Former
methods are known as jumping or leap frog method while the latter is fixed reference
method [10]. A condition applied on leap frog approach is that the reference sensors may
land in nodal points of one or more modes of the structure. In the simulation and
experimental part of this thesis work, fixed reference method is employed.

Ambient excitations level changes with time i.e. it is of non stationary nature. An extra
effort is needed in gluing the different patches of the mode shapes together. Techniques
used for time efficient processing of multi-patch output only data under non stationary
excitation are not dealt in this thesis work because excitations are considered stationary.
Interested reader is referred to [3] and [27].

26
5 Normalization of mode shapes

Due to clear advantage of technology and easy testing, output only modal analysis has
gained more popularity. As mentioned earlier, to obtain scaled mode shapes we need
modal participation factor which is dependent on the excitation. As excitations are not
measured so modal participation factors could not be obtained. It concludes that we can
get resonance frequencies, damping and unscaled mode shapes. In applications field such
as structural health monitoring, structural response simulation and structural
modification, scaled mode shapes are demanded so it restricts the usage of output only
modal analysis [28].

Some techniques are suggested to get scaled mode shapes. One technique suggested by
[29], based on assumption that partition of the inverse of the mass matrix associated with
the measured coordinates is diagonal. However, the approach gives answers only when
there is a full set of modes, and robustness for a truncated modal space has not been
demonstrated. Another approach was suggested by [30] which is based on a more
extensive testing procedure that involves repeated testing where mass changes are
introduced in the points of the structure where the mode shape is known. This approach
seems more feasible as we need information of those modes only which are to be scaled.
Parloo derived an approximate formula for the scaling factor from some basic sensitivity
relations in linear dynamics.

Same formula was derived using equation of motion [30]. Theme of the equation is that if
we add masses to a structure, it will provide shift in the natural frequencies. So we add
masses to the structure at some discrete locations and we get change in resonances for all
the modes. Resulting frequency shift for added masses is used in the derivation of
formula for the scaling factor.

5.1 DERIVATION

Usually in traditional modal analysis, the modes are scaled such that the modal mass is
equal to one, i.e.

{ψ }T [M ]{ψ } = 1 (5.1)

Where

27
{ψ } is the scaled modal vector
[M] is the mass matrix.

It is important to note that the residue matrix [ A]r for a particular pole λ r is related to the
modal vectors. For an r-th mode of an Nm degree of freedom system:

[A]r = Qr {ψ }r {ψ }r
T
(5.2)

⎡ψ 1ψ 1 ψ 1ψ 2 L ψ 1ψ m ⎤
⎢ψ ψ ψ ψ ψ 2ψ m ⎥⎥
[A]r = Qr ⎢ 2 1 2 2 (5.3)
⎢ ⎥
⎢ ⎥
⎣ψ mψ 1 ψ mψ 2 ψ mψ m ⎦ r

Where,
Qr is the scaling constant that is a function of the scaling of the modal vectors.
Using only the q-th column of the residue matrix:

⎡ A1q ⎤ ⎡ψ 1ψ q ⎤ ⎡ψ 1 ⎤
⎢A ⎥ ⎢ψ ψ ⎥ ⎢ψ ⎥
⎢ 2 q ⎥ ⎢ 1 q ⎥ ⎢ 1⎥
⎢ ⋅ ⎥ ⎢ ⋅ ⎥ ⎢ ⋅ ⎥
⎢ ⎥ = Qr ⎢ ⎥ = Qrψ qr ⎢ ⎥ (5.4)
⎢ Aiq ⎥ ⎢ψ iψ q ⎥ ⎢ψ 1 ⎥
⎢ ⋅ ⎥ ⎢ ⋅ ⎥ ⎢ ⋅ ⎥
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
⎣⎢ Amq ⎦⎥ ⎣⎢ψ mψ q ⎦⎥ ⎢⎣ψ 1 ⎥⎦

For a driving point FRF H qq (ω ) obtained by exciting at point q and measuring the
response at point q, the driving point residue for mode r Aqqr is given by

Aqqr = Qrψ qrψ qr = Qrψ qr2 (5.5)

Equation (4.5) can be used to define modal mass M r that is consistent with modal mass
defined analytically from the mass matrix.

