LAB 4 PATAMSETTI GANESH KUMAR

DISTRIBUTED SYSTEMS S296378

432 Dynamics of Engineering Systems

DISTRIBUTED SYSTEMS
EXPERIMENT

Laboratory Report 4

Group Members (2nd batch)

Ganesh kumar Patamsetti (296378)
Jahid Anwar Mollik (295715)
Hasan Iqbal (296147)

Submitted by: - Submitted to:-

Patamsetti Ganesh Kumar Dr.Stefanija Klaric
S296378

Date of Experiment Date of submission

8th -May-2018 15th-May-2018
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

COLLEGE OF ENGINEERING, IT & ENVIRONMENT

ASSESSMENT COVER SHEET
Student Name Patamsetti Ganesh Kumar
Student ID S296378
Assessment Title Distributed systems experiment
Unit Number and Title ENG432 -Dynamics of Engineering Systems
Lecturer/Tutor Dr.Stefanija Klaric
Date Submitted 15h-May-2018
Date Received

KEEP A COPY

Please be sure to make a copy of your work. If you have submitted assessment work
electronically make sure you have a backup copy.
PLAGIARISM
Plagiarism is the presentation of the work of another without acknowledgement. Students
may use a limited amount of information and ideas expressed by others, but this use must
be identified by appropriate referencing.
CONSEQUENCES OF PLAGIARISM
Plagiarism is misconduct as defined under the Student Conduct By-Laws. The penalties
associated with plagiarism are designed to impose sanctions on offenders that reflect the
seriousness of the University’s commitment to academic integrity.
* By submitting this assignment and cover sheet electronically, in whatever form you are deemed to have made the
declaration set out above.

Approver: Chair of Academic Board Review date declare that all material in this assessment is my
own work except where there is a clear acknowledgement and reference to the work of others. I
have read the University’s Academic and Scientific Misconduct Policy and understand its
implications. *
http://www.cdu.edu.au/governance/documents/3.3academicandscientificmisconduct.doc.

Signed…….PATAMSETTI GANESH KUMAR……...Date……..-15th -May-2018……..…….

LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

Contents
INTRODUCTION ....................................................................................................................................... 4
OBJECTIVES ............................................................................................................................................. 4
APPARATUS USED: .................................................................................................................................. 4
EXPERIMENTAL PROCEDURE: ................................................................................................................. 4
Transverse vibration (accelerometer top-middle):............................................................................. 4
Transverse vibration (accelerometer top-almost end): ...................................................................... 5
Longitudinal vibration (accelerometer sideway): ............................................................................... 5
ANALYTICAL CALCULATION: .................................................................................................................... 5
Longitudinal vibration: ........................................................................................................................ 5
Transverse vibration: .......................................................................................................................... 6
EXPERIMENTAL CALCULATION:............................................................................................................... 7
MATLAB for transverse position when accelerometer at middle-top ................................................ 7
MATLAB for transverse position when accelerometer at near end-top........................................... 10
MATLAB for longitudinal position when accelerometer at sideway ................................................ 12
COMPARISON OF TWO METHODS:....................................................................................................... 14
DISCUSSION QUESTIONS:...................................................................................................................... 15
Methodology..................................................................................................................................... 15
Error between two methods............................................................................................................. 15
5th and 6th transverse analytical results ............................................................................................ 15
Code .............................................................................................................................................. 15
Solution: ........................................................................................................................................ 15
Where same sort of systems are used in industries ......................................................................... 16
How confident were you in predicting natural frequencies of a system without experiment
results? .............................................................................................................................................. 16
Possible ways to increase the natural frequency of an existing structure ....................................... 16
CONCLUSION:........................................................................................................................................ 17
REFERENCES: ......................................................................................................................................... 18
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

INTRODUCTION:
The results are more precise for any system if it has multiple masses, dampers and springs (more
degrees of freedom). When the degrees of freedom are extended without limit, a system is formed
with distributed mass and elasticity. This concept is called the distributed systems. In our
experiment, we consider rectangular beam as the distributed system. As distributed systems have
unlimited vibration frequencies, it is necessary to find the mode at a given point of time. The natural
frequency of the distributed system for longitudinal and transverse can be found by 2 methods
(analytical and experimental).The experimental procedure is by conducting vibration analysis of the
force that was given by hammer and the output given by the accelerometer by conducting the
experiment for both longitudinal and transverse vibrations. The analytical procedure is by using the
Euler-Bernoulli beam theory assumptions and formulas. These 2 methods help to find the natural
frequency of the distributed system at a given time.

