Detection of Cracks in a Cantilever Metal Plate Using

Vibrational Techniques

By
Sharif Rezwan Shuvo
Student Number – 171011005

Supervised By
Dr. Md. Zahid Hossain

A thesis submitted to the Department of Mechanical and Production Engineering (MPE) in

partial fulfillment of the requirement for the degree of Master of Science in Mechanical
Engineering

Department of Mechanical and Production Engineering (MPE)

Islamic University of Technology (IUT)
September 2020

1
 Certificate of Approval
The thesis titled “Detection of Cracks in a Cantilever Plate Using Vibrational Techniques”
submitted by Sharif Rezwan Shuvo bearing student number 171011005 of Academic Year
2019-2020 has been found as satisfactory and accepted as partial fulfillment of the requirement
for the degree of Master of Science in Mechanical Engineering on 30 April, 2020.

Board of Examiners

…………………………….
Dr. Md. Zahid Hossain Chairman
Head, Professor, (Supervisor)
Department of Mechanical and Production Engineering (MPE) and
Islamic University of Technology (IUT) Member
Board Bazar, Gazipur, (Ex-officio)
Bangladesh

…………………………….
Dr. Md. Nurul Absar Chowdhury
Professor,
Department of Mechanical and Production Engineering (MPE) Member
Islamic University of Technology (IUT)
Board Bazar, Gazipur,
Bangladesh

…………………………….
Dr. Annayet Ullah Patwari
Professor,
Department of Mechanical and Production Engineering (MPE) Member
Islamic University of Technology (IUT)
Board Bazar, Gazipur,
Bangladesh

…………………………….
Dr. Md. Abdus Salam Akanda
Professor, Member
Department of Mechanical Engineering (ME), (External)
Bangladesh University of Engineering and Technology (BUET)
Dhaka,
Bangladesh

i
 Candidate’s Declaration

It is hereby declared that this thesis or any part of it has not been submitted elsewhere for the
award of any degree or diploma.

Signature of the Candidate

…………………………….
Sharif Rezwan Shuvo
Department of Mechanical and Production Engineering (MPE)
Islamic University of Technology (IUT)
Board Bazar, Gazipur,
Bangladesh.

Signature of the Supervisor

…………………………….
Dr. Md. Zahid Hossain
Head, Professor,
Department of Mechanical and Production Engineering (MPE)
Islamic University of Technology (IUT)
Board Bazar, Gazipur,
Bangladesh.

ii
 Dedication

This thesis is dedicated to my parents and family.

iii
 Acknowledgement
The author is highly grateful to Almighty Allah for successful completion of the postgraduate

thesis. The journey was not an easy one, rather it contained with lots of hard working, stressful

office hours, many sleepless night and dedication, love and care of the family and the

professors.

The author puts deepest gratitude to his supervisor, Dr. Zahid Hossain, Head, Professor of

Mechanical and Production Engineering (MPE), Islamic University of Technology (IUT) for

his constant guidance, recommendation and inspiration. The author will also not forget the

tremendous effort of Mr. Nagib Mehfuz, Lecturer of Department of Mechanical and Production

Engineering (MPE), Islamic University of Technology (IUT), for every stages of this thesis

work.

Finally, the author is thankful to his beloved mother and family for providing financial and

mental support and encouragement.

Despite of the careful and precise work, author ask forgiveness to readers for any mistakes,

whatsoever, found in his report beforehand.

iv
 Abstract
The main objective of this research is to detect crack at different positions of the thin metal

cantilever plate based on the natural frequency. Most of the researchers have investigated on

beam. Few researchers have investigated on thin plate. The crack depth and crack thickness in

standard convention of beam are considered as crack length and depth of crack respectively in

this research for the thin metal plate. Some theoretical approach considering bilinear stiffness

have been found in some literatures. Numerical and experimental analysis due to crack in

cantilever beam or plate have been investigated by different researchers.

In this research work, a thin plate is considered made of aluminum alloy and structure steel. A

slit crack is considered for different cases of investigations. The crack length, width, depth,

location, etc. are taken as parameters. These are investigated on the change of natural

frequencies through modal analysis due to the mentioned parameters change.

The study confirms that when the crack length affects the most in frequency drop than that of

crack width and crack depth. The drop of frequency is massive when the crack length increases.

However, when the crack width increases the drop of frequency remains constant. Moreover,

thorough crack on a plate shows more drop of frequency than that of a plate with crack, which

is not thorough. Therefore, thorough cracks on plates are more detectable

Along the longitudinal direction (Z-axis), the effect of transverse edge crack is more significant

and visible (with large frequency drop) than that of longitudinal crack (surface defect). The

maximum drop in frequency in both the bending and torsional modes is shifted to free end

along the relative distance on the plate. While along transverse direction (X-axis) of free end

of the thin plate, the effect of longitudinal edge cracks are visible in torsional modes with

massive frequency drop. In this case, the frequency of transverse cracks (surface defects) found

in bending modes is greater than healthy plate and longitudinal curve. Surface defects, i.e., the

v
longitudinal crack along the longitudinal direction and the transverse crack along the transverse

direction, along the edge of the plate remain undetectable. Although, the longitudinal

embedded crack remain undetectable, the transverse embedded crack is traceable by the aid of

bending modes only. Angle between 75o to 105o of the inclined cracks are more prone to

detection. The effect of stiffness on the crack position on the plate near a bend or a twist (found

in mode shape) is greater than any other region. Crack located in this region is more likely to

be detected.

In static structural analysis along the longitudinal direction, stress concentration and total

deformation is maximum while the first natural frequency is the minimum at fixed end. While

at free end stress and deformation becomes minimum but first natural frequency becomes

maximum. It is also found that, for healthy plate and plate with crack near free ends (with

relative distance > 70%), the maximum von-Mises stress is located near fixed region of the

plate. While for crack positions near fixed end and at mid-region (0% ≤ relative distance ≤

70%), the maximum von-Mises stress is located at the tip of the crack. It is also observed that

the pattern of normalized frequency curve in metal (structural steel) and metal alloy (aluminum

alloy) are same. An analytical FEM approach has been performed for without crack condition

and numerical validation with published paper has been done and found in good agreement.

vi
 Table of Contents

Certificate of Approval ............................................................................................................. i

Candidate’s Declaration ..........................................................................................................ii

Dedication ............................................................................................................................... iii

Acknowledgement ................................................................................................................... iv

Abstract ..................................................................................................................................... v

Table of Contents ...................................................................................................................vii

List of Tables ............................................................................................................................ x

List of Figures .......................................................................................................................... xi

Nomenclature and Terminology .......................................................................................... xiv

Chapter 1 Introduction .......................................................................................................... 1

1.1 Definition of Crack and Breathing Crack ........................................................................ 2

1.2 Why Crack Detection is Important? ................................................................................. 2

1.3 Research Objectives ......................................................................................................... 2

1.4 Possible Outcomes ........................................................................................................... 3

1.5 Outline of Methodology ................................................................................................... 3

1.5.1 Step 1 ............................................................................................................................. 3

1.5.2 Step 2 ............................................................................................................................. 4

1.5.3 Step 3 ............................................................................................................................. 4

Chapter 2 Literature Review ................................................................................................. 5

vii
Chapter 3 Modeling, Numerical Analysis and Problem Specification.............................. 11

3.1 Modeling of Cantilever Plate ......................................................................................... 12

3.2 Numerical Analysis ........................................................................................................ 14

3.2.1 Modal Analysis ........................................................................................................... 14

3.2.2 Static Structural Analysis ............................................................................................ 15

3.2.3 Boundary Condition and Meshing .............................................................................. 16

3.3 Convention for Measuring Plate and Beam Dimension ................................................. 17

3.4 Problem Specification .................................................................................................... 18

Chapter 4 Results of Numerical Simulation and Discussion ............................................. 21

4.1 Validation of the Scientific Work .................................................................................. 22

4.2 Theoretical Calculation for Bending Vibration Beam Element Theory......................... 26

4.3 Case-1: Change of Transverse Edge Crack Positions along Z-axis ............................... 28

4.4 Case-2: Change of Transverse Crack Lengths on the Edge along Z-axis ...................... 34

4.4.1 Further Study on Different Crack Lengths .................................................................. 38

4.5 Case-3: Change of Transverse Crack Widths on the Edge along Z-axis ....................... 42

4.5.1 Further Study on Different Crack Widths ................................................................... 46

4.6 Case-4: Study on Effect of Normalized Frequency due to Change of Crack Depths

along Y-axis ......................................................................................................................... 49

4.7 Case-5: Transverse Edge Crack Positions along Z-axis with Structural Steel ............... 52

4.8 Case-6: Change of Embedded Transverse Crack Positions along Mid Z-axis............... 55

4.9 Case-7 and Case-8: Change of Longitudinal Crack at Edge and Mid-Positions along Z-

axis ....................................................................................................................................... 58

viii
 4.10 Case-9 and Case-10: Change of Longitudinal and Transverse Edge Crack Positions

along X-axis on Free End ..................................................................................................... 61

4.11 Case-11 and Case-12: Inclined Crack Positions at Relative Distance of 21% and 79%

along Z-axis .......................................................................................................................... 65

4.12 Equivalent von-Mises Stress and Total Deformation .................................................. 69

4.13 Mode Shapes, Stress Concentration and Total Deformation at Three Regions ........... 71

4.13.1 Stress Concentration and Total Deformation on Healthy Plate ................................ 72

4.13.2 Mode Shape of First Natural Frequency ................................................................... 72

4.13.3 Stress Concentration and Total Deformation Near Fixed End Region ..................... 74

4.13.4 Stress Concentration and Total Deformation at Mid-Region .................................... 76

4.13.5 Stress Concentration and Total Deformation Near Free End .................................... 77

4.13.6 Other Mode Shapes at Three Regions ....................................................................... 78

Chapter 5 Detection of Cracks.............................................................................................. 85

5.1 Detection of Cracks of all Cases .................................................................................... 86

5.2 Experimentation ............................................................................................................. 89

Chapter 6 Conclusion and Recommendation ...................................................................... 90

6.1 Conclusion ...................................................................................................................... 91

6.2 Recommendation ............................................................................................................ 94

References ............................................................................................................................... 95

ix
 List of Tables

Table 3.1: Dimension and Material Properties of the Plate and the Crack .............................. 13

Table 3.2: Important Steps of Numerical Analysis Procedures ............................................... 14

Table 3.3: Nomenclature for Beam and Crack According to Standard Convention ............... 17

Table 3.4: Brief Case Descriptions of all the Cases................................................................. 19

Table 4.1: Validation Result of Specimen-2 & Specimen-3.................................................... 23

Table 4.2: Validation Result of Specimen-4 & Specimen-5.................................................... 24

Table 4.3: Key Findings of Case-2 for Different Crack Lengths ............................................ 37

Table 4.4: Seven Samples of Crack Dimension with Various Crack Lengths ........................ 38

Table 4.5: Frequency Drop for Cracks at Three Distinct Regions for Different Crack Length

.................................................................................................................................................. 41

Table 4.6: Key Findings of Case-3 for Different Crack Widths .............................................. 45

Table 4.7: Five Samples of Crack Dimension with Various Crack Widths ............................ 46

Table 5.1: Summary of Crack Detection in the Research ........................................................ 86

x
 List of Figures

Figure 3.1: SOLIDWORKS Design of Thin Metal Plate ........................................................ 12

Figure 3.2: Schematic Diagram for Indicating Fixed End and Application of Load on Metal

Plate.......................................................................................................................................... 15

Figure 3.3: Meshing on Metal Plate ......................................................................................... 16

Figure 3.4: Convention for Measuring Beam and Crack Dimension ...................................... 17

Figure 3.5: Standard Used in this Research for Thin Plate ...................................................... 18

Figure 3.6: Illustration of All the Cases ................................................................................... 19

Figure 4.1: Illustration of Specimen from Scientific Paper [3] (a) Specimen 1# Plate with No

Crack (b) Specimen 2# Plate with Crack Size 20mm (c) Specimen 3# Plate with Crack Size

40mm (d) Specimen 4# Plate with Crack Size 60mm (e) Specimen 5# Plate with Crack Size

70mm ....................................................................................................................................... 22

Figure 4.2: Comparison of Mode Shape for Specimen-1 ........................................................ 25

Figure 4.3: Comparison of Mode Shape for Specimen-5 ........................................................ 25

Figure 4.4: Uniform Beam Element [9] ................................................................................... 26

Figure 4.5: Schematic Diagram for the Change in Transverse Crack Positions along Z-axis. 28

Figure 4.6: Mode Shape Deformation of a Healthy Metal Plate ............................................. 30

Figure 4.7: Normalized Frequency over Relative Distance along Z-axis ................................ 33

Figure 4.8: Schematic Diagram for the Change of Transverse Crack Length on the Edge

along Z-axis ............................................................................................................................. 34

Figure 4.9: Normalized Frequency over Relative Distance along Z-axis for Transverse Crack

Length of 10mm, 20mm and 30mm ........................................................................................ 36

Figure 4.10: Normalized Frequency Curve over Different Crack Length to Plate Length Ratio

.................................................................................................................................................. 40

xi
Figure 4.11: Schematic Diagram for the Change of Transverse Crack Width on the Edge

along Z-axis ............................................................................................................................. 42

Figure 4.12: Normalized Frequency over Relative Distance along Z-axis for Transverse

Crack Width of 0.5mm, 1mm and 1.5mm ............................................................................... 44

Figure 4.13: Normalized Frequency Curve over Different Crack Width to Plate Length Ratio

.................................................................................................................................................. 48

Figure 4.14: Normalized Frequency Curve over Different Crack Depth to Plate Thickness

along Y-axis ............................................................................................................................. 51

Figure 4.15: Normalized Frequency Between Aluminum Alloy (Case-1) and Structural Steel

(Case-5) .................................................................................................................................... 54

Figure 4.16: Schematic Diagram for the Change in Embedded Transverse Crack Positions

along Mid Z-axis of the Plate................................................................................................... 55