{ψ }Tr [M ]{ψ }r = Mr (5.6)

28
Rather than defining modal mass in terms of an absolute quantity, Aqqr , modal mass is
defined based on Qr which is a relative quantity. If the scaling of {ψ }r is changed, Qr will
be altered accordingly. Therefore, the r-th modal mass M r of a multi-degree of freedom
system is defined as,

1
Mr = (5.7)
j 2Qr ω r

ψ prψ pr
Mr = (5.8)
j 2 A pqr ω r

Where,
M r = Modal mass
Qr = Modal scaling constant
ω r = Damped natural frequency
From equation (5.7), the modal scaling constant Qr is given by

1
Qr = (5.9)
j 2M rω r

Since, the modes, in traditional modal analysis, are scaled such that the modal mass is
equal to one,

1
Qr = (5.10)
j 2ω r

Recalling equation (5.5), the driving point residue for mode r, Aqqr , is given by

Aqqr = Qrψ qrψ qr = Qrψ qr2 (5.11)

Putting the value of Qr in equation (5.10) into (5.5), the scaled residues becomes

1
Aqqr = ψ qrψ qr (5.12)
j 2ω r

29
The scaled modal vector is {ψ }r , is given by
1
{ψ }r = {A}r (5.13)
Qrψ qr

Usually in output-only modal analysis the mode shapes are just scaled to unity, i.e.

{φ }T {φ } = 1 (5.14)

1
{φ }r = {A}r (5.15)
{A}r 2

Where,

{A}r 2
is the vector norm of {A}r

m
{A}r 2
= ∑A iqr
*
Aiqr (5.16)
i =1

5.1.1 Scaling factor

The scaled and unscaled mode shapes are related by the equation,

{ψ } = {φ } (5.17)
{φ } [M ]{φ }
T

Where
{ψ } is the scaled mode shape,
{φ } is the unscaled mode shape.
Equation (5.3) implies that the scaling factor is expressed as

30
 1
α= (5.18)
{φ } [M ]{φ }
T

and

{ψ } = α {φ } (5.19)

The equation to estimate the scaling factor is derived from the basic equation of motion
of a structure subjected to a force {F (t )}, i.e.:

[M ]{&z&} + [C ]{z&} + [K {z }] = {F (t )} (5.20)

Where,

[M ] is the mass matrix,

[C ] is the damping matrix,
[K ] is the stiffness matrix.
The classical eigenvalue equation in case of no damping is

[M ]{ψ 1 }ω12 = [K ]{ψ 1 } (5.21)

Where,
{ψ 1 } is the mode shape,
ω1 is the natural frequency of any of the modes of the problem related to the mass
matrix M.

If a mass change is made so that the mass matrix becomes [M ] + [ΔM ], then the Eigen
value equation becomes

([M ] + [ΔM ]){ψ 2 }ω12 = [K ]{ψ 2 } (5.22)

31
Where,
{ψ 2 } and ω 2 are now the modal parameters of the modified problem.

Subtracting equation (5.8) from (5.7) then gives the equation to approximate in order to
obtain an equation for the unknown scaling factor.

[M ]({ψ 1 }ω12 − {ψ 2 }ω 22 ) − [ΔM ]{ψ 2 }ω 22 = [K ]({ψ 1 } − {ψ 2 }) (5.23)

If it is assumed that the mass change is so small that the mode shape does not change
significantly, i.e.

{ψ 2 } ≅ {ψ 1 } = {ψ } (5.24)

Hence, equation (4.9) becomes

[M ]{ψ }(ω12 − ω 22 ) = [ΔM ]{ψ }ω 22 (5.25)

Pre-multiplying equation (5.25) by {ψ } results in,

T

{ψ }T [M ]{ψ }(ω12 − ω 22 ) = {ψ }T [ΔM ]{ψ }ω 22 (5.26)

Considering the orthogonality of the modes,

{ψ }T [M ]{ψ } = 1

Hence, equation (5.26) becomes

(ω12 − ω 22 ) = {ψ } [ΔM ]{ψ }ω 22

T
(5.27)

Now, recall that in output-only modal testing the modal vector length is scaled to unity
(5.14)

32
 {φ }T {φ } = 1

Finally, combining equation (4.19) and (4.27) we obtain,

(ω12 − ω 22 ) = α 2 {φ } [ΔM ]{φ }ω 22

T
(5.28)

And the unknown scaling factor can be obtained as

α=
(ω 1
2
− ω 22 ) (5.29)
ω {φ } [ΔM ]{φ }
2
2
T

When using equation (5.29) to determine a scaling factor, only the mode shape and the
natural frequency of that particular mode have to be known. Equation (5.29) gives exact
results when the matrix [ΔM ] is proportional to the mass matrix [M ] since in this case
the modes remain unmodified. [28].
In equation (5.29), both the modified and unmodified mode shapes can be used.
However, the better results are obtained using the unmodified shapes, i.e.