OBJECTIVES:
 To find the natural frequencies of the distributed system using the transfer function for
hammer and accelerometer placed at middle of beam for transverse vibrations.
 To find the natural frequencies of the distributed system using the transfer function for
hammer and accelerometer placed at almost end of beam for transverse vibrations.
 To find the natural frequencies of the distributed system using the transfer function for
hammer and accelerometer placed at side of beam for longitudinal vibrations.
 To calculate the analytical vibration frequencies using the theory.
 Compare results of both methods.

APPARATUS USED:
Mild steel rectangular beam (free-free arrangement)

Vernier callipers

Measuring tape

Hammer with rubber tip sensor (Kister Company for transverse vibrations)

Hammer with steel tip sensor (Kjael and Bruer Company for longitudinal vibrations)

Accelerometer sensor

DLC controller and monitor

EXPERIMENTAL PROCEDURE:
Transverse vibration (accelerometer top-middle):
The distributed system (mild-steel rectangular beam) is arranged in free-free position with the help
of hooks on both sides. The measurements of the beam are noted. An accelerometer sensor which is
connected to the PLC (Programmable Logic Controller) is placed at the top surface and at the centre
of the beam. Now, a force is induced with the help of rubber tip hammer on the top surface of one
end of beam and the vibration is sensed by connecting the tip sensor to the PLC controller. The tip of
the hammer was made of rubber and it was flat. The vibrations produced will induce the piezo
electric materials of the accelerometer to generate an electric impulse which is forwarded to the
PLC. The PLC controller converts these input voltage signals to the digital signals where a plot is
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

displayed in the monitor with accelerometer and impact hammer sensor data. This data is saved for
further analysis in MATLAB to determine the natural frequencies of the system at a given time.

Transverse vibration (accelerometer top-almost end):

The arrangement is same as of the transverse vibration arrangement like discussed before except
the accelerometer is placed at almost near end of the beam. The impact is given at the other end
and the readings are recorded as previous process. The natural frequencies are found by studying
the Bode plot of MATLAB by using the generated sensor data while conducting the experiment.

Longitudinal vibration (accelerometer sideway):

The arrangement for inducing longitudinal vibration is same as like transverse vibration except 2
changes. The accelerometer sensor is placed sideway of one end of beam and impact created by
hammer on the other sideway of the beam. The hammer used to induce the vibration is different for
both transverse and longitudinal vibrations. Steel tipped hammer with hexagonal faced is used in
case of longitudinal vibration analysis. Sensor readings are recorded and saved to analyse in the
MATLAB and to determine the natural frequencies of the system at a given point of time.

ANALYTICAL CALCULATION:
Longitudinal vibration:
(From 10.3-1 of Mechanical Vibration, William J Palm III)

The distributed system considered for experiment is of free-free arrangement. Hence the
calculations are to be made for free-free arrangement.

Properties of the beam (mild-steel)

Density, ρ=7800 kg/m3

Young’s Modulus, E= 2*1011 Pa

Longitudinal equation of motion,

𝐸
C2=
𝜌
2∗10^11
C2=
7800

C=5063.697 m/sec

The frequencies can be measured by,

𝑛𝜋𝑐
ωn= where n=0, 1,2,3….
𝐿
From the above,
1∗𝜋∗5063.697
ω1= =15908.07 rad/sec=2531.848 Hz
1
(The calculations are provided in the excel sheet attached)
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

Frequency,ωn rad/sec Hz
ω1 15908.07 2531.848
ω2 31816.15 5063.697
ω3 47724.22 7595.545
ω4 63632.29 10127.39
ω5 79540.36 12659.24
ω6 95448.44 15191.09

The above table gives the first 6 natural frequencies for longitudinal vibration

Transverse vibration:
(From 10.4-2 of Mechanical Vibration, William J Palm III)

The distributed system considered for experiment is of free-free arrangement. Hence the
calculations are to be made for free-free arrangement.

Properties of the beam (mild-steel)

Density, ρ=7800 kg/m3

Young’s Modulus, E= 2*1011 Pa

Length of beam, L=1 m

Breadth of beam, b=0.025 m

Height of beam, h=0.025 m

𝑏ℎ^3 0.025∗0.025^3
Moment of inertia, I= = =3.2552*10-8 kg-m2
12 12
Area of the beam, A=L*b*h=1*0.025*0.025=0.000625 m3

Transverse equation of motion,

𝐸𝐼
C2=
𝜌𝐴

2∗1011 ∗3.2552∗10−8
C2=
7800∗0.000625

C=36.544 m/sec

The first 4 values of Beta are taken from the text book and next 2 values are calculated in the
MATLAB (code provided as an attachment as it is further required for discussion questions).