Figure 4.17: Difference in Normalized Frequency Between Transverse Edge Crack (Case-5)

and Transverse Mid-Crack (Case-6) along Z-axis ................................................................... 57

Figure 4.18: Schematic Diagram for the Change in Positions along the Z-axis of the Plate for

Longitudinal Edge Crack (Case-7) and Embedded Mid-Crack (Case-8) ................................ 58

Figure 4.19: Difference in Normalized Frequency Between Longitudinal Edge Crack (Case-

7) and Longitudinal Mid-Crack (Case-8) along Z-axis ........................................................... 60

Figure 4.20: Schematic Diagram for the Change in Longitudinal (Case-9) and Transverse

(Case-10) Crack Positions along the X-axis on Free End........................................................ 61

Figure 4.21: Difference in Normalized Frequency Between Longitudinal (Case-9) and

Transverse (Case-10) Crack Positions along the X-axis on Free End ..................................... 64

Figure 4.22: Schematic Diagram for the Change in Inclined Crack at x/L=21% (Case-11) and

x/L=79% (Case-12) along the Z-axis....................................................................................... 65

xii
Figure 4.23: Difference in Normalized Frequency Between Change in Inclined Crack at

x/L=21% (Case-11) and x/L=79% (Case-12) along the Z-axis ............................................... 68

Figure 4.24: Maximum Equivalent von-Mises Stress and Deformation ................................. 69

Figure 4.25: Maximum Deformation for Modal Analysis ...................................................... 70

Figure 4.26: Normalized Frequency of Mode 1 and Maximum Equivalent von-Mises Stress71

Figure 4.27: Deformation and von-Mises Stress of a Healthy Plate from Static Structural

Analysis.................................................................................................................................... 72

Figure 4.28: Mode Shape of First Natural Frequency at Three Regions ................................. 73

Figure 4.29: Deformation and von-Mises Stress of a Crack at Relative Distance of 9.1% Near

Fixed End from Static Structural Analysis .............................................................................. 75

Figure 4.30: Deformation and von-Mises Stress of a Crack at Relative Distance of 50% at

Mid-Region from Static Structural Analysis ........................................................................... 76

Figure 4.31: Deformation and von-Mises Stress of a Crack at Relative Distance of 84.8% at

Free End from Static Structural Analysis ................................................................................ 78

Figure 4.32: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of 9.1%

Near Fixed End ........................................................................................................................ 80

Figure 4.33: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of 50% at

Mid-Region .............................................................................................................................. 82

Figure 4.34: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of 84.8%

at Free End ............................................................................................................................... 84

xiii
 Nomenclature and Terminology
Particulars Nomenclature
Length of the Plate L or l
Width of the Plate B
Thickness of the Plate H
Length of the Crack a
Width of the Crack b
Depth of the Crack c
Location of Crack along Z-axis x
Location of Crack along X-axis y
Natural Frequency f or 𝜔
Natural Frequency of healthy plate fo
Normalized Frequency f/fo or p
Relative Distance along Z-axis x/L or n
Relative Distance along X-axis y/B
Crack Length to Plate Length Ratio a/L
Crack Width to Plate Length Ratio b/L
Crack Depth to Plate Thickness Ratio c/H
Mass of the beam m
Density of the material 𝜌
Cross section Area A
Young’s Modulus E
Aluminum Alloy AA
Structural Steel SS

xiv
 Chapter 1
Introduction

1
1.1 Definition of Crack and Breathing Crack
A crack is a discontinuity of stress concentration in a solid body creating some separation extent

(crack length and depth) leading to metal failure (crack propagation) due to material separation

by opening or sliding effects. For a beam to fail, cracks can occur due to increased shear stress

or bending stress, compression failure (mechanical or thermal fatigue) and corrosion (corrosion

fatigue). The number, position (embedded or edged), size (crack depth and length) and mode

(opening, in-plane shear and out-of-plane shear) of the crack plays important role in detection

of the crack. In mechanical vibration, considering only linear behavior of a crack might bring

unrealistic result because it becomes insufficient to describe behavior of the crack. In our

research we will consider, “Breathing Phenomena” in a crack. This progression is non-linear

in nature consisting of alternating opening and closing of crack depending on the damping

characteristic of the experimental material.

1.2 Why Crack Detection is Important?

Crack detection is one of the important aspect of structural safety. It is also essential in

designing stage to know the behavior of the material based on natural frequency for with and

without crack condition. Early detection of crack in an object could avoid severe accidents

from occurring hence save a lot of currency. There are two ways to detect cracks, first the

destructive testing and the second is the non-destructive testing. Destructive testing is carried

out until the objects fails to find out the behavior of the material. While non-destructive testing

undergoes investigation without any physical damage of the object. This analysis based on

natural frequency could be a useful tool in reducing failure rate, accidents and costs.

1.3 Research Objectives

In this research, vibration methods are used to study different types of cracks, positon of cracks

and identification of their status are verified. The linear characteristics are only considered to

identify the crack initiation and progression. Moreover, the frequency drop and gain at different

2
beam locations are also considered to identify the most significant position and orientation of

the crack along the plate. In short, our research will focus on:

a. To study intensively on different types of cracks at various locations on the plate using
vibration analysis

b. To verify simulation results with theoretical and experimental results.

c. Intensive study to crack identification due to parametric changes.

The benefits of our research will add advantage to improve the technique currently used and to

make certain applications like cantilever constructions, buildings, bridges etc. safer and more

durable.

1.4 Possible Outcomes

From our research, we are looking forward to following outcomes:

a. From different types of cracks and their parametric change, crack identification will be
possible.

b. Experimental results will establish the simulation result.

c. This work will have a good contribution in identifying cracks intensively.

1.5 Outline of Methodology

We have outlined our research into three phase:
1.5.1 Step 1
SolidWorks will be used to design a thin rectangular metal plate. For the crack design, initially

a small through rectangular slit is considered. ANSYS Workbench will be used to simulate the

design imported from SolidWorks through finite element modelling (FEM). Different

boundary condition of the plate will be applied. The position of the slit will be located at the

fixed end of the plate for primary validation purpose. For modal analysis, the natural frequency

of the healthy rectangular plate without crack will be simulated and natural frequencies of first

twenty modes will be recorded. Similarly, the natural frequencies of the rectangular plate with

3
crack will be noted to compare the variation. Once validation is completed, the slit position

will be changed, depending on the type of operations and simulation conditions and first six

modes of natural frequencies will be recorded.

1.5.2 Step 2
ANSYS 16.0 workbench will be used to perform static structural analysis. For different cases,

maximum von-Mises stress and total deformation will be recoded. The findings will be

compared with the modal frequency curves to establish any relationships for changing crack

positions along beam, crack orientations or crack sizes.

Harmonic analysis will also be performed through simulation. The results will be found in bode

plot. To check for the non-linearity of the design, the range of frequency of first three natural

frequencies are considered and increase the solution interval. If any abrupt amplitude change

is not observed then system is linear, in that case increasing the width of the crack to achieve

non-linearity. Secondary peak for non-linear response located at the corresponding natural

frequency multiplied by the factor 3, 5 and 7 times or 1/3, 1/5 and 1/7 times, known as the sub

and super-harmonic response respectively.

1.5.3 Step 3
The experimental setup will be prepared using aluminum alloy as the material with the exact

dimension is used for simulation. The experimental equipment and apparatus are used for the

experiment will be: Clamping Device (Vice), ECL 202e Driver (Sensor Controller), U12 Non-

Ferrous (Sensor), PCB 2007 mini-shaker (Shaker), Oscilloscope and a Power Supply. The

results obtained from modal and harmonic analysis will be compared with the simulation

results.

4
 Chapter 2
Literature Review

5
Before 1970s, most of the vibrational techniques researchers used were associated with

boundary conditions and derivation of single edge fatigue crack ignoring the opening and

closing phenomena of the crack. Afterwards, researcher put emphasis on this issue to make it

more realistic. Shen and Chu [1] had introduced a fatigue crack in a uniform beam in the form

of breathing crack, which open when the normal strain at the crack tip is positive and vice

versa. The objective was to investigate the contact effect of the crack on the structural dynamic

behavior and its changes with time. A governing equation was derived for the beam by applying

Galerkin procedure to formulate a bilinear stiffness equation of motion for each vibration mode

for a simply supported beam. The result deduced that the changes in dynamic behavior is non-

linear and can be used to identify the crack size and crack location. Chondros et al. [2] used

different boundary conditions on a simply supported beam with a breathing crack. The bilinear

characteristics of the stiffness system were solved separately over their respective domain.

Open cracks shows larger natural frequency than fatigue breathing crack. The experimentation

results with aluminum beam matches with the analytical finding. Tao et al. [3] investigated a

thin cantilever rectangular plate with thorough crack on the fixed end edge to measure the

change of natural frequencies and mode shapes. The plate was made of aluminum. First, the

time-average electronic speckle pattern interferometry (ESPI) was used as experimental

techniques. PZT actuator was fixed with the face surface to measure the natural frequency. This

experiment determined non-linear vibration response of cracked plate. The technique provides

whole field and real time measurement for vibration analysis. Later, the experimental data with

validated with the finite element modeling (FEM) calculation using ANSYS. The study also

showed the principal mode shape of super-harmonic vibration but failed to describe the

detection of crack in the structure. This scientific paper is used for validation purpose.

Orhan [4] showed single and two V-shaped edge cracks were assessed on top and bottom

surface of the cantilever beam. Natural frequencies were calculated using free vibration

6
numerically. The sinusoidal forced vibration was applied on free end to measure the harmonic

response of the beam. The numerical analysis was done with ANSYS for healthy and cracked

beam. It was found that as the crack depth increased natural frequency decreased. Natural

frequency of top crack was more than bottom crack. Moreover, for constant crack depth, as

crack location increased natural frequency increased. Natural frequency of bottom crack was

more than top crack. Modal analysis was found more effective in identifying two cracks, while

harmonic analysis showed better approach in single crack detection. Tufisi et al. [5] conducted

numerical analysis to achieve on six natural frequencies and stress distribution for various crack

positions along the 2% range of the fixed cantilever beam made of structural steel. It was

observed that the stress concentration was disturbed by introduction of crack and its movement

along the positions on the beam. Two linear regression curves were plotted to show first natural

frequency and deflection. These curves had a good match against fixed end crack positions

found from numerical analysis. It was shown that the highest frequency drop for a crack is

located at the fixed end but result obtained could be misleading for detecting crack unless

regression curve is plotted to clarify both deflection and natural frequency. Shinde and Katerkar

[6] investigated on a single transverse crack for various crack depths and crack locations

experimentally on a cantilever beam with mild steel beam. The experiment was performed by

dropping a spherical object 400mm above at 50mm of free end and first three natural frequency

were observed. From graphical representations of crack depth against natural frequency and

crack length against natural frequency were plotted. It was conclude the natural frequencies

decreases with increase in crack depth and the first natural frequency increases as the crack

position moves away from fixed end and vice versa.

Praisach et al. [7] had developed analytical calculus valid for calculating natural frequency of

a damaged cantilever beam (effective for all Euler-Bernoulli Type damaged beam). They

considered crack on the damaged beam as open on the entire width of the beam. Later the crack

7
was placed at different locations of the beam and simulations were conducted. The researchers

concluded the derived analytical calculus matches the FEM modal analysis. The study could

give a clear understanding on material behavior based on static and dynamic response. Another

important finding was that due to change of crack position, there were considerable amount of

relative shift in natural frequency. The thoroughly investigations have not been found properly.

Gillich et al. [8] investigated the effect of natural frequencies due to the change of location and

dimension of the cracks for a cantilever beam. For this purpose, the influence of mass loss and

stiffness change were considered due to presence of transverse crack in a cantilever beam. The

study also concentrates these influences on breathing cracks in open and closing stages along

the cracked beam (i.e. geometrical discontinuity) both analytically and with numerical analysis.

An equation was derived to calculate the frequency of the beam, which showed loss of mass

analytically. This equation was found in terms of mass of the beam and squared of mode shape

of curvature. The length of the crack is same of the width of the beam in this investigation.

This investigation mainly focused on the change of natural frequencies due to the location of

the crack for the beam. Mode shapes changes are not investigated thoroughly due to the change

of crack.

There was only 0.2% variation when the numerical simulation found from ANSYS was

compared with the derived equation. Another set of equations were generated in terms of mass

of the beam and squared of mode shape of curvature. These were used to analytically calculate

the natural frequency of damaged beam with decrease in stiffness (with and without loss of

mass). In numerical simulation, a relation between frequency and damaged width were shown.

A linear and asymptotic relationship were found for larger and smaller frequencies

respectively. Another finding was found in relation with frequency and relative beam length

(x/L). Breathing crack frequency (analytical) is more than frequency of open crack (FEA). The

8
mathematical equations proposed were used to estimate the frequency shifts due to structural

changes in beam, when the crack was open or with breathing crack or crack with mass loss.

Andreaus and Baragatti [11] worked on breathing crack and its detection using aluminum alloy

and steel beam. Instead of beam, this research deals with thin plate of the same materials.

Charalambides and Fang [13] worked with a cantilever beam with horizontal embedded cracks

subjected to transverse force. Lie et al. [14] investigated on a cantilever beam with closed

embedded horizontal crack. In this research, both transverse embedded crack and longitudinal

embedded crack are used along the mid axis of the Z-axis. Soliman [15] worked on impact of

crack inclination angle on a cracked beam. While Jena and Parhi [16] made parametric study

on response analysis of different structures of cracks – one of which is inclined type in their

study. This research includes inclined cracks to analyze the behavior of crack at different

positions. Gillich et al. [17] developed an improved frequency evaluation algorithm based on

damaged location indicator and damage signature for early detection of damage in beam-like

structures. Barad et. al. [18] showed that in crack detection the effect of crack location and

crack depth are vital because they significantly affect the natural frequency. In this research,

the crack length represents as the crack depth by the researcher. Elshamy et al. [19] obtain

similar results. They not only performed it using finite element analysis but also experimentally

validate it. They determined the transverse breathing crack would introduce flexibility in

cantilever beam where the system stiffness and natural frequency would reduce. Gillich et al.