α=
(ω 2
1 − ω 22 ) (5.30)
ω {φ1 } [ΔM ]{φ1 }
2
2
T

or both the unmodified and the modified mode shapes, i.e.:

α=
(ω 2
1 − ω 22 ) (5.31)
ω {φ1 } [ΔM ]{φ 2 }
2
2
T

The results of scaling factor obtained with equation (5.30) are independent of the type of
normalization used whereas normalization to the length should be used with equation
(5.31)

33
5.2 Things to Remember

Mathematical equation for estimating scaling factor shows that we have to add masses at
defined locations. We don’t have the actual mass, damping and stiffness values for the
defined locations before starting experiment. The added mass should not be very large as
it can distort the mode shape. Mass which we add could seem smaller but at the location,
at which we add it, may produce different effect as compared to other location. In this
way, mode shapes could be erroneous. Adding large masses could cause local damage to
the structure. Mass addition also affects stiffness of the structure.

Frequency shifts produced due to mass addition should be significant. Previous research
in the field of output only modal testing shows that frequency shifts up to 1% or 2% is
considered to be fit to provide good results.

Mass addition should be avoided at node locations. Resulting change in resonance

frequency for the particular mode will be very small even for large change in mass. The
target is to produce change in resonance frequency with small addition of masses because
large change in mass affect mode shapes i.e. mode shapes is deteriorated.

As a conclusion, we have to add masses with small magnitude and should avoid adding
them at nodes of the structure.

Multiple mass additions i.e. if we add different masses at more than one location could
improve the results. In this way, mass adding effect distributes among different locations
and we could get reasonable frequency shifts for different modes.

34
6 Simulation

This chapter deals with application of the theory of Output-Only modal analysis on four
degrees of freedom system whose mass, stiffness and damping matrices are known.
System is excited by Gaussian white noise. Responses are simulated and modal
parameters are extracted. Moreover, normalization of mode shapes is carried for scaling
the mode shapes. Results are compared with the traditional modal analysis.

6.1 Description of System

A four degrees of freedom lightly damped system shown in figure (6.1) is selected for
simulation with following parameters.

Table 61 Parameters of the 4DOF system analyzed in simulation

Masses Magnitude Damping Magnitude Spring Magnitude
(kg) Constants (Ns/m) Stiffness (Nm)
m1 225 c1 600 k1 1500000
m2 175 c2 140 k2 900000
m3 80 c3 175 k3 500000
m4 42 c4 150 k4 450000
k5 200000

Given system is treated as non proportional damped system. Modal parameters extraction
for such system with given M, C and K matrices is done by using state space vectors
model. As theoretical results used here are just for comparison and the scope of the thesis
deal with the methods used in output only modal analysis so the interested reader is
referred to Chapter 6 of [32] or Chapter 9 [33].

Theories described in earlier chapters are coded in Matlab tool box. Required results are
interpreted and the figures are plotted for comparing theory and output-only modal
analysis.

35
 Figure 6.1 4 DOF system used in simulation

Given system is excited by Gaussian white noise containing two million samples. Force
vector is filtered using fourth order low pass Butterworth filter. Responses are simulated
and cross correlations between the responses are evaluated by keeping response at 2nd
mass as a reference. In a similar way, cross power spectral densities are computed from
cross correlations. Mode indicator function is obtained from cross power spectral
densities. MIF is used as a background in the stability diagram for estimation of poles.
Poles are estimated using Complex exponential method. Finally, residues are estimated.
All these methods are coded in matlab tool box.

Modal vectors obtained so far are not scaled. Following mode normalization technique,
masses of 4kg and 8kg are added at location DOF1 and DOF4. Same procedure as shown
above is followed to obtain the poles and residues for the case of added masses. At this
stage, we get poles and residues for with and without addition of masses to the system.
Using these parameters, we get the scaling factor to scale our mode shapes using equation
(5.29).