β1 0
β2 4.73004
β3 7.8532
β4 10.99561
β5 14.1372
β6 17.2788
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

ωn=cβn2
Now,

ω1=36.544*(0)2=0

Frequency,ωn rad/sec Hz
ω1 0 0
ω2 817.611 130.1268271
ω3 2253.774 358.6993004
ω4 4418.305 703.1951536
ω5 7303.716 1162.422527
ω6 10910.49 1736.458343
The above table gives the first 6 natural frequencies for transverse vibration

The first 6 natural frequencies of the distributed systems are found out for both longitudinal and
transverse vibrations.

EXPERIMENTAL CALCULATION:
MATLAB for transverse position when accelerometer at middle-top
closeall; clear all; clc;

%% accelerometer data
load'transverse_2.txt'
Fs =5000;%sample frequency from experiment

% Draw your data in time domain

t=linspace(0,2,10000);
output=((transverse_2(:,2))./9.49);%Divided with sensitivity of
accelerometer
plot(t,output);%plotting time in terms of samples(10000 for 2 secs)
xlabel('Time(secs)');
ylabel('Acceleration(m/s^2)');
title('Time domain when accelerometer placed at middle(Transverse)');

%% fft of the input

L=length(transverse_2);%10000 samples
input=((transverse_2(:,1))./9.08);%divided with sensitivity of hammer(Grey
Kistler)
NFFTInput = 2^(nextpow2(length(transverse_2)));
BFFTInput = fft(input,NFFTInput);
f = Fs*(0:NFFTInput-1)/NFFTInput;

% do the fft for the output

NFFTOutput=2^(nextpow2(length(transverse_2)));
BFFTOutput=fft(output,NFFTOutput);

%transfer function
TF = 20*log10(BFFTOutput./BFFTInput); %Creating a experimental transfer
function
figure;
semilogx(f,TF,'b');%Using to plot logarithmic magnitude values in Y axis
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378
[TF2, F] = tfestimate(input,output,[],[],[],Fs);%Estimating the transfer
functions for given input(hammer) and output(accelerometer) till the
samples end
holdon;
semilogx(F,20*log10(abs(TF2)),'r');

title('Bode plot when accelerometer placed at middle(Transverse)');

xlabel('Frequency(Hz)');
ylabel('Magnitude(dB)');
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

MATLAB for transverse position when accelerometer at near end-top

closeall; clear all; clc;

%% accelerometer data
load'transverse_6.txt'
Fs =5000;%sample frequency from experiment

% Draw your data in time domain

t=linspace(0,2,10000);
output=((transverse_6(:,2))./9.49);%Divided with sensitivity of
accelerometer
plot(t,output);%plotting time in terms of samples(10000 for 2 secs)
xlabel('Time(secs)');
ylabel('Acceleration(m/s^2)');
title('Time domain when accelerometer placed at end(Transverse)');

%% fft of the input

L=length(transverse_6);%10000 samples
input=((transverse_6(:,1))./9.08);%divided with sensitivity of hammer(Grey
Kistler)
NFFTInput = 2^(nextpow2(length(transverse_6)));
BFFTInput = fft(input,NFFTInput);
f = Fs*(0:NFFTInput-1)/NFFTInput;

% do the fft for the output

NFFTOutput=2^(nextpow2(length(transverse_6)));
BFFTOutput=fft(output,NFFTOutput);

%transfer function
TF = 20*log10(BFFTOutput./BFFTInput); %Creating a experimental transfer
function
figure;
semilogx(f,TF,'b');%Using to plot logarithmic magnitude values in Y axis

[TF2, F] = tfestimate(input,output,[],[],[],Fs);%Estimating the transfer

functions for given input(hammer) and output(accelerometer) till the
samples end
holdon;
semilogx(F,20*log10(abs(TF2)),'r');

title('Bode plot when accelerometer placed at end(Transverse)');

xlabel('Frequency(Hz)');
ylabel('Magnitude(dB)');
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

MATLAB for longitudinal position when accelerometer at sideway

closeall; clear all; clc;

%% accelerometer data
load'long_2.txt'
Fs =5000;%sample frequency from experiment

% Draw your data in time domain

t=linspace(0,2,10000);
output=((long_2(:,2))./9.49);%Divided with sensitivity of accelerometer
plot(t,output);%plotting time in terms of samples(10000 for 2 secs)
xlabel('Time(secs)');
ylabel('Acceleration(m/s^2)');
title('Time domain for longitudinal arrangement');