[20] processed mathematical association among deflection, stored energy and natural

frequencies for detection of cracks. It was found that, when the crack is located where the

bending moments are not maximum, the effect on the natural frequency is not weakened once

it is compared with the healthy beam with the normalized square of the mode shape of curvature

of the crack.

9
In this research work, detection of crack in a thin rectangular plate is investigated. The type of

crack analyzed here is thorough on the edge of the plate. The thorough parametric investigation

has been done in this research work to understand the change of natural frequencies due to the

slit crack. Mode shapes have been also investigated to explain the phenomenon properly.

Modal analysis has been done through ANSYS to determine the natural frequencies and mode

shapes. The mode shapes and normalized frequency response curves are considered in

determining the presence of the crack. Two types of modes, namely bending modes and

torsional modes, have been observed. Validation of published paper has been made and a

theoretical investigation is done for the healthy (without crack) plate. The results have been

found in a very good agreement.

10
 Chapter 3
Modeling, Numerical Analysis and Problem Specification

11
3.1 Modeling of Cantilever Plate
In this research, SOLIDWORKS Premium 2019 is used to model of the thin metal plate. The

dimension of the plate is 165mm x 80mm x 1.43mm. This remains fixed throughout the

research. The plate drawn on SOLIDWORKS is shown in Figure 3.1.

(a) Healthy Metal Plate

(b) Cracked Metal Plate

Figure 3.1: SOLIDWORKS Design of Thin Metal Plate

12
On the cantilever thin plate, a thorough crack of different dimensions and orientation is taken

to meet certain research scenario. The crack positions are also changed along Z-axis

(longitudinal direction) and X-axis (transverse direction). Different cases are considered for

investigation shown in the next chapter. Some of the important parameters of the plate and the

crack are shown in the table 3.1 along with nomenclature and values. The basic dimension of

crack is 20mm x 1mmx 1.43mm.

Table 3.1: Dimension and Material Properties of the Plate and the Crack

Particulars Value/mm Material Value

Standard Range Properties
Plate Properties Aluminum Alloy
Length of the Plate 165 165 Density 2700 kg/m3
Width of the Plate 80 80 Young’s Modulus 70 GPa
Thickness of the Plate 1.43 1.43 Poisson’s Ratio 0.33

Crack Properties Structural Steel

Length of the Crack 20 10-70 Density 7850 kg/m3

Width of the Crack 1 0.5-20 Young’s Modulus 200 GPa

Depth of the Crack 1.43 0.143-1.43 Poisson’s Ratio 0.3

Change of Crack Location Along Axis

Location of Crack along Z-axis x
Location of Crack along X-axis y
Crack Orientation
Transverse Crack Crack vertical to Z-axis
Longitudinal Crack Crack vertical to X-axis
Inclined Crack Crack at various angles 0o to 180o
Crack Movement Axis
Longitudinal Direction Z-Axis
Transverse Direction X-Axis
Depth Direction Y-Axis

13
3.2 Numerical Analysis
In this research, ANSYS 16.0 Workbench is used for conducting modal analysis and static
structural analysis. Some important steps are taken into account are shown below.

Table 3.2: Important Steps of Numerical Analysis Procedures

STEP 1 : Create New Project at ANSYS Workbench

STEP 2 : Open: (a) Modal (b) Static Structural
STEP 3 : Edit Engineering Data to add Material Properties & Type
STEP 4 : Import Geometry from SOLIDWORKS
STEP 5 : Open Model of Each Analysis
STEP 6 : Add Mesh & Generate
STEP 7 : Add analysis specification at Analysis Settings
STEP 8 : Define Boundary Conditions
STEP 9 : Add Input Load or Force
STEP 10 : Define Solution Information
STEP 11 : Solve
STEP 12 : Collect Data as per recommendation

Aluminum alloy is used for validation. Later both structural steel and aluminum alloy are used
for particular cases for the numerical analysis.

3.2.1 Modal Analysis

Initially, SolidWorks is used to design a thin rectangular metal plate. For the crack design,

specification mentioned on the previous chapter is used. The design is imported from

SolidWorks to ANSYS 16.0 workbench to simulate through finite element modelling (FEM).

Table 3.2 shows the flow of the processes. Boundary conditions and mesh size are added on

simulation settings. For validation, the position of the crack is located at the fixed end of the

plate. The natural frequency of the healthy rectangular plate is recorded because this is utilized

for comparing the natural frequency of cracked plate. Once validation is complete, the crack

positions are changed, depending on the operations and simulation conditions. In modal

analysis, first six natural frequencies are extracted.

14
3.2.2 Static Structural Analysis
Static structural analysis is conducted using ANSYS 16.0 workbench to illustrate the stress

concentration throughout the thin metal plate. In this case, a fixed end is defined and a small

load of 1N on negative Y-axis applied at the free end. This are shown in the Figure 3.2. The

Maximum von-Mises stress and total deformation are recorded when the position of the crack

along the plate is changed. These findings are compared with the result of modal analysis to

establish any relationships.

Figure 3.2: Schematic Diagram for Indicating Fixed End and Application of Load on
Metal Plate

15
3.2.3 Boundary Condition and Meshing
For setting up a cantilever plate, we defined a fixed end. This is shown in the Figure 3.2. In a

thin healthy metal plate, there are 13483 nodes and 1860 elements. For a cracked plate with

crack length 20mm, number of nodes are from 13666 to 14527 and the number of elements are

from 1874 to 2007. Solid186 element type is used. The advantage of this element type is that

it is functional for three-dimensional nodes for higher order. It displays quadratic displacement

nature for 20 nodes having three degree of freedom per node, i.e., translation in nodal x, y and

z directions. The element is valid for plasticity, hyper-elasticity, creep, stress stiffening, large

deflection, and large strain capabilities. It also has mixed formulation capability for simulating

deformations of nearly incompressible elastoplastic materials, and fully incompressible hyper-

elastic materials.

Figure 3.3: Meshing on Metal Plate

In ANSYS, the relevance center of the meshing i.e., mesh size, is set to fine mode during

simulation in which the minimum edge length is 1mm. Though medium and coarse mesh size

only deviate by about 1% in compared to fine mode, fine mesh size is used throughout the

research for higher precision.

16
3.3 Convention for Measuring Plate and Beam Dimension
In most scientific papers, the dimensions of beam with crack are as shown in the Figure 3.4[12].

In this case, the nomenclature used are put in Table 3.3.

Table 3.3: Nomenclature for Beam and Crack According to Standard Convention

Nomenclature Description
l Length of the beam
w Width of the beam;
Thickness of the Crack
d Depth of the beam
dc Depth of the crack
xc Location of the crack

The standard convention is to use full thickness of the crack along the width of the beam as

shown in Figure 3.4 (b).

(a) ISO View

(b) RHS View

Figure 3.4: Convention for Measuring Beam and Crack Dimension

17
In Figure 3.5, dimensions used for this research are illustrated; where L, B and H – length,

width and thickness of the plate respectively. While a, b and c are length, width and depth of

crack respectively. All the nomenclatures are added in Nomenclature and Terminology section.

The thickness of the beam and thickness of the crack are along the same direction as shown in

3.5 (a). Comparing both figures, it is noticed that the standard crack depth dc (Figure 3.4) is in

the direction of crack length a in this research (Figure 3.5). Furthermore, the standard crack

thickness w (standard beam width) is in the direction of depth of crack c in this research.

(a) ISO View

(b) RHS View

Figure 3.5: Standard Used in this Research for Thin Plate

3.4 Problem Specification

In this study, normalized frequency is used over relative distance of crack along longitudinal

and transverse direction of the plate. A comparison is made keeping the natural frequency of

the healthy plate as a base. Set of simulations are conducted with different crack length, crack

width and crack depth along the edge of the plate. Moreover, the study also include similar

study on metal and metal alloys to note any difference. The orientation of the crack is also

changed from transverse to longitudinal along the edge and mid-axis.

18
 3 5
1

6
2
4

7
8

Figure 3.6: Illustration of All the Cases

Table 3.4: Brief Case Descriptions of all the Cases

Case Case Description Legends

Number on Fig 3.6
Case-1 Change of Transverse Edge Crack Positions along Z-axis 1
(Longitudinal Direction)
Case-2 Change of Transverse Crack Lengths on the Edge along Z-axis 1
(Longitudinal Direction)
Case-3 Change of Transverse Crack Widths on the Edge along Z-axis 1
(Longitudinal Direction)
Case-4 Study of Effect on Normalized Frequency due to Change of Crack 1
Depths along Y-axis (along Depth Direction)
Case-5 Change of Transverse Edge Crack Positions along Z-axis 1
(Longitudinal Direction) with Structural Steel
Case-6 Change of Embedded Transverse Crack Positions along Mid Z- 2
axis
Case-7 Change of Longitudinal Edge Crack (Surface Defect) Positions 3
along Z-axis (Longitudinal Direction)
Case-8 Change of Embedded Longitudinal Crack Positions along Mid Z- 4
axis
Case-9 Change of Longitudinal Edge Crack Positions along X-axis 5
(Transverse Direction)
Case-10 Change of Transverse Edge Crack (Surface Defect) Positions 6
along X-axis (Transverse Direction)
Case-11 Inclined Crack Positions at Relative Distance of 21% on Z-axis 7
Case-12 Inclined Crack Positions at Relative Distance of 79% on Z-axis 8

19
The inclined cracks at different angles are placed at two distinct positions. The first scenario is

for inclined cracks at the free end and the other one is at the fixed end, to check for any outcome.

Finally, a static structure analysis is conducted to visualize the stress concentration and total

deformation when a little load is applied at free end.

20
 Chapter 4
Results of Numerical Simulation and Discussion

21
4.1 Validation of the Scientific Work
In this section, results of the validation with a scientific paper is overviewed. The title of the

paper is “Investigation on Non-Linear Vibration Response of Thin Plates with Crack Using

Electronic Speckle Pattern Interferometry” [3] published in the “The International Conference

on Experimental Mechanics 2018 (ICEM18), Brussels, Belgium”. The author of the paper is

Nan Tao, Yinhang Ma, Hanyang Jiang, Meiling Dai and Fujun Yang. The dimension of the

plate is 165mm x 80mm x 1.43mm. Taking aluminum alloy as the material of the simulation,

the properties of it depicted on Table 4.1. Specimen 1 is the healthy plate, while from

specification 2 to specification 5; the crack length is 20mm, 40mm, 60mm and 70mm

respectively. Figure 4.1 shows all the specimens mentioned below. In all the cases, the crack

is placed at the fixed end of the cantilever plate.

Figure 4.1: Illustration of Specimen from Scientific Paper [3]

(a) Specimen 1# Plate with No Crack (b) Specimen 2# Plate with Crack Size 20mm (c)
Specimen 3# Plate with Crack Size 40mm (d) Specimen 4# Plate with Crack Size 60mm
(e) Specimen 5# Plate with Crack Size 70mm

From Table 4.1 and Table 4.2, we can conclude that the natural frequency found in the scientific

paper and the result obtained in our simulation is very close. The average percentage error is
22
from 0.80 to 0.86, which is less than 1%. This means our simulation work validates with the

scientific work in the paper. Moreover, the mode shapes of our results are similar with the

paper shown in Figure 4.2 and Figure 4.3.

Table 4.1: Validation Result of Specimen-2 & Specimen-3

Natural Frequency
Specimen-1 Specimen-2 Specimen-3
(No Crack) (Crack Size 20mm) (Crack Size 40mm)
Result Obtained Result Obtained Result Obtained
% % %
Mode from from from from from from
Error Error Error
Paper[3] ANSYS Paper[3] ANSYS Paper[3] ANSYS
1 44.176 44.596 0.95 42.698 43.084 0.90 38.359 38.722 0.95

2 191.79 193.36 0.82 182.52 184 0.81 159.33 160.74 0.88

3 274.96 277.56 0.95 266.34 268.7 0.89 249.52 251.73 0.89

4 623.69 628.87 0.83 589.87 594.74 0.83 489.99 494.3 0.88

5 771.77 778.97 0.93 748.34 754.91 0.88 703.39 709.34 0.85

6 1195 1205.2 0.85 1113.4 1122.8 0.84 898.68 905.84 0.80

7 1265.2 1274.3 0.72 1258.1 1267.3 0.73 1251 1260.4 0.75

8 1527 1541.5 0.95 1480.1 1492.7 0.85 1314.5 1324.3 0.75

9 1690.4 1703.3 0.76 1614.9 1627.5 0.78 1529.8 1541.6 0.77

10 1968.4 1985.5 0.87 1844.5 1859.9 0.83 1717.2 1730.2 0.76

11 2352 2370.6 0.79 2233.1 2251.2 0.81 2051.7 2067 0.75

12 2546.2 2569 0.90 2413.5 2432.8 0.80 2294.3 2312.1 0.78

%
Error
0.86 0.83 0.82

23
Table 4.2: Validation Result of Specimen-4 & Specimen-5

Natural Frequency
Specimen-4 Specimen-5
(Crack Size 60mm) (Crack Size 70mm)
Result Result
Obtained from Obtained from
Mode from % Error from % Error
ANSYS ANSYS
Paper[3] Paper[3]
1 31.802 31.397 1.27 25.662 25.95 1.12

2 122.07 123.26 0.97 98.036 99.082 1.07

3 229.2 231.2 0.87 216.11 218.01 0.88

4 385.16 388.28 0.81 350.88 353.62 0.78

5 659.81 665.18 0.81 632.32 637.55 0.83

6 815.589 821.62 0.74 797.7 803.53 0.73

7 1218.6 1228.2 0.79 1142.4 1151.7 0.81

8 1258.9 1268 0.72 1251.4 1260.6 0.74

9 1474 1485 0.75 1424.4 1434.7 0.72

10 1631.6 1644.1 0.77 1579.5 1591.3 0.75

11 1926.1 1940.9 0.77 1822.3 1835.9 0.75

12 2099.3 2114.8 0.74 2069.9 2084.8 0.72

13 2554.2 2573.4 0.75 2521.1 2540.2 0.76

14 2662.1 2680.8 0.70 2571.9 2589.3 0.68

15 3026.7 3049 0.74 3010.2 3032.2 0.73

Average
0.81 0.80
% Error

24
 Number of Modes
1 2 3 4 5 6 7 8 9 10 11 12

(a) FEM from Scientific Paper[3]

Number of Modes
1 2 3 4 5 6 7 8 9 10 11 12

(b) Results obtained from ANSYS 16.0

Figure 4.2: Comparison of Mode Shape for Specimen-1

Number of Modes
1 2 3 4 5 6 7 8 9 10 11 12

(a) FEM from Scientific Paper[3]

Number of Modes
1 2 3 4 5 6 7 8 9 10 11 12

(b) Results obtained from ANSYS 16.0

Figure 4.3: Comparison of Mode Shape for Specimen-5

25
4.2 Theoretical Calculation for Bending Vibration Beam Element Theory
Three are two types of modes shapes, namely bending modes and torsional. In bending mode,

the motion of the plate is observed in the linear direction along the axis of motion. In torsional

vibration, motion of the plate is observed at angular motion along the axis of rotation. First,

third and fifth mode shapes are the bending modes. While, second, fourth and sixth mode

shapes are torsional modes.