36
6.2 Results And Comparisons
Cross correlation between responses at 1st mass and 2nd mass is shown in figure.

Figure 6.2 Cross correlation between Response at DOF 1 and DOF 2

Calculated cross correlations are truncated i.e. half of them are taken because they are of
order twice that of impulse response. Stability diagram which shows the location of
resonances is shown in figure (6.3).

Figure 6.3 Stability diagram for poles estimation using complex exponential

37
Estimated poles along with cross power spectra matrices are used to find residues. These
residues are unscaled which means that modal vectors are needed to be scaled. Scaling
factor is derived by using mode normalization technique defined in chapter 5. This is
done by adding small masses of magnitude 4 kg and 8 kg at DOF 1 and DOF 4
respectively. As a result of mass addition, we get a shift in the resonance frequencies is
obtained. For the modified structure (structure loaded with masses), we get new poles and
residues. Using the evaluated parameters in equation (5.31), we get the scaling factor to
scale the residues. Hence, we get scaled FRFs.

Theoretical frequency response functions are calculated using original M, C and K

matrices. Comparisons of few modal parameters are shown in figures and tables below.

Figure 6.4 Comparison of Theoretical and Estimated H12

38
Figure 6.5 Comparison of Theoretical and Estimated H22

Figure 6.6 Comparison of response time histories

39
Table 6.2 Comparison of theoretical and estimated resonance frequencies, damping and
scaling factor

Modes Frequency Hz Damping Scaling

Theory Estimated Theory Theory Estimated

Estimated
1 7.3746 7.3749 0.4770 0.4754 0.0113 0.0119
2 12.6748 12.6726 1.3625 1.4195 0.0133 0.0128
3 18.6701 18.6744 1.4807 1.4748 0.0065 0.0064
4 23.3312 23.3348 2.2629 2.2875 0.0154 0.0127

Theoretical and estimated mode shapes are compared in terms of percent error using the
following equation

Theoretical − Estimated
%error = × 100 (6.1)
Theoretical

Table 6.3 Comparison of estimated and theoretical mode shapes in terms of relative error
Relative Error of Mode shapes (%)
Mode 1 2 3 4
DOF1 0.3448 0.0655 0.1218 1.0533
DOF2 0.0699 0.1737 0.0880 0.1318
DOF3 0.0553 0.0552 1.7281 0.0770
DOF4 0.0285 0.0200 0.2134 0.0686

Comparisons of different modal parameters show very small deviations between output
only and theoretical results. It is concluded that theory works.

An important point worth noting is that the system is an ideal system which has discrete
masses. In real life, situation is very different as compared to simulation world as we
don’t have defined M, C and K matrices and we measure on limited discrete points. So it
depends on skills in different aspects. Next chapter deals with output only modal analysis
on a bolt channel joint structure.

40
7 Application to a Bolt Channel Joint Structure

7.1 Introduction

In this chapter, Output only modal analysis technique is applied to a bolt channel joint
structure. It consists of two beams, joined together in longitudinal and transverse
directions by bolt channel joints. Longitudinal beam is fixed at one end. Both the beams
have same cross section dimensions. Structure is excited by random forces with an
electrodynamic shaker. Normalization of mode shapes is carried out by adding masses on
the structure in accordance with the theory explained in chapter 5.

All the required calculations are done with the help of Matlab toolbox. For the purpose of
validation of results, force data is also measured along with response. 1st and 2nd
resonances of the structure are analyzed. Modal parameters obtained from output only
modal analysis are compared with modal parameters obtained from traditional modal
analysis.

7.2 Equipment Setup

In this section equipment setup is discussed. Force signal is produced in Matlab with a
function randn. Signal is fed to a combination of low and high pass filters. Filter output is
fed to amplifier. Amplifier feeds the signal to electrodynamic shaker to excite the
structure. Excitation from the shaker is measured through a force sensor. Response data
in terms of acceleration is measured by the accelerometers. In one patch, three
accelerometers are used. Data from accelerometers and force sensor is saved in computer
through a four channels data acquisition tool. Table 7.1 shows the numbering of
equipments used in the experiment. Figure 7.1 shows the equipments setup.