%% fft of the input

L=length(long_2);%10000 samples
input=((long_2(:,1))./22.65);%divided with sensitivity of
hammer(Bruel&Kjaer)
NFFTInput = 2^(nextpow2(length(long_2)));
BFFTInput = fft(input,NFFTInput);
f = Fs*(0:NFFTInput-1)/NFFTInput;

% do the fft for the output

NFFTOutput=2^(nextpow2(length(long_2)));
BFFTOutput=fft(output,NFFTOutput);

%transfer function
TF = 20*log10(BFFTOutput./BFFTInput); %Creating a experimental transfer
function
figure;
semilogx(f,TF,'b');%Using to plot logarithmic magnitude values in Y axis

[TF2, F] = tfestimate(input,output,[],[],[],Fs);%Estimating the transfer

functions for given input(hammer) and output(accelerometer) till the
samples end
holdon;
semilogx(F,20*log10(abs(TF2)),'r');

title('Bode plot for longitudinal arrangement');

xlabel('Frequency(Hz)');
ylabel('Magnitude(dB)');
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

The time domains and the bode plots are made using the MATLAB by analysing the sensor data of
accelerometer and impact hammer.
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

COMPARISON OF TWO METHODS:

A comparison is made between the 2 methods in the excel sheet. The error percentages are very
minimal which shows that the values are more accurate and precise. The minimum error is also due
to non-isolation of system to surroundings, levelling of the beam while conducting experiment,
impact given by the hammer, human errors and constant values taken for mild steel. The values of
frequencies for longitudinal is very high in experimental and cannot be seen on graph because the
steel tip hammer was not able to induce the force that can create more than 5000 Hz frequency. The
method and the approach are correct and the hammer needs to be replaced to see more natural
longitudinal vibrations on the graph. The transverse vibrations are almost perfect compared to the
analytical values as the error percentage is very low.

Transverse vibration(accelerometer at middle)

Error
Frequency,ωn(Hz) Analytical Experimental percentage
ω1 0 0
ω2 130.1268271 128.2 1.480730131
ω3 358.6993004 352.8 1.644636709
ω4 703.1951536 682.4 2.957237904
ω5 1162.422527 1119 3.735520117

Transverse vibration(accelerometer at end)

Error
Frequency,ωn(Hz) Analytical Experimental percentage
ω1 0 0
ω2 130.1268271 127 2.402907384
ω3 358.6993004 355.2 0.975552605
ω4 703.1951536 681.2 3.127887544
ω5 1162.422527 1115 4.079629071

Longitudinal vibration(accelerometer at
sideway)
Error
Frequency,ωn(Hz) Analytical Experimental percentage
ω1 2531.848418 2483 1.92935791
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

DISCUSSION QUESTIONS:
Methodology
The methodology is discussed in brief in the experiment procedure section.

Error between two methods

Comparison is made between two methods and well discussed in the previous section.

5th and 6th transverse analytical results

In the textbook there are only first 4 values of β are given. In order to find the 5th and 6th transverse
frequency we need to use the MATLAB code. (MATLAB code from learn line and textbook page No
628, Mechanical Vibration, William J Palm III)

Code:
closeall; clear all; clc;
%% Solving cosbnL*coshbnL = 1

f = @(x) cos(x).* cosh(x) - 1;

nb = 1;
for k = 0:25
if f(k)*f(1+k)<0
nb = nb+1;
zerocross(nb) = k;
end
end

first= fzero(f, zerocross(1))

first1= fzero(f, [zerocross(1) zerocross(1)+1])
second=fzero(f, zerocross(2))
third=fzero(f, zerocross(3))
fourth=fzero(f,zerocross(4))
fifth=fzero(f,zerocross(5))
sixth=fzero(f,zerocross(6))
xl = linspace(0,25);
fl = f(xl);

plot(xl,fl)

Solution:

First = 0

first1 = 0

Second = 4.7300

Third 7.8532

Fourth = 10.9956

Fifth = 14.1372

Sixth 17.2788
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

>>

The β values of 5th and 6th are 14.1372 and 17.2788.