In this section, the natural frequencies for this first three bending mode are calculated

theoretically. Here, the Euler-Bernoulli beam model is used where the cross-section is

perpendicular to the bending line. Shear deformation is not taken into consideration. The beam

element has two nodal points and two degrees of freedom in each node: the translation and

rotation along the x-direction. There are four equations of motion. Therefore, the mass and

stiffness matrix are 4x4 matrix which is given below.

Figure 4.4: Uniform Beam Element [9]

26
The mass matrix is
156 22𝑙 54 −13𝑙
𝜌𝐴𝑙 22𝑙 4𝑙 2 13𝑙 3𝑙 2
[𝑚] = [ ]
420 54 13𝑙 156 −22𝑙
−13𝑙 −3𝑙 2 −22𝑙 4𝑙 2

The stiffness matrix is

12 6𝑙 −12 6𝑙
𝐸𝐼 6𝑙 4𝑙 2 −6𝑙 2𝑙 2
[𝑘] = 3 [ ]
𝑙 −12 6𝑙 12 −6𝑙
6𝑙 2𝑙 2 −6𝑙 4𝑙 2

Then Finite element method was applied in this beam to get a more accurate result. So, the total

beam is divided into 20 elements. Then common nodal points between two elements are taken

as one nodal point. Finally, 42x42 mass and stiffness matrix is formed. For the boundary

condition of a cantilever beam, the deflection and slope are zero so the matrix reduced to 40x40

matrix. For free vibration the equation of motion is

{[𝑘] − 𝜔2 [𝑚]}𝑥 = 0

Where ω is the angular velocity and by divided it with 2π, the natural frequencies are found.

The first three bending frequencies are 42.24 Hz, 269.39 Hz and 759.75 Hz, which matched

with the natural frequencies found from ANSYS simulation: 44.6 Hz, 277.56 Hz and 778.59

Hz in good agreement.

27
4.3 Case-1: Change of Transverse Edge Crack Positions along Z-axis
In this study, we use aluminum alloy as the material. Material properties like, density, Young’s

modulus, Poisson’s ratio, along with plate and crack properties are mentioned in the Table 4.1.

In this case, the transverse edge crack positions are moved from fixed to free end i.e. along Z-

axis or in longitudinal direction. The crack positions are changed at 5mm intervals. The

dimension of the crack is 20mm x 1mm x 1.43mm. Schematic diagram of the movement of

crack is illustrated in Figure 4.5.

Figure 4.5: Schematic Diagram for the Change in Transverse Crack Positions along Z-
axis

Two types of modes shapes found from ANSYS simulation, namely bending modes and

torsional modes shown in Figure 4.6. In bending mode, the motion of the plate is observed in

the linear direction along the axis of motion. In torsional vibration, motion of the plate is

observed at angular motion along the axis of rotation. First, third and fifth mode shapes are the

bending modes. While, second, fourth and sixth mode shapes are torsional modes. For first

bending mode, there is one nodal point or (or bends). For second and third bending modes, two

and three nodal points exist respectively. Similarly, for torsional modes there one, two and

three nodal points (or twists) for first, second and third torsional modes respectively.

28
 (a) First Mode (First Bending Mode)

(b) Second Mode (First Torsional Mode)

(c) Third Mode (Second Bending Mode)

29
 (d) Fourth Mode (Second Torsional Mode)

(e) Fifth Mode (Third Bending Mode)

(f) Sixth Mode (Third Torsional Mode)

Figure 4.6: Mode Shape Deformation of a Healthy Metal Plate

30
The Figure 4.7 compares the first six natural frequencies of a healthy plate with plates having

crack at various positions along the edge. It is found that for all the six modes there are

significant frequency drop at relative distance from 0 to 12% (near fixed end) of the plate. For

torsional mode, the most significant frequency drop is 7.4% at relative distance 84.8% on 6th

mode. For bending modes, the most significant frequency drop is 5.88% at relative distance

9.1% on 1st mode. The study also shows that the torsional modes give significantly larger drop

(overall 3.3 to 5%) in frequencies than bending modes (overall 2 to 2.9%).

For bending modes (first, third and fifth modes), it is observed that the maximum drop in

frequency compared to the healthy plate shifted along the relative distance. For first bending

mode, at 9.1% relative distance, the frequency drop is 5.9%. For second bending mode, 4.4%

frequency drop is found at relative distance 54.5%. For third bending mode the frequency drop

is 5.0% at relative distance 69.7%. Similar pattern is observed in torsional modes as well. For

all the six modes at certain relative distance at the free end region, the natural frequency of the

cracked plate demonstrates higher natural frequency than healthy plate. In this research, the

relative distance of crack where these two curves crosses is called “Intersection Point”. For a

20mm crack, it occurs at relative distance of 78.8% in first mode, at 93.9% in second and third

modes and at 99.4% for rest of the modes.

31
 1.01
Normalized Frequency, f/fo 1

0.99

0.98

0.97
Healthy
0.96
Cracked
0.95

0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.01
Normalized Frequency, f/fo

0.99

0.98

0.97 Healthy
Cracked
0.96

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.01
Normalized Frequency, f/fo

0.99

0.98
Healthy
0.97 Cracked

0.96

0.95
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

32
 1.01
Normalized Frequency, f/fo 1

0.99

0.98

0.97
Healthy
0.96
Cracked
0.95

0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.01
Normalized Frequency, f/fo

0.99

0.98

0.97 Healthy
Cracked
0.96

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.01
1
Normalized Frequency, f/fo

0.99
0.98
0.97
0.96 Healthy
0.95 Cracked
0.94
0.93
0.92
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.7: Normalized Frequency over Relative Distance along Z-axis

33
4.4 Case-2: Change of Transverse Crack Lengths on the Edge along Z-axis
This is a similar study done in case-1 for three different transverse crack lengths of 10mm,

20mm (case-1) and 30mm with aluminum alloy. The other properties of the material and plate

dimensions are same as depicted in Table 3.1. The movement of crack is at 5mm intervals

along the edge of the Z-axis. Figure 4.8 illustrates the schematic diagram of the movement of

crack of different lengths along longitudinal direction.

Figure 4.8: Schematic Diagram for the Change of Transverse Crack Length on the Edge
along Z-axis
Figure 4.9 shows the results of this case scenario. Similar trend and pattern are observed in all

the six normalized frequency curves with different variations of frequency drop. It is found that

the larger the crack length the higher is the frequency drop. For bending modes, the maximum

significant frequency for the crack length of 10mm, 20mm and 30mm are 1.8%, 5.9% and

11.4% respectively at relative distance 9.1% on the first mode. For torsional modes, the

maximum significant frequency for the above mentioned crack lengths are 1.4%, 7.4% and

20% for relative distance of 81.8%, 84.8% and 90.9% respectively on the 6th mode. It is also

observed that for small cracks (a≤10mm), fixed end region has more drop in frequencies than

mid-region and free end regions.

34
 1.02
Normalized Frequency, f/fo 1

0.98

0.96
Healthy
0.94
a=10mm
0.92
a=20mm
0.9 a=30mm
0.88

0.86
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.02

1
Normalized Frequency, f/fo

0.98

0.96
Healthy
0.94
a=10mm
0.92
a=20mm
0.9 a=30mm
0.88

0.86
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.02
Normalized Frequency, f/fo

0.98
Healthy
0.96
a=10mm
0.94 a=20mm
a=30mm
0.92

0.9
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

35
 1.02
Normalized Frequency, f/fo 1
0.98
0.96
0.94 Healthy
0.92 a=10mm
0.9 a=20mm
0.88 a=30mm

0.86
0.84
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.02
Normalized Frequency, f/fo

0.98

0.96 Healthy

0.94 a=10mm
a=20mm
0.92
a=30mm
0.9

0.88
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.05
Normalized Frequency, f/fo

0.95
Healthy
0.9
a=10mm
0.85 a=20mm
a=30mm
0.8

0.75
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.9: Normalized Frequency over Relative Distance along Z-axis for Transverse
Crack Length of 10mm, 20mm and 30mm

36
The Table 4.3 below shows the intersection point on the normalized frequency curve of the

first mode. As the crack length increases, the intersection point moves towards the free end. To

determine the exact relative distance of the intersection point we deduced regression equation

of sixth polynomial for the curve of the first mode and found the solution as 70.07%, 74.6%

and 77.40% for 10mm, 20mm and 30mm respectively. These points in this case also seems to

move towards free end with increases crack length.

The regression equations for crack length 10mm, 20mm and 30mm are respectively:

For a=10mm: p =1.6849n6-5.7195n5+7.6358n4-5.0438n3+1.6602n2-0.2057n+0.9902

For a=20mm: p = 3.7773n6 - 13.067n5 + 17.926n4 - 12.312n3 + 4.2663n2 - 0.5517n + 0.9649

For a= 30mm: p = 5.1548n6 - 18.087n5 + 25.393n4 - 18.077n3 + 6.5767n2 - 0.8774n + 0.9238

Here, p = f/fo (normalized frequency), and n=x/L (relative distance)

Table 4.3: Key Findings of Case-2 for Different Crack Lengths

Region on Cantilever Plate % of Max Frequency Drop on Sixth Mode

a=10mm a=20mm a=30mm
Fixed End (x/L: 0-15%) 1.8% 7.0% 16.1%

Mid-Region (x/L: 42-55%) 1.4% 7.0% 16.8%

Free End (x/L: 85-99%) 1.3% 7.4% 20%

Methods Relative Distance of Intersection Pint on First Mode

a=10mm a=20mm a=30mm
Normalized Frequency Curve 72.7% 78.8% 81.8%
(x=120mm) (x=130mm) (x=135mm)
Regression Equation 70.07% 74.60% 77.40%
(x=115.61mm) (x=123.09mm) (x=127.71mm)

37
4.4.1 Further Study on Different Crack Lengths
The study is conducted to find the relationship between normalized frequencies over various

crack length to plate length ratio (a/L). In this case, seven samples are taken, shown in Table

4.4, for three different crack locations: 15mm (near fixed end), 82.5mm (mid-region) and

140mm (near free end), keeping crack width and crack depth constant.

Table 4.4: Seven Samples of Crack Dimension with Various Crack Lengths

Crack Length Crack Width Crack Depth

Sample
a/mm b/mm c/mm a/L
1 70 1 1.43 0.4242

2 60 1 1.43 0.3636

3 40 1 1.43 0.2424

4 20 1 1.43 0.1212

5 15 1 1.43 0.0909

6 10 1 1.43 0.0606

7 5 1 1.43 0.0303

Figure 4.10 illustrates the frequency drop of six modes for changing crack length. It is found

that for all six modes in three distinct regions (fixed end, mid-region and free end) there is

increase in frequency drop with increase in crack length and all these modes are below the

frequency line of the healthy plate. The first two modes show a good pattern where the fixed

region have more drop in frequency than free end. The fifth and sixth mode have more drop at

free end than fixed end. In third and fourth mode, the drop in these regions overlap with each

other and cannot be predicted at mid-region.

38
 1.1
Normalized Frequency, f/fo
1

0.9
No Crack
0.8
x/L=9.1%
0.7 x/L=50%
x/L=84.8%
0.6

0.5
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(a) First Mode

1.1
Normalized Frequency, f/fo

0.9

0.8 No Crack

0.7 x/L=9.1%
x/L=50%
0.6
x/L=84.8%
0.5

0.4
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(b) Second Mode

1.1
Normalized Frequency, f/fo

0.9
No Crack
x/L=9.1%
0.8
x/L=50%
x/L=84.8%
0.7

0.6
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(c) Third Mode

39
 1.1
Normalized Frequency, f/fo
1

0.9
No Crack
0.8
x/L=9.1%
0.7 x/L=50%
x/L=84.8%
0.6

0.5
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(d) Fourth Mode

1.05
Normalized Frequency, f/fo

0.95

0.9 No Crack

0.85 x/L=9.1%
x/L=50%
0.8
x/L=84.8%
0.75

0.7
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(e) Fifth Mode

1.05
1
Normalized Frequency, f/fo

0.95
0.9
0.85 No Crack
0.8 x/L=9.1%
0.75 x/L=50%
0.7 x/L=84.8%

0.65
0.6
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
a/L

(f) Sixth Mode

Figure 4.10: Normalized Frequency Curve over Different Crack Length to Plate Length
Ratio

40
Table 4.5 shows the maximum significant difference in frequency drop for cracks in three

regions. Fixed end region have more frequency drop than free end for a particular crack

dimension.