41
 Table 7.1 Numbering of equipment used in experiment

NUMBERS ITEMS
1 Structure under test
2 Electrodynamic shaker
3 Data Acquisition tool
4 System for storing measured data
5 System for producing excitation signal
6 Low and High pass Filters
7 Amplifier
8 Accelerometer
9 Force sensor

Figure 7.1 Equipment setup for experiment

42
7.3 Experimental Setup

Structure is divided into eleven points where responses are to be measured. Length of the
transverse beam is 1000 millimeters while longitudinal beam is 840 millimeters.
Transverse beam is divided into 7 locations while longitudinal beam is divided in to 4
locations. Distances between the measurement locations could
be seen in the figure7.2.

Figure 7.2 Locations for measurement of response data

Acceleration in the vertical direction will be measured with accelerometers. Location 1 is

selected as reference DOF. Structure is excited at location 1by an electrodynamic shaker.

Data acquisition tool has four input channels. One channel corresponds to measurement
of excitation force through out the experiment. Location 1 is selected to be the reference;
hence, second channel is dedicated to the acceleration data acquisition of location 1. The
remaining two channels are dedicated to the roving locations. Table (7.2) gives the
explanation of measurement points in different patches.

43
 Table 7.2 Locations numbering per patch

Patches Locations

1 1 (reference) 2 3
2 1 (reference) 4 5
3 1 (reference) 6 7
4 1 (reference) 8 9
5 1 (reference) 10

This experimental setup is used for both the conditions i.e. with and without mass loading
of structure.

7.3.1 Procedure

Traditional modal analysis and output only modal analysis techniques are applied to the
structure. It follows that excitation force is also measured along with responses in each
patch.

Matlab function randn is used to produce uniformly distributed numbers (white noise).
Two million samples are produced. The generated signal is then filtered through a band
pass filter between 15 Hz and 45 Hz. High pass and low pass filters are combined to
produce the band pass filtered signal. Signal is fed to the shaker through Matlab function
sound with sampling frequency 8192 Hz. Shaker excitation is measured through force
sensor while acceleration data is measured through accelerometers. Measured force
signal along with acceleration signals from the available locations in particular patch are
fed to the data acquisition tool. Data is stored in the computer through data acquisition
tool with sampling frequency of 2000 Hz. Time length of 254 seconds was used which
meant that recorded signals will have length of 508000 samples.

According to theory, force spectrum should be flat within filtered frequency range
because signal produced was Gaussian white noise. Force spectrum was plotted which
show force drop outs at certain frequencies shown in the figure (7.3).

44
 Figure 7.3 Spectrum of the excitation signal

Force spectrum produced conflicts with the theory. The two major force drop outs are in
the frequency range of 19 Hz – 23 Hz and 24 Hz – 31 Hz.

A method of PSD compensation was devised1 which produce a time signal from a given
PSD spectrum. Basic idea was to plot a spectrum with triangles in the opposite direction
to force drop outs at particular frequencies. These triangles will compensate for the drop
outs in the force spectrum and as a result, a flat spectrum could be produced. A signal of
2097152 samples was produced which was fed as force signal to the structure.

Acquired data consisted of 409600 samples which mean that we have 25 equal blocks of
length 16384 samples in each. Force spectrum obtained after compensation of PSD is
shown in the figure (7.4).

1 The method is coded in Matlab with a function psdsynt, by Saven Edutech AB provided by my supervisor
Professor Kjell Ahlin

45
 Figure 7.4 Spectrum of the excitation signal after psd compensation

Figure (7.4) shows that we succeeded in compensating for the severe force drop outs in
the spectrum. Still certain drop outs are visible which conflicts with the theory and more
intensive study is required for compensating the remaining force drop outs. The same
signal is used to excite the structure for all the locations in patches. Analysis procedure
shown in figure 2.4 is followed to produce cross correlations and cross power spectral
densities. Stability diagram produced for calculation of poles is shown in the figure 7.5

Figure 7.5 Stability diagram for location of poles

46
Un Scaled Frequency response functions are shown in the figure 7.6

Figure 7.6 Unscaled FRFs using response data

For the purpose of comparison and validation of modal parameters estimation, force
signal is also acquired along acceleration signals for every patch. Frequency response
functions are obtained by traditional modal analysis technique summarized in figure 2.2.
Matlab function tfestimate is used calculate the FRFs.

7.3.2 Estimation of Mode Shapes

Residues obtained from response data are not scaled. According to theory defined in
chapter 5, point masses are added on the structure to calculate the scaling factor.
Locations 2 and 6 are selected for adding point masses of 100 grams and 210 grams
respectively.