We know for transverse vibrations, ωn=C*βn2

We already know c=36.544 m/sec

So the equation becomes,

ω5=36.544*β52=36.544*(14.1372)2=7303.699 rad/sec=1162.419 Hz
ω6=36.544*β62=36.544*(17.2788)2=10910.46 rad/sec=1736.45 Hz
Where same sort of systems are used in industries
 It is essential to know the correct natural frequencies of a machine to select the best damper
possible to limit the vibrations passing to the surroundings. This method is mainly helpful to
determine the frequencies and thereby isolation of vibration.
 These systems are mainly used to reduce the vibrations in the rotating equipment like
compressors, turbines using the mass imbalance due to gravitational force.
 Fouling in pipes is found out using this type of impact hammer test using the mass and
elasticity distribution.
 Detection of cracks can be done using this system as a part of NDT techniques.

How confident were you in predicting natural frequencies of a system without

experiment results?
It is possible to detect the natural frequencies of a system without conducting the experiment but
the system should be bound to the following assumptions as mentioned in the Euler-Bernoulli beam
theory like(page no 619, Mechanical Vibration, William J Palm III)

 No axial load
 Beam thickness to width ratio should be small
 Uniform properties of beam
 Beam should be linear, isotropic and should obey principles of Hooke’s law

It is not possible in real life scenarios to have the systems with all the conditions satisfied so the
experimental method is mostly used depending on the scenarios and is cross checked with the
analytical results just like in our experiment to make sure that we are in the right track of finding the
correct natural frequencies. Large deviations in the values prove that the procedure used is at fault.

Possible ways to increase the natural frequency of an existing structure

In general,
𝑘
n=√
𝑚

Where n is natural frequency, k is stiffness and m is mass.

This relation gives us the understanding that the mass of the system should be reduced or the
stiffness of the system should be increased in order to improve the natural frequency of the system.

The other relation that can be used for distribution systems in terms for finding natural frequency
are,
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

n=c*2 and
𝐸𝐼
c=√
𝐴

Where n is natural frequency, c is propagation speed, and  is the lowest normalised frequencies
when divided in terms of its length.

So, increasing the moment of inertia and Young’s modulus or decreasing the density and area of the
system increases the natural frequency of the structure.

CONCLUSION:
The 2 methods that are used to find the natural frequency of a system are well studied. The
assumptions and the conditions to be followed to use the analytical method is well understood. The
use of transfer functions like tfestimate, fft to determine the sensor data for accelerometer and
impact hammer are studied. The results are almost the same for both the methods proves that the
system was well adjusted to all the assumptions. The selection of hammers is critical for both
longitudinal and transverse as the hammers should be able to induce a force that gives the desired
frequency range. Usage of this method in industries for fouling and crack detection is observed in
practical research.
LAB 4 PATAMSETTI GANESH KUMAR
DISTRIBUTED SYSTEMS S296378

REFERENCES:
PALM, W. J. 2007. Mechanical vibration, Hoboken, NJ, Hoboken, NJ : John Wiley

Perso.univ-rennes1.fr. (2018). [online] Available at: https://perso.univ-

rennes1.fr/lalaonirina.rakotomanana-ravelonarivo/Stokey_chapter7.pdf [Accessed 13 May 2018].

Www-mdp.eng.cam.ac.uk. (2018). [online] Available at: http://www-

mdp.eng.cam.ac.uk/web/library/enginfo/cueddatabooks/materials.pdf [Accessed 13 May 2018].

Properties, L. (2018). Semilogarithmic plot - MATLAB semilogx- MathWorks Australia. [online]

Au.mathworks.com. Available at: https://au.mathworks.com/help/matlab/ref/semilogx.html
[Accessed 13 May 2018].

Au.mathworks.com. (2018). Transfer function estimate - MATLAB tfestimate- MathWorks Australia.

[online] Available at: https://au.mathworks.com/help/signal/ref/tfestimate.html [Accessed 13 May
2018].

Bksv.com. (2018). [online] Available at: https://www.bksv.com/media/doc/br0094.pdf [Accessed 13

May 2018].

Homepages.cae.wisc.edu. (2018). [online] Available at:

http://homepages.cae.wisc.edu/~zhous/papers/final.pdf [Accessed 13 May 2018].

Lemos, Luiz & Neto, J & Silva, J.J. & Rocha Neto, Jose Sergio. (2011). Fouling detection using hammer
impact test and wireless comunication. Conference Record - IEEE Instrumentation and Measurement
Technology Conference. 1-5. 10.1109/IMTC.2011.5944168.

Silva, J., Neff, F., Lima, A. and Rocha Neto, J. (2011). Hammer impact test applied for fouling
detection in pipelines. Sba: Controle & Automação Sociedade Brasileira de Automatica, 22(6),
pp.620-630.