Table 4.5: Frequency Drop for Cracks at Three Distinct Regions for Different Crack
Length

Region on Cantilever Significant Difference in Frequency Drop (%)

Plate (In Sample 1)
Mode 1 Mode 2 Mode 5 Mode 6

Fixed End 48.1% 56% 17% 23.3%

(x/L=9.1%)
Mid-Region 22.7% 44.5% 20.3% 21.7%
(x/L=50%)
Free End 1.7% 26.1% 29.3% 32.3%
(x/L=84.8%)

41
4.5 Case-3: Change of Transverse Crack Widths on the Edge along Z-axis
This is a similar study done in case-1 for three different transverse crack widths of 0.5mm,

1mm (case-1) and 1.5mm with aluminum alloy. The other properties of the material and plate

dimensions are same as depicted in Table 3.1. The movement of crack is at 5mm intervals

along the edge of the Z-axis. Figure 4.11 illustrates the schematic diagram of the movement of

crack of different width along longitudinal direction.

Figure 4.11: Schematic Diagram for the Change of Transverse Crack Width on the
Edge along Z-axis

The results displayed in the Figure 4.12 shows that as the crack widths changes there is very

little variations of frequency drop in all the six normalized frequency curves. These curves have

similar pattern as in case-1. For bending modes, the maximum significant frequency for the

crack width of 0.5mm, 1.0mm and 1.5mm are 5.7%, 5.9% and 6.0% respectively at relative

distance 9.1% on the 1st mode. For torsional modes, the maximum significant frequency for

the crack width of 0.5mm, 1.0mm and 1.5mm are 7.1%, 7.4% and 7.6% respectively at relative

distance 84.8% on the 6th mode.

42
 1.01
Normalized Frequency, f/fo 1

0.99

0.98
No Crack
0.97
b=0.5mm
0.96
b=1.0mm
0.95 b=1.5mm
0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.01

1
Normalized Frequency, f/fo

0.99

0.98
No Crack
0.97
b=0.5mm
0.96
b=1.0mm
0.95 b=1.5mm
0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.01
Normalized Frequency, f/fo

0.99
No Crack
0.98
b=0.5mm
0.97 b=1.0mm
b=1.5mm
0.96

0.95
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

43
 1.01
Normalized Frequency, f/fo 1

0.99

0.98
No Crack
0.97
b=0.5mm
0.96
b=1.0mm
0.95 b=1.5mm
0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.01
Normalized Frequency, f/fo

0.99

0.98 No Crack

0.97 b=0.5mm
b=1.0mm
0.96
b=1.5mm
0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.02
1.01
Normalized Frequency, f/fo

1
0.99
0.98
0.97 No Crack
0.96 b=0.5mm
0.95 b=1.0mm
0.94
b=1.5mm
0.93
0.92
0.91
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.12: Normalized Frequency over Relative Distance along Z-axis for Transverse
Crack Width of 0.5mm, 1mm and 1.5mm

44
The frequency drop remains almost equal for varying the crack width over the Z-axis. It is

observed that all the regions in the plate, free end, mid-region and fixed end, have almost

similar frequency drop ranging from 6.8% to 7.6% given on Table 4.6. The maximum

significance frequency drop compared to the frequency of healthy plate for 0.5mm, 1mm and

1.5mm are 7.1%, 7.4% and 7.6% at 6th mode at relative distance of 84.8% for all the cases.

Table 4.6: Key Findings of Case-3 for Different Crack Widths

Region on Cantilever Plate % of Max Frequency Drop on Sixth Mode

b=0.5mm b=1mm b=1.5mm
Fixed End (x/L: 0-15%) 6.8% 7.0% 7.4%
Mid-Region (x/L: 42-55%) 6.8% 7.0% 7.2%
Free End (x/L: 85-99%) 7.1% 7.4% 7.6%

Methods Relative Distance of Intersection Point on First Mode

b=0.5mm b=1mm b=1.5mm
Normalized Frequency Curve 81.8% 78.8% 75.8%
(x=135mm) (x=130mm) (x=125mm)
Regression Equation 75.84% 74.60% 72.37%
(x=125.14mm) (x=123.09mm) (x=119.40mm)

The regression equations for crack width 0.5mm, 1mm and 1.5mm are respectively:

For b=0.5mm: p = 3.748n6 - 12.968n5 + 17.792n4 - 12.22n3 + 4.2354n2 - 0.5512n + 0.9669

For a=1mm: p = 3.7773n6 - 13.067n5 + 17.926n4 - 12.312n3 + 4.2663n2 - 0.5517n + 0.9649

For a=1.5mm: p = 3.7842n6 - 13.099n5 + 17.981n4 - 12.353n3 + 4.2788n2 - 0.5494n + 0.9632

Here, p = f/fo (normalized frequency), and n=x/L (relative distance)

The Table 4.6 also shows the intersection point on the normalized frequency curve of the first

mode. As the crack width increases, the intersection point moves towards the fixed end, unlike

the case of increasing the crack length. To determine the exact relative distance of the

intersection point we deduced regression equation of sixth polynomial for the curves of the

45
first mode and found the solution as 75.84%, 74.60% and 72.37% for 0.5mm, 1mm and 1.5mm

respectively. These points in this case also seems to move towards fixed end with increases

crack width.

4.5.1 Further Study on Different Crack Widths

The study is conducted to find the relationship between normalized frequencies over various

crack width to plate length ratio (b/L). Table 4.7 shows five samples that are taken for three

different crack locations: 15mm (near fixed end), 82.5mm (mid-region) and 140mm (near free

end). In this case, crack length and crack depth are kept constant.

Table 4.7: Five Samples of Crack Dimension with Various Crack Widths

Crack Length Crack Width Crack Depth

Sample
a/mm b/mm c/mm b/L
1 20 1 1.43 0.0061
2 20 5 1.43 0.0303
3 20 10 1.43 0.0606
4 20 15 1.43 0.0909
5 20 20 1.43 0.1212

Figure 4.13 illustrates the frequency drop of six modes for changing crack width in three

distinct regions of the cantilever plate. The frequency drop at mid-region, the normalized

frequency curve of fifth mode and sixth mode cannot be explained for the random overlapping

of the curves. In the first four mode, it is observed that as the crack width increases the

frequency drop increases at fixed region. On the other hand, there is frequency gain at free end.

All the six modes consistently drop below the frequency of the healthy plate over the

subsequent increase in modes. The maximum significant difference in frequency drop is found

at fixed region for mode 1 Sample 5 (b=20mm) with the value of 9.7%.

46
 1.06
Normalized Frequency, f/fo 1.04
1.02
1
0.98 No Crack
0.96 x/L=9.1%
0.94 x/L=50%
0.92 x/L=84.8%

0.9
0.88
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(a) First Mode

1.08
1.06
Normalized Frequency, f/fo

1.04
1.02
1 No Crack
0.98 x/L=9.1%
0.96 x/L=50%
0.94 x/L=84.8%

0.92
0.9
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(b) Second Mode

1.03

1.02
Normalized Frequency, f/fo

1.01

1
No Crack
0.99
x/L=9.1%
0.98
x/L=50%
0.97 x/L=84.8%
0.96

0.95
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(c) Third Mode

47
 1.04
Normalized Frequency, f/fo
1.02

1
No Crack
0.98
x/L=9.1%
0.96 x/L=50%
x/L=84.8%
0.94

0.92
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(d) Fourth Mode

1.01
Normalized Frequency, f/fo

0.99
No Crack
0.98
x/L=9.1%
0.97 x/L=50%
x/L=84.8%
0.96

0.95
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(e) Fifth Mode

1.02
Normalized Frequency, f/fo

0.98
No Crack
0.96
x/L=9.1%
0.94 x/L=50%
x/L=84.8%
0.92

0.9
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
b/L

(f) Sixth Mode

Figure 4.13: Normalized Frequency Curve over Different Crack Width to Plate Length
Ratio

48
4.6 Case-4: Study on Effect of Normalized Frequency due to Change of Crack Depths
along Y-axis
The study is conducted, to find the relationship between normalized frequencies over various

crack depths, to plate thickness (c/H) along Y-axis (depth direction). Table 4.8 shows five

samples that are taken for three different crack locations: 15mm (near fixed end), 82.5mm

(mid-region) and 140mm (near free end). In this case, crack length and crack width are kept

constant.

Table 4.8: Five Samples of Crack Dimension with Various Crack Depths
Crack Length Crack Width Crack Depth
Sample
a/mm b/mm c/mm c/H
1 20 1 1.43 1.0000

2 20 1 1.0725 0.7500

3 20 1 0.715 0.5000

4 20 1 0.3575 0.2500

5 20 1 0.143 0.1000

Figure 4.14 illustrates the frequency drop of six modes for changing crack depth in three

distinct regions of the cantilever plate. For all the six modes, there is no frequency drop until

the crack depth is 75% of the plate thickness (i.e., c/H=75%). It is also found that curve

generated up to this point is parallel to the line of the healthy plate. After this point there is a

rapid frequency drop until the crack is thorough (i.e., c/H=100%).

For bending modes, the maximum frequency drop for thorough crack (sample 1) and non-

thorough crack (sample 2) is 5.9% and 0.87% respectively near fixed end on the 1st mode. For

torsional modes, the maximum frequency drop for thorough crack (sample 1) and non-thorough

crack (sample 5) is 7.4% and 0.92% respectively near free end on the 6th mode.

49
 1.01
Normalized Frequency, f/fo 1

0.99

0.98
No Crack
0.97
x/L=9.1%
0.96
x/L=50%
0.95 x/L=84.8%
0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(a) First Mode

1.01
Normalized Frequency, f/fo

0.99

0.98 No Crack

0.97 x/L=9.1%
x/L=50%
0.96
x/L=84.8%
0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(b) Second Mode

1.01
Normalized Frequency, f/fo

0.99
No Crack
0.98
x/L=9.1%
0.97 x/L=50%
x/L=84.8%
0.96

0.95
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(c) Third Mode

50
 1.02
Normalized Frequency, f/fo 1.01

0.99
No Crack
0.98
x/L=9.1%
0.97
x/L=50%
0.96 x/L=84.8%
0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(d) Fourth Mode

1.01

1.005
Normalized Frequency, f/fo

0.995
No Crack
0.99
x/L=9.1%
0.985
x/L=50%
0.98 x/L=84.8%
0.975

0.97
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(e) Fifth Mode

1.02
1.01
Normalized Frequency, f/fo

1
0.99
0.98
No Crack
0.97
x/L=9.1%
0.96
x/L=50%
0.95
x/L=84.8%
0.94
0.93
0.92
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
c/H

(f) Sixth Mode

Figure 4.14: Normalized Frequency Curve over Different Crack Depth to Plate
Thickness along Y-axis

51
4.7 Case-5: Transverse Edge Crack Positions along Z-axis with Structural Steel
This is a similar study like case-1 where transverse edge cracks are changed along Z-axis. The

only difference here is that the material is changed to structural steel. Table 3.1 shows the

properties of the cantilever plate and the crack. The orientation of crack position is transverse

in nature. Figure 4.15 shows the normalized frequency over the relative distance for six modes.

The result found in case-1 with aluminum alloy is plotted in the same figure with case-5 with

structural steel to compare any differences. Schematic diagram of the movement of crack is

illustrated in Figure 4.5.

It is found that for all the six modes there are significant frequency drop at relative distance

from 0 to 12% (near fixed end) of the plate. For torsional mode, the most significant frequency

drop is 7.3% at relative distance 84.8% on 6th mode. For bending modes, the most significant

frequency drop is 5.8% at relative distance 9.1% on 1st mode. For a 20mm crack, the

intersection point occurs at relative distance of 78.8% in first mode, 93.9% in second and third

modes and at 99.4% for rest of the modes.

All natural frequencies of aluminum alloy are found greater than that of structural steel along

Z-axis showing that the former has a higher stiffness. Moreover, normalized frequency curves

of both these cases (case-1 and case-5) follow the same pattern and trend as shown in the Figure

4.15. The normalized curves in torsional modes share the same line of progression with little

difference. On the other hand, though the normalized curve of the bending modes have the

same pattern, their progression takes place at certain gap. As both these materials share the

common trends, structural steel is used in later cases in the research. AA stands for Aluminum

Alloy and SS stand for Structural Steel in the figures and tables below.

52
 1.03
Normalized Frequency, f/fo 1.02
1.01
1
0.99
Healthy AA
0.98
Healthy SS
0.97
Cracked AA
0.96
Cracked SS
0.95
0.94
0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.01
Normalized Frequency, f/fo

0.99

0.98 Healthy AA

0.97 Healthy SS
Cracked AA
0.96
Cracked SS
0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.02
Normalized Frequency, f/fo

1.01

0.99 Healthy AA

0.98 Healthy SS
Cracked AA
0.97
Cracked SS
0.96

0.95
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

53
 1.02
Normalized Frequency, f/fo 1.01
1
0.99
0.98 Healthy AA
0.97 Healthy SS
0.96 Cracked AA
0.95 Cracked SS

0.94
0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.02

1.01
Normalized Frequency, f/fo

0.99
Healthy AA
0.98
Healthy SS
0.97
Cracked AA
0.96 Cracked SS
0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.02
1.01
Normalized Frequency, f/fo

1
0.99
0.98
Healthy AA
0.97
Healthy SS
0.96
Cracked AA
0.95
Cracked SS
0.94
0.93
0.92
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.15: Normalized Frequency Between Aluminum Alloy (Case-1) and Structural
Steel (Case-5)

54
4.8 Case-6: Change of Embedded Transverse Crack Positions along Mid Z-axis
This is a similar study mentioned in case-5 where the transverse cracks are moved along the

edge of the Z-axis using structural steel. In this case, material used is structural steel. However,

the crack positions are changed along the mid-line of the Z-axis. This type of crack, which is

not at the edge of the plate, are known as embedded crack. This is shown in the Figure 4.16.

Table 3.1 shows the properties of the cantilever metal plate and the crack.