Same procedure is used as above and data is acquired in different patches. Scaling factor
is obtained using equation 5.30 in chapter 5.

47
7.4 Comparison of Results

Natural frequencies and damping ratios for the first two modes are compared in table 7.3
while magnitudes of residues are compared in table 7.4.

Table 7.3 Resonance frequencies and Damping ratios from traditional and output only modal
analysis techniques
Modes Frequency Hz Damping

Traditional Estimated Traditional Estimated

1 20.8408 20.9728 0.0037 0.0066
2 27.1031 27.2036 0.0049 0.0029

Table 7.4 Residues comparison obtained from Traditional modal analysis and Output Only
modal analysis
Modes 11 21 31 41
Trad OPO Trad OPO Trad OPO Trad OPO

1 0.0017 0.0019 0.0015 0.0017 0.0014 0.0015 0.0012 0.0014

2 0.0022 0.0028 0.0013 0.0016 0.0005 0.0006 0.0003 0.0003

Modes 51 61 71 81
Trad OPO Trad OPO Trad OPO Trad OPO

1 0.0011 0.0013 0.0010 0.0013 0.0009 0.0012 0.0008 0.0009

2 0.0010 0.0012 0.0017 0.0021 0.0026 0.0033 0.0002 0.0002

48
 Modes 91 101
Trad OPO Trad OPO

1 0.0004 0.0005 0.0001 0.0001

2 0.0001 0.0001 0.0000 0.0000

Frequency response functions are compared for different locations shown in the
following figures 7.7 – 7.10.

Figure 7.7 Comparison of FRF (input at DOF1 and output at DOF1), from traditional
and output only modal analysis technique

49
Figure 7.8 Comparison of FRF (input at DOF1 and output at DOF2), from traditional and output
only modal analysis technique

Figure 7.9 Comparison of FRF (input at DOF1 and output at DOF3), from traditional and output
only modal analysis technique

50
Figure 7.10 Comparison of FRF (input at DOF1 and output at DOF4), from traditional and output
only modal analysis technique

Figure 7.11 Comparison of FRF (input at DOF1 and output at DOF5), from traditional and output
only modal analysis technique

51
Correlation diagram for the mode shapes obtained from traditional and output only modal
analysis are shown.

Figure 7.12 Correlation of 1st mode shape for Traditional and Output Only modal analysis

Figure 7.13 Correlation of 2nd mode shape for Traditional and Output Only modal analysis.
Numbers show the output DOF

52
7.5 Discussions

Error regarding resonance frequencies is very small. Structure is very low damped and
the damping estimates deviation is smaller as well. Correlation for the 1st mode shape
gives very nice results but results deviate for the second mode shape.

To explain the difference, we look back at our assumption. We assumed the excitation to
be white noise. More precisely speaking, a completely flat spectrum through out the
desired frequency range is not required. Flat spectrum is required in the vicinity of
resonance of the structure. Uneven spectrum in the vicinity of resonance produces
skewness in the frequency response function. Force spectrum shown in figure (7.4) shows
unevenness in the vicinity of second resonance. This effect produces deviation in the
results.

Another assumption was that masses added on the structure are point masses and
magnitude of masses should be small enough to make a small frequency shift keeping the
mode shapes unchanged. So mode shapes before and after adding masses should have
perfect correlation. Figures (7.14) and (7.15) show the correlation for both conditions.

Figure 7.14 Correlation of 1st mode shape for with and without mass loading on the structure

53
 Figure 7.15 Correlation of 2nd mode shape for with and without mass loading on the structure

Figure 7.14 is showing correlation between the modal vectors with and without mass
addition for the first mode shape. Figure shows a nice correlation for all the locations
except locations 6 and 5.

Second mode shape correlation for both conditions deviate much at various locations as
could be seen in figure (7.15). It is concluded that the condition of no change in mode
shapes is not valid completely here. To improve results, the structure should be loaded
with masses of less magnitude. Results for deviation of 2nd mode shape are according to
expectation because shift in the 2nd resonance is more as compared to the 1st resonance.
Also, 2nd resonance could only be affected if we load the structure at end locations on the
transverse beam because it is torsional mode. Location 1 and 2 were not suitable because
they are very near to the free ends and loading of structure at these positions could disturb
the equilibrium position of the structure. For these reasons, location 2 and 6 were selected
for mass loading. If we select location 3 and 5 then we need more magnitude of mass in
order to decrease the second resonance frequency because we are more near to the node
for torsional mode which is near to the center of the transverse beam. Due to limited
number of measurement points we paid for the placement of masses. If we increase the
number of measurement points on the transverse beam then better options could be
produced.