Figure 4.16: Schematic Diagram for the Change in Embedded Transverse Crack
Positions along Mid Z-axis of the Plate
Figure 4.17 compares the first six normalized frequencies of a healthy plate with the plates

having crack. In bending modes, the normalized frequency curves of embedded crack and edge

crack follow the same pattern. However, the frequency drop of edge crack is massive than that

of mid-crack. The bending modes in mid-cracks have frequency drop near the fixed end. With

the subsequent increase in bending mode, the maximum frequency drop moves towards the

free end. For bending modes, the maximum frequency drop is 4.2% at relative distance 6.1%

on 1st mode. In torsional modes of embedded crack has no significant frequency drop. This is

because the angle of twist in this axis is zero. The maximum frequency drop in torsional mode

is 0.26% at relative distance 90.9% on 6th mode. This phenomenon could be used to predict

the presence of embedded crack with transverse orientation. In Figure 4.17, the embedded

crack is mentioned as “Mid Crack”.

55
 1.01
Normalized Frequency, f/fo 1

0.99

0.98

0.97 Healthy

0.96 Edge Crack

Mid Crack
0.95

0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.01
Normalized Frequency, f/fo

0.99

0.98
Healthy
0.97
Edge Crack
0.96 Mid Crack

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.01
Normalized Frequency, f/fo

0.99

0.98 Healthy
Edge Crack
0.97
Mid Crack
0.96

0.95
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

56
 1.01
Normalized Frequency, f/fo 1

0.99

0.98

0.97 Healthy

0.96 Edge Crack

Mid Crack
0.95

0.94

0.93
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.01
Normalized Frequency, f/fo

0.99

0.98
Healthy
0.97
Edge Crack
0.96 Mid Crack

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.01
1
Normalized Frequency, f/fo

0.99
0.98
0.97
Healthy
0.96
Edge Crack
0.95
Mid Crack
0.94
0.93
0.92
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.17: Difference in Normalized Frequency Between Transverse Edge Crack
(Case-5) and Transverse Mid-Crack (Case-6) along Z-axis

57
4.9 Case-7 and Case-8: Change of Longitudinal Crack at Edge and Mid-Positions along
Z-axis
Case-7 describes the change of longitudinal edge crack positions along longitudinal direction

(Z-axis). The longitudinal edge crack is generally stated as surface defect. On the other hand,

case-8 describes the change of embedded longitudinal crack positions along mid Z-axis. Figure

4.18 illustrates schematic diagram these two cases. In these cases, the crack positions are

changed at 10mm intervals including fixed and free end of the plate. The details of plate and

crack properties are in the Table 3.1.

(a) Longitudinal Edge Crack (Case-7) (b) Embedded Longitudinal Crack (Case-8)
Figure 4.18: Schematic Diagram for the Change in Positions along the Z-axis of the
Plate for Longitudinal Edge Crack (Case-7) and Embedded Mid-Crack (Case-8)

Figure 4.19 shows the comparison between a longitudinal edge crack (Case-7) and an

embedded longitudinal crack (Case-8) of the first six modes. In the same figure, the embedded

crack is mentioned as “Mid Crack”.

For longitudinal edge crack, the frequency drop is observed at fixed end, while frequency gain

is observed at free end. However, it is found that for both bending and torsional modes there

are very little frequency drop all over the plate. The most significant frequency drop for bending

mode is 0.29% at relative distance 6.1% on 1st mode. While for torsional mode, it is 0.25% at

relative distance 0% on 2nd mode. This frequency drop is negligible and this type of surface

defect remain undetectable in this research.

58
 1.004
Normalized Frequency, f/fo 1.003
1.002
1.001
1
0.999
Healthy
0.998
Edge Crack
0.997
0.996 Mid Crack
0.995
0.994
0.993
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(a) First Mode

1.006

1.004
Normalized Frequency, f/fo

1.002

0.998 Healthy

0.996 Edge Crack

Mid Crack
0.994

0.992

0.99
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(b) Second Mode

1.002
Normalized Frequency, f/fo

1.001

0.999 Healthy
Edge Crack
0.998
Mid Crack
0.997

0.996
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(c) Third Mode

59
 1.01
Normalized Frequency, f/fo 1

0.99

0.98
Healthy
0.97
Edge Crack
0.96 Mid Crack

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(d) Fourth Mode

1.001
Normalized Frequency, f/fo

1.0005

0.9995
Healthy
0.999
Edge Crack
0.9985 Mid Crack

0.998

0.9975
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(e) Fifth Mode

1.01
1
Normalized Frequency, f/fo

0.99
0.98
0.97
0.96 Healthy
0.95 Edge Crack
0.94 Mid Crack
0.93
0.92
0.91
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, x/L

(f) Sixth Mode

Figure 4.19: Difference in Normalized Frequency Between Longitudinal Edge Crack
(Case-7) and Longitudinal Mid-Crack (Case-8) along Z-axis

60
In bending modes, the maximum frequency drop for the embedded longitudinal crack is only

0.58% at relative distance 0% on the 1st mode. In torsional modes, the most significant

frequency drop is 7.9% at the relative distance 87.9% (free end) on the 6th mode. These drops

are very abrupt and sharp. This is the only crack position where the presence of embedded

longitudinal mid-crack could be confirmed. It is also noticed that the magnitude of this

frequency drop at free end increases with mode. Other than the free end, the maximum

frequency drop is 1.7% at the relative distance of 84.8% on the 6th mode.

4.10 Case-9 and Case-10: Change of Longitudinal and Transverse Edge Crack Positions
along X-axis on Free End
Case-9 describes the change of longitudinal edge crack positions along transverse direction (X-

axis) on the free end of the plate. On the other hand, case-10 describes the change of transverse

edge crack positions along transverse direction (X-axis) on the free end of the plate. The

transverse edge crack on this axis act as a surface defect. Figure 4.20 illustrates schematic

diagram these two cases. The longitudinal crack positions (case-9) and the transverse crack

positions (case-10) are changed at 5 mm and 10mm intervals respectively. The details of plate

and crack properties are in the Table 3.1.

(a) Longitudinal Edge Crack (Case-9) (b) Transverse Edge Crack (Case-10)
Figure 4.20: Schematic Diagram for the Change in Longitudinal (Case-9) and
Transverse (Case-10) Crack Positions along the X-axis on Free End
Figure 4.21 shows the comparisons between the change in crack positions of longitudinal edge

crack (case-9) and a transverse edge crack (case-10) along X-axis for the first six modes. Here

61
the relative distance along X-axis is considered as the ratio of the width of the plate and the

width of the plate (y/B).

For the longitudinal edge crack, it is found that, the bending modes have very little frequency

change. The maximum frequency drop for this mode is 0.3% throughout the 1st mode. While

for the torsional modes there are significant frequency drop from the initial edge position to

the mid-position. After this point, there is significant frequency gain until the crack position

reaches the final edge. The maximum frequency drop for torsional modes is 8.9% at relative

distance 37.5% and 62.5% on the 6th mode.

Except the first and sixth mode, all the normalized frequency curves are parabolic when

roughly 10% of both edges are not considered. The presence of a longitudinal crack at the free

end can be estimated considering the fact that the bending modes having natural frequencies

higher (or almost align themselves) than the healthy plate.

For the transverse edge crack (surface defect), the frequency drop for both bending and

torsional modes is very low. For bending modes, the maximum frequency drop is 0.3%

throughout the 1st mode. For torsional modes, the maximum frequency drop is 0.14% at

relative distance 50% on the 6th mode.

For the bending modes, normalized frequency curves are higher than the frequencies of the

healthy plate. In the torsional modes for transverse edge crack, the normalized curves are

parabolic with symmetry along the middle of the width of plate. The troughs of these parabolas

lies slightly below the frequencies of the healthy plate on the same points.

The existence of a transverse edge crack on the free end can be estimated by the presence of

three parabolic curves for torsional modes and the three normalized curves higher than that of

frequencies of healthy plate for bending modes.

62
 1.0035
Normalized Frequency, f/fo 1.003

1.0025

1.002

1.0015 Healthy

1.001 Transverse
Longitudinal
1.0005

0.9995
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(a) First Mode

1.006

1.004
Normalized Frequency, f/fo

1.002

0.998 Healthy

0.996 Transverse
Longitudinal
0.994

0.992

0.99
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(b) Second Mode

1.0035

1.003
Normalized Frequency, f/fo

1.0025

1.002

1.0015 Healthy

1.001 Transverse
Longitudinal
1.0005

0.9995
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(c) Third Mode

63
 1.01
Normalized Frequency, f/fo 1

0.99

0.98
Healthy
0.97
Transverse
0.96 Longitudinal

0.95

0.94
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(d) Fourth Mode

1.004
1.0035
Normalized Frequency, f/fo

1.003
1.0025
1.002
1.0015 Healthy
1.001 Transverse
1.0005 Longitudinal
1
0.9995
0.999
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(e) Fifth Mode

1.02
Normalized Frequency, f/fo

0.98

0.96 Healthy
Transverse
0.94
Longitudinal
0.92

0.9
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Relative Distance, y/B

(f) Sixth Mode

Figure 4.21: Difference in Normalized Frequency Between Longitudinal (Case-9) and
Transverse (Case-10) Crack Positions along the X-axis on Free End

64
4.11 Case-11 and Case-12: Inclined Crack Positions at Relative Distance of 21% and
79% along Z-axis
Case-11 describes the inclined crack positions near the fixed end, at relative distance of 21%

of the plate along Z-axis. While case-12 describes the inclined crack positions near free end, at

relative distance of 79% of the plate along Z-axis. Figure 4.22 illustrates schematic diagram of

these two cases. In both these cases, the angular interval is 15o for the span of 180o. The details

of plate and crack properties are in the Table 3.1

(a) Inclined Crack at x/L=21% (Case-11) (b) Inclined Crack at x/L=79% (Case-12)

Figure 4.22: Schematic Diagram for the Change in Inclined Crack at x/L=21% (Case-
11) and x/L=79% (Case-12) along the Z-axis

The study of inclined crack location is taken at relative distance 21% (at 34.5mm mark) from

fixed end because this region is found to have maximum frequency drop in first mode of

transverse edge crack (found in case-1 to case-4). On the other hand, the study of inclined crack

location is taken at relative distance 79% (at 130mm mark) from fixed end because in this

region the intersection point in first mode exceeds the natural frequency of a healthy plate

(found in case-1 to case-4).

From Figure 4.23, for the inclined crack at relative distance 21%, it is found that, the shape of

the normalized curves are parabolic or slightly distorted parabolic with the trough in the region

60o to 120o. The maximum frequency drop for bending modes is 4.6% at an angle of 75o on the

65
1st mode. While for torsional modes, the maximum frequency drop is 5.4% at an angle of 90o

on the 2nd mode.

From Figure 4.23, for the inclined crack at relative distance 79%, it is found that, the shape of

the normalized curves are parabolic or slightly distorted parabolic (except the first mode) with

the trough in the region 60o to 90o. The maximum frequency drop for bending modes is 4.5%

at an angle of 75o on the 5th mode. While for torsional modes, the maximum frequency drop

is 6.4% at an angle of 90o on the 6th mode.

In both these cases, for all the modes, the lowest frequency drop is at the starting and ending

point of the curve where the crack angles are 0o and 180o. These are the location where the

crack align longitudinally at the edge. The frequency drop at first and second mode are

dominant at relative distance 21%, while it is dominant at fifth and sixth modes at relative

distance 79%.

66
 1.01
Normalized Frequency, f/fo
1

0.99

0.98 Healthy
x/L=79%
0.97
x/L=21%
0.96

0.95
10
20
30
40
50
60
70
80
90
0

100
110
120
130
140
150
160
170
180
Angle

(a) First Mode

1.01
Normalized Frequency, f/fo

0.99

0.98
Healthy
0.97
x/L=79%
0.96 x/L=21%

0.95

0.94
10
20
30
40
50
60
70
80
90
0

100
110
120
130
140
150
160
170
180

Angle

(b) Second Mode

1.002
1
Normalized Frequency, f/fo

0.998
0.996
0.994
0.992
Healthy
0.99
x/L=79%
0.988
0.986 x/L=21%
0.984
0.982
0.98
80
10
20
30
40
50
60
70

90
0

100
110
120
130
140
150
160
170
180

Angle

(c) Third Mode

67
 1.01
Normalized Frequency, f/fo 1

0.99

0.98

0.97 Healthy

0.96 x/L=79%
x/L=21%
0.95

0.94

0.93
10
20
30
40
50
60
70
80
90
0

100
110
120
130
140
150
160
170
180
Angle

(d) Fourth Mode

1.01
Normalized Frequency, f/fo

0.99

0.98 Healthy
x/L=79%
0.97
x/L=21%
0.96

0.95
10
20
30
40
50
60
70
80
90
0

100
110
120
130
140
150
160
170
180

Angle

(e) Fifth Mode

1.01

1
Normalized Frequency, f/fo

0.99

0.98

0.97 Healthy

0.96 x/L=79%
x/L=21%
0.95

0.94

0.93
10
20
30
40
50
60
70
80
90
0

100
110
120
130
140
150
160
170
180

Angle

(f) Sixth Mode

Figure 4.23: Difference in Normalized Frequency Between Change in Inclined Crack at
x/L=21% (Case-11) and x/L=79% (Case-12) along the Z-axis

68
4.12 Equivalent von-Mises Stress and Total Deformation
This is a further study of case-1 based on Static Structural Analysis using ANSYS 16.0

Workbench. In this case, the applied load is 1N at downward direction (negative Y-axis) at the

mid-point of the free end of the cantilever plate, as shown in Figure 3.2. From this study,

maximum and minimum of equivalent von-Mises stress and maximum total deformation is

performed for the crack length of 20mm.

In Figure 4.24, the maximum stress rises to peak at relative distance of 9.1% and gradually

drops linearly to a lowest point at 78.8% of relative distance. The curve then becomes parallel

and align itself over the maximum stress line of the healthy plate. The maximum total

deformation curve for static structural analysis follows the same pattern as maximum von-

Mises stress. The curve becomes parallel to the maximum deformation of a healthy plate for a

of relative distance of 90.9% to 97%. After that, the curve takes a sharp rise in deformation at

the free end. Here the total deformation and maximum von-Mises stress have similar pattern.

1.18 25

Max Equivalent von-Mises Stress, Pa (10^6)

1.16

20
1.14
Deformation, m (10^3)

1.12
15
1.1

1.08
10
1.06

1.04
5

1.02

1 0
0.00

0.06

0.12

0.18

0.24

0.30

0.36

0.42

0.48

0.52

0.58

0.64

0.70

0.76

0.82

0.88

0.94

0.99

Relative Distance, x/L

Healthy Deformation Cracked Deformation

Figure 4.24: Maximum Equivalent von-Mises Stress and Deformation

69
 Both total deformation and the modal deformation is observed along the Y-axis at the free

end. The modal deformation of six modes found in Figure 4.25, also represents the rise in

deflection and deformation at free end. The torsional modes typically illustrates the

deflection behavior at the free end region of the cantilever plate. The deformation for bending

modes less than torsional modes.

11
25

10.5 23

Deformation, m (Mode 2, 4 & 6)

Deformation, m (Mode 1, 3 & 5)

21
10

19
9.5
17
9
15

8.5
13

8 11
0.00

0.06

0.12

0.18

0.24

0.30

0.36

0.42

0.48

0.52

0.58

0.64

0.70

0.76

0.82

0.88

0.94

0.99
Relative Distance, x/L

Mode 1 Mode 3 Mode 5 Mode 2 Mode 4 Mode 6

Figure 4.25: Maximum Deformation for Modal Analysis

Another interesting relationship is found between natural frequency of mode 1 and maximum

von-Mises stress illustrated in Figure 4.26. Both the curves reaches their respective lowest and

highest point at the same relative distance of 9.1%. These curves either cross or become parallel

to their respective line of healthy plate at relative distance of 78.8%. This means that the stress

curve and normalized frequency curve of mode 1 is reflection of one another.

70
 1.01 25

Max Equivalent von-Mises Stress, Pa (10^6)

1

0.99 20
Normalized Frequency, f/fo

0.98

0.97 15

0.96

0.95 10

0.94

0.93 5

0.92

0.91 0
0.00

0.06

0.12

0.18

0.24

0.30

0.36

0.42

0.48

0.52

0.58

0.64

0.70

0.76

0.82

0.88

0.94

0.99
Relative Distance, x/L

Healthy Mode Mode 1 Healthy Max Stress Cracked Max Stress

Figure 4.26: Normalized Frequency of Mode 1 and Maximum Equivalent von-Mises

Stress

4.13 Mode Shapes, Stress Concentration and Total Deformation at Three Regions
The static stress analysis has been performed to observe the stress concentration due to the

change of the position of the crack. As it is not harmonically excited, only the first mode

condition can be observed.

In this section, brief discussion of the deformation, stress concentration and first six mode

shapes are carried out on cracked and healthy plate. For cracked plate, three distinct points are

selected. From the fixed end region, crack on 15mm mark on the plate is considered. This has

a relative distance of 9.1%, which shows the maximum frequency drop in mode 1. The mid-

point on the plate i.e., 82.5mm mark is taken which has the relative distance of 50%. From the

free end region on the plate, 140mm mark is considered. This has a relative distance of 84.8%,

where intersection point from mode 1 exceeds or intersect the line of the healthy plate.

71
4.13.1 Stress Concentration and Total Deformation on Healthy Plate
Figure 4.27 shows the total deformation and von-Mises Stress of a healthy plate when Static

Structural Analysis is performed. It is understood that the deformation is maximum at free end

when the stress concentration is maximum at fixed end of the healthy plate. The magnitude of

stress on the healthy plate is 6.79 MPa. The minimum stress is found at the free end of the plate

covering almost 15% of the area.

(a) Deformation when 1N force applied downwards at Free end

(b) von-Mises Stress

Figure 4.27: Deformation and von-Mises Stress of a Healthy Plate from Static
Structural Analysis
4.13.2 Mode Shape of First Natural Frequency
Figure 4.28 illustrates the mode shapes of first natural frequency of for crack at relative

distance 9.1% (near fixed end), 50% (mid-region) and 84.8% (near free end)

72
 (a) Mode 1 at relative distance of 9.1% near fixed end

(b) Mode 1 at relative distance of 50% at Mid-region

(c) Mode 1 at relative distance of 84.8% near fixed end

Figure 4.28: Mode Shape of First Natural Frequency at Three Regions

73
Although, the sketch of the mode shape of first natural frequency of the three regions look the

same, the drop in frequency is different. The maximum drop in frequency is near fixed end,

where the bend is maximum. Any crack at this area would cause the stiffness to decrease

resulting in the massive drop in frequency. Moreover, due to the presence of crack in this region

breathing phenomena is observed, i.e., opening and closing of the crack.

Along the mid-region, from Figure 4.26, it is understood that there would be rise in frequency.

This happens because the effect on stiffness increases on the crack when any crack is found in

this area. Another explanation is that due to decrease in breathing effect of the crack, the

closeness of the crack decreases, as a result the stiffness of the plate increases in this region

causing the natural frequency to rise.

It is discussed earlier that there exist an intersection point at free end where the normalized

frequency curve crosses the frequency line of the healthy plate. At the free end, the natural

frequency remain constant at maximum value and above the frequency line of the healthy plate.

This happens because the effect is mass is more than that of the stiffness. This happens due to

the increased deflection when the plate is loaded statically with its own mass at this area. As a

result, the stored energy increase and hence increase in natural frequency. Moreover, the effect

of breathing phenomena is decreased in this region than in the fixed end and mid-region.

4.13.3 Stress Concentration and Total Deformation Near Fixed End Region
From Figure 4.29, it is found that when crack is located near the fixed end region, the maximum

von-Mises stress is located at the inner tip of the crack. At relative distance 9.1%, the plate

reaches its maximum stress concentration of 22.27 MPa that is the highest stress value recorded

for any crack position along the plate. The corresponding frequency drop is highest as found

in Figure 4.26. In theory, the mass located near the fixed end carries the highest amount of

strain energy. In this region the bending moment and the mode shape of curvature also reaches

74
the maximum value. When any piece of mass removed near the fixed end, balance of these

properties is hampered and there is a massive drop in frequency in this region.

From Figure 4.24 the total deformation is the highest value for any crack positions along the

plate. The minimum stress is found at the free end of the plate covering almost 50% of the area

of it.

(a) Deformation when 1N force applied downwards at Free end

(b) von-Mises Stress

Figure 4.29: Deformation and von-Mises Stress of a Crack at Relative Distance of 9.1%
Near Fixed End from Static Structural Analysis

75
4.13.4 Stress Concentration and Total Deformation at Mid-Region
From Figure 4.30, it is found that when crack is located near the mid- region, the maximum

von-Mises stress is located at the inner tip of the crack. This magnitude of this stress is lower

(14.09Mpa) than that is found near fixed end. Along with decrease in stress the deformation

decreases and natural frequency increases. A possible explanation is that due to decrease in

breathing effect of the crack, the closeness of the crack decreases, and the stiffness of the plate

increases in this region causing the natural frequency to rise. The minimum stress is found at

the free end of the plate covering almost 33% of the area of it.

(a) Deformation when 1N force applied downwards at Free end

(b) von-Mises Stress

Figure 4.30: Deformation and von-Mises Stress of a Crack at Relative Distance of 50%
at Mid-Region from Static Structural Analysis

76
4.13.5 Stress Concentration and Total Deformation Near Free End
From Figure 4.31, it is found that when crack is located near the free end region, the maximum

von-Mises stress is located near the fixed end of the cracked plate. This magnitude of this stress

is lowest (6.79MPa) than that is found near fixed end and in mid-region. The magnitude of the

stress for this cracked plate and the healthy plate is the same and shows similar deformation.

For rest of the crack locations until the end of free end the magnitude remains constant as

shown in Figure 4.26. As the stress becomes constant in this region, the total deformation and

the natural frequency of mode 1 also remain constant in this region. This is the region the

intersection point is located where the natural frequency of the cracked plate intersects the

equation of natural frequency line of the healthy plate.

In this region, the breathing effect reaches to minimum and deflection of the plate becomes

maximum. A possible explanation for increased natural frequency is that due to increased

deflection of the free end due the presence of crack, the plate is loaded statically with its own

mass. As a result, the stored energy increase and hence increase in natural frequency. The

minimum stress is found at the free end of the plate covering almost 10 to 15% of the area.

77
 (a) Deformation when 1N force applied downwards at Free end

(b) von-Mises Stress

Figure 4.31: Deformation and von-Mises Stress of a Crack at Relative Distance of
84.8% at Free End from Static Structural Analysis
4.13.6 Other Mode Shapes at Three Regions
The explanation of mode shapes are already discussed on section 4.2 and 4.3. Moreover,

section 4.13.2 explains the first mode shape of the three regions of the plate. Rest of the mode

shapes (from mode 2 to mode 6) is explained in this section.

Figure 4.32 shows the rest of the mode shapes (from mode 2 to mode 6) of crack located at

relative distance of 9.1% near fixed end. Similar to healthy plate, for first bending mode and

torsional mode there is one nodal point each. For second bending and torsional mode, there are

two nodal points and so on. The nodal point increases for subsequent increase in number mode

each mode.

78
(b) Modal Deformation at Mode 2

(c) Modal Deformation at Mode 3

(d) Modal Deformation at Mode 4

79
 (e) Modal Deformation at Mode 5

(f) Modal Deformation at Mode 6

Figure 4.32: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of
9.1% Near Fixed End

Figure 4.33 shows the rest of the mode shapes (from mode 2 to mode 6) for crack located at

relative distance of 50% at mid-region. The linear motion in bending modes and the angular

motion in torsional modes behave the same way as explained in section 4.2 and 4.3. The

number of nodal points also remains the same as before. However, the effect of breathing

phenomena decreases in this region than in the fixed end.

80
(b) Modal Deformation at Mode 2

(c) Modal Deformation at Mode 3

(d) Modal Deformation at Mode 4

81
 (e) Modal Deformation at Mode 5

(f) Modal Deformation at Mode 6

Figure 4.33: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of
50% at Mid-Region

Free: Figure 4.34 shows the rest of the mode shapes (from mode 2 to mode 6) for crack located

at relative distance of 84.8% at free end region. The linear motion in bending modes and the

angular motion in torsional modes behave the same way as explain before. The number of

nodal points also remains the same as before. However, the effect of breathing phenomena is

decreased further in this region than in the fixed end and mid-region.

82
(b) Modal Deformation at Mode 2

(c) Modal Deformation at Mode 3

(d) Modal Deformation at Mode 4

83
 (e) Modal Deformation at Mode 5

(f) Modal Deformation at Mode 6

Figure 4.34: Other Mode Shapes (Mode 2 to Mode 6) of Crack at Relative Distance of
84.8% at Free End

84
 Chapter 5
Detection of Cracks

85
 5.1 Detection of Cracks of all Cases
Based on the problem specifications mentioned on section 3.4 and Figure 3.6, the Table 5.1 is

prepared for the detection of crack on the metal plate. This table is self-explanatory and

summarized the percentage of maximum frequency drop for the presence of crack on the plate.

The cracks are detectable based on simulation and mode shapes.

Table 5.1: Summary of Crack Detection in the Research

Frequency Drop
Case Case Description Bending Mode Torsional Mode Remarks
1 Transverse Edge Crack Have Distinct Pattern Have Distinct Pattern Detectable
along Z-axis (AA)
Max Frequency Drop 5.9% at x/L=9.1% on 1st Mode 7.4% at x/L=84.8% on 6th Mode
2 Change of Transverse Similar Pattern as Case-1 Similar Pattern as Case-1 Detectable when
Edge Crack Lengths Larger the Crack Length: Larger the Crack Length: Crack Length is
along Z-axis -Higher the Frequency Drop -Higher the Frequency Drop more
-Intersection Point Moves -Intersection Point Moves
towards Free End towards Free End
Max Frequency Drop 1.8%, 5.9% & 11.4% 1.4%, 7.4% & 20% respectively
(a=10,20,30mm) respectively at x/L=9.1% at x/L=81.8%, 84.8%, 90.9%
on 1st Mode on 6th Mode
3 Change of Transverse Similar Pattern as Case-1 Similar Pattern as Case-1 Detectable when
Edge Crack Widths Larger the Crack Width: Larger the Crack Width: Crack Length is
along Z-axis -No Effect on Frequency Drop -No Effect on Frequency Drop more irrespective
-Intersection Point Moves -Intersection Point Moves of crack width
towards Fixed End towards Fixed End
Max Frequency Drop 5.7%, 5.9% & 6.0% respectively 7.1%, 7.4% & 7.6% respectively
(b=0.5,1.0,1.5mm) at x/L=9.1% on 1st Mode at x/L=84.8% on 6th Mode
4 Change of Crack Very Little change until crack is Very Little change until crack is Detectable when
Depths along Y-axis thorough thorough thorough
Max Frequency Drop 0.87% at Fixed End (c/H<1) 0.92% at Free End (c/H<1)
5.9% at Fixed End (c/H=1) 7.4% at Free End (c/H=1)
on 1st Mode on 6th Mode
5 Transverse Edge Crack Have Distinct Pattern Have Distinct Pattern Detectable
along Z-axis (SS) Similar to Case-1 Similar to Case-1
Max Frequency Drop 5.8% at x/L=9.1% on 1st Mode 7.3% at x/L=84.8% on 6th Mode