Measured frequency response function needs to be improved. Frequency resolution was

very much limited and more data is required to improve the results. Frequency resolution
used in this experiment is 0.4833 which is more than the difference in the resonance

54
frequencies obtained from traditional and output only modal analysis. 409600 samples for
each measurement were acquired which means that for acquired sampling frequency of
2000 Hertz, we get 25 blocks containing 16384 samples each. To treat the random error
in signals, large number of averages is required. It means that more measurement time is
required. It will aid us in increasing block size and averages. Using these options, we
could get rid of the bias and random errors involved in the experiment.

55
CLOSING REMARKS

In this thesis, Output-Only modal analysis concept is discussed. Theoretical justifications

of different techniques to obtain modal parameters are outlined. Theory was implemented
in a simulation of 4 DOF system and experimental analysis of bolt channel joint
structure.

Results obtained from Output-Only modal analysis were compared with traditional
Modal analysis. Modal parameters obtained in both cases were found in a very good
agreement with each other. Results from this thesis work motivate the readers that the
proposed technique is applicable to wide range of structures excited by ambient
excitations. One disadvantage with ambient excitation is that most of energy is
concentrated in the lower frequency range. In this case, lower modes are the only modes
which are well excited. In such cases, accurate estimation of modes requires exciting the
structure with artificial exciters.

Applications of the technique are outlined in chapter 2. Many of the clear edges of
traditional modal analysis are covered by Output-Only modal analysis e.g. scaling factor,
accuracy, FE model updating etc. It marks that very soon Output-Only modal analysis
will become a widespread tool for solving many engineering problems.

56
REFERENCES

1 Maia, N. and Silva, J., Theoretical and Experimental Modal Analysis, John Wiley
& Sons, 1997

2 Clarence W de Silva, Vibration Fundamentals and Practice, CRC Press LLC

3 http://svibs.com/documentation/oma_modal.htm

4 Palle Anderson, Identification of Civil Engineering structures using ARMA

models, PhD thesis, Aalborg University, Denmark

5 Anders Brandt, Introductory Noise & Vibration Analysis, Saven Edu Tech AB
&The Department of Telecommunications and Signal Processing, Blekinge
Institute of Technology Sweden, 2003

6 Crawford, R. & Ward, H.S., Determination of the Natural Period of Buildings,

Bulletin of the Seismological Society of America, Vol.54, No.6, p.1743-1756,1964

7 Trifunac, M.D., Comparison between Ambient and Forced Vibration Experiments,

Earthquake Engineering and Structural Dynamics, Vol.1, p.133-150, 1972

8 Mc Lamore, V.R., Hart, G. & Stubbs, I.R., Ambient Vibration of Two Suspension
Bridges, Journal of the Structural Division, ASCE, Vol.97, N.ST10, p.2567-
2582,1971

9 Abdel-Ghaffar, A.M., Vibration Studies and Tests of a Suspension Bridge,

Earthquake Engineering and Structural Dynamics, Vol.6, p.473-496, 1978

10 Felber, A., Development of a Hybrid Bridge Evaluation System, Ph.D. Thesis,

University of British Columbia (UBC), Vancouver, Canada, 1993

11 Brincker, R., Zhang, L. & Andersen, P., Modal Identification from Ambient
Responses using Frequency Domain Decomposition, Proc. 18th International
Modal Analysis Conference, Kissimmee, USA, 2001

12 Brincker, R., Ventura, C. & Andersen, P., Damping Estimation by Frequency

Domain Decomposition, Proc. 19thInt. Modal Analysis Conference, San Antonio,
USA, 2000

13 Brincker, R., Krenk, S., Kirkegaard, P.H. & Rytter, A., Identification of the
Dynamical Properties from Correlation Function Estimates, Bygningsstatiske
Meddelelser, Danish Society for Structural Science and Engineering, Vol.63, N.1,
p.1-38, 1992

57
14 Asmussen, J.C., Modal Analysis based on the Random Decrement Technique.
Application to Civil Engineering Structures, Ph.D. Thesis, Univ. Aalborg, 1997