86
 Frequency Drop
Case Case Description Bending Mode Torsional Mode Remarks
6 Embedded Transverse Similar Pattern as Case-1 Very Low Detectable
Crack along Mid Z-axis
Max Frequency Drop 4.2% at x/L=6.1% on 1st Mode 0.26% at x/L=90.9% on 6th Mode
7 Longitudinal Edge Very Low Very Low Undetectable
Crack along Z-axis (Surface Defect)
Max Frequency Drop 0.29% at x/L=6.1% on 1st Mode 0.25% at x/L=0% on 2nd Mode
8 Embedded Longitudinal Very Low Very Low but Sharp Drop at Only Detectable only
Crack along Mid Z-axis Free End at Free End
Max Frequency Drop 0.58% at x/L=0% on 1st Mode 1.7% at x/L=84.8% on 6th Mode
7.9% at x/L=87.9% on 6th Mode
(Free End)
9 Longitudinal Edge Very Low Visible at Mid-Region Detectable at
Crack along X-axis Mid Region
Max Frequency Drop 0.3% throughout the 1st Mode 8.9% at y/B=38.1% & 63.1% on
6th Mode
10 Transverse Edge Crack Very Low Very Low Undetectable
along X-axis (Surface Defect)
Max Frequency Drop 0.3% throughout the 1st Mode 0.14% at y/B=50%
on 6th Mode

11 Inclined Crack Moderate (First two Modes) Moderate (First two Modes) Detectable when
Positions near Fixed Inclined Crack at
End on Z-axis 60o to 120o
Max Frequency Drop 4.6% at Angle=75o on 1st Mode 5.4% at Angle=90o on 2nd Mode
12 Inclined Crack Moderate (Last two Modes) Moderate (Last two Modes) Detectable when
Positions near Free End Inclined Crack at
on Z-axis 60o to 90o
Max Frequency Drop 4.5% at Angle=75o on 5th Mode 6.4% at Angle=90o on 6th Mode

87
It is seen that surface defects (case-7 and case-10) for both longitudinal and transverse direction

could not be traced from the research. While the inclined cracks (case-11 and case-12) within

the angle between 60o to 120o could only be traced if the crack positions are within the nodal

region. Torsional modes (case-6) have no effect on embedded transverse crack at mid-axis

along longitudinal direction because of the twisting moment is zero. The embedded crack

would only be detectable through bending modes, with the limitation that the crack position

must be near nodal region. On the other hand, embedded longitudinal crack at mid-axis along

longitudinal direction (case-8) would remain undetectable throughout the axis. The only crack

location where crack detection is possible is at the free end of the plate. It is only found on the

torsional modes where the frequency drop is abrupt and sharp on the normalized frequency

curve. Crack, which is not completely thorough, (case-4) would remain undetectable in the

research. Crack width have negligible effect on frequency drop (case-3). Therefore, the change

in crack length plays vital role in detection of crack. The larger the crack length the higher the

frequency drop (case-2).

Considering all the above and the Figure 4.6 into account, it is observed that number of bends

(i.e. number of nodes) in case of bending increases with increased modes. Similar things

happen in case of torsional modes for increased number of twists. Any crack that is located in

this region experiences massive frequency drop than any other regions. This is because the

influence of stiffness of the plate is greater in this region.

88
5.2 Experimentation
From the modal analysis, it is expected that the significance difference of 6% to 10% and more

could be detectable through experimentation. Moreover, carrying out harmonic analysis could

result change of amplitudes in different positions and using frequency response function (FRF),

amplitude and natural frequency could be obtained. Besides, the variation of amplitude for

cracked plate and the healthy plate could be attained and distinguished. The obtained results

could be validated through experimentation and could be used for detection of crack, although

the change of natural frequency is not very high. Unfortunately, the research only considered

the results from numerical simulation of different cases. Due to outbreak of COVID-19, no

experimentations were conducted. If experimentation would have been done, the results

attained could be verified for further analysis. In that case, harmonic analysis could have been

incorporated to authenticate the results analytically in some specific cases with more significant

differences for detection of cracks.

89
 Chapter 6
Conclusion and Recommendation

90
6.1 Conclusion
Most of the scientific research is conducted with cantilevered cracked beam with different

approaches into account for detecting cracks. These include bilinear effect, modal analysis and

harmonic analysis along with mathematical derivation. The research is inspired because very

few scientific works are investigated on thin metal cantilever plates for detection of crack.

Detailed parametric study is conducted on an edge crack, which is thorough in nature. The

research consists of several cases. Studies of different crack orientations (transverse,

longitudinal and inclined) and crack sizes (crack length, crack width and crack depth) are done

over relative distance along the longitudinal (Z-axis), transverse (X-axis) and depth (Y-axis)

directions. Different materials (aluminum alloy and structural steel) is also taken into account.

The study of each of these cases are simulated on ANSYS 16.0 workbench and designed over

SolidWorks. The results of numerical analysis (modal analysis) is correlated with stress

concentration, deformation, deflection and breathing effect of the crack on the plate. Taking

the natural frequency of the healthy plate as the base of the comparison, any deviation due to

presence in crack is noted and to identify the most significant locations on the plate where the

change is maximum. The work has been validated with a scientific paper [3] of Tao et al.

For a transverse crack along longitudinal (Z-axis) direction (case-1), the change of frequency

is more in torsional modes than bending modes. In both bending and torsional modes, the

maximum frequency drop shift towards the free end with the increased number of modes.

In other observation, it is found that as the crack length increases (case-2), the frequency drop

increases and the intersection point moves towards the free end. This is found in both

simulation results and solutions with regression equations. For all six modes in three distinct

regions (fixed end, mid-region and free end), there is increase in frequency drop with increase

in crack length (further study). Moreover, for larger crack length, there is always more

frequency drop at fixed end (48.1% and 56%) than in mid-region (22.7% and 44.5%) and

91
lowest at free end (1.7% and 26.1%) when considering mode 1 and 2 only. On the other hand,

the frequency drop is higher at free end (29.3% and 32.3%) when considering mode 5 and 6.

It is noticed (in case-3) that as the crack width increases, the frequency drop remains constant

and the intersection point moves towards the fixed end. This is confirmed in both simulation

results and solutions with regression equations. When the crack width increases, in further

study of case-3, the frequency drop increases at fixed region and frequency gain increases at

free region for mode 1 and 2. The maximum significant difference is 9.7% and 9.0% at fixed

end for mode 1 and 2 respectively.

When the crack depth is increased (in case-4), the frequency drop remains constant until crack

depth to plate thickness (c/H) ratio becomes 75% and more. After this point, there is a rapid

drop of frequency drop until the crack is thorough.

It is deduced that metal (case-5) and metal alloy (case-1) have the same normalized frequency

curve as they follow same pattern and trend plotted over relative distance. The torsional modes

found in normalized frequency curves have the same line of progression with little difference.

While the bending modes found in normalized frequency curves take place at certain gap

though they follow the same pattern of progression. In this case, natural frequency of aluminum

alloy is greater than structural steel indicating greater stiffness in aluminum alloy. Both bending

and torsional modes, the maximum drop in frequency shifts towards the free end with increase

in mode.

It is observed that at all the locations the frequency drop in transverse edge crack (case-5) is

massive than embedded mid-crack (case-6). For transverse mid-crack, torsional modes have

little significance with maximum frequency drop of 0.26% at relative distance 90.9%. While

in bending modes, the frequency curves of embedded mid-crack have the same pattern as edge

92
crack but the frequency drop lower (maximum frequency drop 4.2% on 1st mode) than

transverse edge crack (maximum frequency drop 5.8% on 1st mode).

The maximum frequency drop in bending modes is 0.29% for longitudinal edge crack (case-7)

and 0.58% for embedded longitudinal mid-crack (case-8). These are located near fixed end on

1st mode. While in torsional modes, maximum frequency gain is observed at free end in case-

7. However, in case-8, sharp frequency drop is noticed. Both these cracks are undetectable.

Only the embedded longitudinal mid-crack can be detected at free end with the presence of

sharp frequency drop in torsional modes.

The bending modes of both longitudinal edge crack (case-9) and transverse edge crack (case-

10) is negligible along the transverse direction. The frequency drop for longitudinal edge crack

at mid-region is detectable when torsional mode is considered. The maximum frequency drop

is 8.9% on 6th mode. The transverse edge crack (surface defect) remains undetectable.

Most of the frequency curves for inclined crack at relative distance 21% (case-11) and 79%

(case-12) are parabolic or slightly distorted parabolic. Considering the torsional modes, the

lowest frequency drop is found at the starting and ending point of the normalized frequency

curve where the crack angles are 0o and 180o. These are the locations where the crack align

longitudinally at the edge (surface defect). On the other hand, the maximum frequency drop is

found when the crack angle is 90o, the crack location where the crack orientation is transverse

at the edge. For inclined crack near fixed end, mode 1 and 2 are dominant and maximum

significant difference found within this region. While near free end, mode 5 and 6 are dominant

and maximum significant difference found within this region.

93
6.2 Recommendation
As crack length plays an important role, further analysis should be done with increased crack

length. Moreover, static structural analysis should be done of these increased crack lengths, to

understand the behavior of crack more. Experimentation of some specific cases where cracks

are more prone to detection should be performed first to verify the obtained simulation results.

Harmonic analysis should be performed to get wide range of amplitude. A comparison could

be made to note the major differences in detection of cracks in a cantilever beam and thin

cantilever plate.

94
 References
1. Shen, M.-H. H. and Chu, Y. C., “Vibrations of Beams with a Fatigue Crack”,
Computers & Structures, Vol. 45, Issue 1, 17 September 1992, pp. 79-93.

2. Chondros, T.G., Dimarogonas, A.D. and Yao, J., “Vibration of Beam with a Breathing
Crack”, Journal of Sound and Vibration, 239 (1), 2001, pp. 57-67.

3. Tao, N., Ma, Y., Jiang, H., Dai, M. and Yang, F., “Investigation on Non-Linear
Vibration Response of Cantilevered Thin Plates with Crack Using Electronic Speckle
Pattern Interferometry”, The International Conference on Experimental Mechanics
2018 (ICEM18), Brussels, Belgium, 2 (8), 1–5 July 2018, pp. 539.

4. Orhan, S., “Analysis of Free and Forced Vibration of a Cracked Cantilever Beam”,
NDT & E International, Volume 40, Issue 6, September 2007, pp. 443–450.

5. Tufisi, C., Gillich, G.-R. and Aman, A.T., “The Effect of a Crack Near the Fixed End
on the Natural Frequencies of a Cantilever Beam”, Vibroengineering PROCEDIA, Vol.
23, 2019, pp. 37-42.

6. Shinde, Y.D. and Katerkar, S.D., “Vibration Analysis of Cantilever Beam with Single
Crack using Experimental Method”, International Journal of Engineering Research &
Technology (IJERT), Vol. 3, Issue 5, May – 2014, pp. 1644-1648.

7. Praisach, Z.I., Minda, P.F., Gillich, G.-R. and Minda, A.A., “Relative Frequency Shift
Curves Fitting Using FEM Modal Analyses”, Proceedings of the 4th WSEAS
International Conference on Finite Differences - Finite Elements - Finite Volumes -
Boundary Elements, April 2011, pp. 82-87.

8. Gillich, G.-R., Wahab, M.A., Praisach, Z.I. and Ntakpe, J.L., “The Influence of
Transversal Crack Geometry on the Frequency Changes of Beams”, International
Conference on Noise and Vibration Engineering, September 2014, pp. 485-498.

9. Rao, Singiresu S.. Mechanical Vibration, Fourth Edition, Pearson, 2004.

10. Anderson, T.L., Fracture Mechanics - Fundamentals and Applications. Third Edition,
Taylor and Francis.

95
11. Andreaus, U. and Baragatti, P., “Fatigue Crack Growth, Free Vibrations, and Breathing
Crack Detection of Aluminium Alloy and Steel Beams”, The Journal of Strain Analysis
for Engineering Design, 44 (7), 2009, pp. 595-608.

12. Al-Waily, M. and Deli, A., “A Suggested Analytical Solution of Buckling Investigation
for Beam with Different Crack Depth and Location Effect”, International Journal of
Energy and Environment, Vol.7, Issue 3, 2016, pp. 201-216.

13. Charalambides, P.G. and Fang, X.M., “The Mechanics of a Cantilever Beam with an
Embedded Horizontal Crack Subjected to an End Transverse Force, Part A: Modelling.
Mechanics”, Materials Science & Engineering, Vol 5, 2016.

14. Liu, J., Zhu, W.D., Charalambides, P.G., Shao, Y.M., Xu, Y.F. and Fang, X.M., “A
Dynamic Model of a Cantilever Beam with a Closed, Embedded Horizontal Crack
Including Local Flexibilities at Crack Tips”, Journal of Sound and Vibration, Vol. 382
(10), November 2016, pp. 274-290.

15. Soliman, E.S.M.M, “Influence of Crack Inclination Angle on Isotropic Cracked

Cantilever Beam”, J Fail. Anal. and Preven., Vol. 20, 2020, pp. 1065–1080.

16. Jena S.P. and Parhi D. R., “Response Analysis of Cracked Structure Subjected to
Transit mass – a Parametric Study”, Journal of Vibroengineering, Vol. 19, Issue 5,
2017, pp. 3243-3254.

17. Gillich, G.-R., Maia, N.M.M., Mituletu, I.-C., Praisach, Z.-I, Tufoi, M. and Negru, I.,
“Early Structural Damage Assessment by Using an Improved Frequency Evaluation
Algorithm”, Latin American Journal of Solids and Structures, 12(12), 2015, pp. 2311-
2329.

18. Barad, K. & Sharma, D. and Vyas, V., “Crack Detection in Cantilever Beam by
Frequency based Method”, Procedia Engineering, 51, 2013, pp. 770-775.

19. Elshamy, M., Crosby, W.A. and Elhadary, M., “Crack Detection of Cantilever Beam
by Natural Frequency Tracking using Experimental and Finite Element Analysis”,
Alexandria Engineering Journal, Vol. 57, Issue 4, December 2018, pp. 3755-3766.

20. Gillich, G.-R., Tufoi, M., Korka, Z., Stanciu, E. and Petrica, A., “The Relations between
Deflection, Stored Energy and Natural Frequencies, with Application in Damage
Detection”, Romanian Journal of Acoustics and Vibration, 13, 2016.

96