15 Peeters, B., System Identification & Damage Detection in Civil Engineering, PhD
thesis, Dept. of Civil Engineering, Katholieke Unviversiteit Leuven, Belgium
2000

16 Peeters, B. & De Roeck, G., Reference based stochastic subspace identification

for output-only modal analysis, Mech. Syst. Signal Process., 13 No. 6. pp 855-
878. 1999

17 Brown, D.L., Allemang, R.J., Zimmerman, R. & Mergeay, M., Parameter

Estimation Techniques for Modal Analysis, SAE Technical Paper Series,
N.790221, 1979

18 Vold, H., Kundrat, J., Rocklin, G.T. & Russel, R., A Multi-Input Modal
Estimation Algorithm for Mini-Computers, SAE Technical Paper Series,
N.820194, 1982

19 Van Overschee, P. & De Moor, B., Subspace Identification for Linear Systems:
Theory, Implementation, Applications, Kluwer Academic Publishers, The
Netherlands, 1996

20 Rodrigues, J., Brincker, R. & Andersen, P., Improvement of Frequency Domain

Output-Only Modal Identification from the Application of the Random Decrement
Technique, Proc. 23rd Int. Modal Analysis Conference, Deaborn, USA, 2004

21 Asemota, Osamede. Output only modal testing, Masters Thesis, Department of

Mech. Engineering, Blekinge Institute of Technology, Sweden (2005)

22 Peeters, B., Vanhollebeke, F. & Van der Auweraer, H., Operational Polymax for
Estimating the Dynamic Properties of a Stadium Structure during a Football
Game, Proc. 23rd Int. Modal Analysis Conference, Orlando, USA, 2005

23 Douglas L. Jones, Module, http://cnx.org/content/m12014/latest/

58
24 Hermans L., Guillaume P. and Van der Auweraer H., A frequency-domain
maximum likelihood approach for the extraction of modal parameters from
output-only data’, In Proc. of ISMA 23, Noise and Vibration Engineering,
K.U.Leuven, Belgium, 1998

25 James, G. H., Carne T.G. and Lauffer, J. P., The natural excitation technique
(NExT) for modal parameter extraction from operating structures, Modal
Analysis: the International Journal of Analytical and Experimental Modal, 1995

26 Ahlin, K., Brandt, A., Experimental Modal Analysis in Practice, Saven Edutech
AB, Täby, Sweden, 2001

27 Parloo, E., Application of frequency-domain system identification techniques in

the field of Operational Modal Analysis, PhD thesis, Vrije Universiteit Leuven,
Brussels, Belgium, 2003

28 Brincker, Rune and Andersen, Palle, A way of getting scaled mode shapes in
Output Only Modal Testing, In Proc. of the International Modal Analysis
Conference (IMAC), Kissimmee, Florida, USA. 2003

29 Dionisio Bernal and Burcu Gunes, Damage localization in Output-Only systems:

A Flexibility Based Approach. In Proc. of the International Modal Analysis
Conference (IMAC) XX, Los Angeles, California, pp 1185-1191. February 2002

30 Parloo, E., Verboven, P. Guillaume, Overmeire, M.Van, Sensitivity-Based

Operational Mode Shape Normalization. In Mech. Systems and Signal Proc., pp
757-767, 16(5), 2002

31 Ahlin, Kjell, Scaled Mode Shapes in Output Only Modal Testing – A simple
Example, SVIB AU1 Sympsium Riksgränsen April, 2005

32 FS Tse, IE Morse and RT Hinkle, Mechanical Vibrations- Theory and

Applications, Allyn and Bacon, Boston, Massachusetts, 1978

33 Meirovitch, L., Analytical Methods in Vibration, McMillan Co., New York, 1971

34 Krämer C., de Smet C.A.M. and Peeters B., Comparison of Ambient and Forced
Vibration Testing of Civil Engineering Structures, In Proceedings of the 17th
International Modal Analysis Conference (IMAC17), pp 1030-1034, Kissimmee,
USA, 1999

59
Department of Mechanical Engineering, Master’s Degree Programme Telephone: +46 455-38 55 10
Blekinge Institute of Technology, Campus Gräsvik Fax: +46 455-38 55 07
SE-371 79 Karlskrona, SWEDEN E-mail: ansel.berghuvud@bth.se