Fatigue Cracks On Welded Steel Plates Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Man's desire to live a contented life has led to a need towards the advancement in technology and innovation. With this advancement, successes and failures have occurred which have caused both loss of life and property; on the other hand these failures have acted as agents of change through which advancement in science technology and innovation of new products, systems and processes, thus helping the betterment of our lives. Understanding of how these failures occur is the task of forensic engineers. The work of a forensic engineer is to investigate projects that do not provide the expected quality of performance for the expected period [1].

Fritz Simons 1986 [2], gives the following four reasons for structural failures, which he aptly calls, "the four horse-men of the engineering appocalypse", as ignorance, avarice, negligence and incompentence. Although failure can be attributed to these casues, natural disasters also have a role in failure of structures.In 2005, Hurricane Katrina casued damage and loss of life when it struck New Orleans.Failure of levees which held back the sea water are attributed to systems design failure by a report issued by the American Society Of Structural Engineers [3].

Disasters or engineering failures tend to spur advacment in innovation and technology. In 2000, Air France Concorde Flight Number 4590 crushed killing all 109 passengers and crew on board. The crush was attributed to a piece of metal that had fallen from a previous flight that had taken of a few minutes earlier, which punctured the aircrafts tyre, causing pieces to tear through the fuel tank casuing the plane to explode. The accident led to modifications to the Concorde, including more secure electrical controls, Kevlar lining to the fuel tanks, and specially developed, burst-resistant tires.

To be able to find the cause or causes of failure, forensic engineers have at their disposal various methods that assist in systems, structural or process failures analysis. From simple observation using their eyesight to the use of x-ray or optical microscope for the identification of trace evidence, as well as photography, are just but a few of the methods that forensic engineers use to determine causes of failure. Other techniques include Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Fractography, Nuclear Magnetic Resonance (NMR) etc. To better understand the failure of structures, various studies have been carried out and the application of advanced investigative methods employed [4-5]. A key feature in all of these methods of forensic analysis is the identification of the failure mechanism by examining the failed component or process, with the choice of method used depending upon the nature of the material under examination. It is thus a prerequisite for a forensic engineer to have a firm grip on various methods employed for material failure analysis as well as properties of materials and material standards.

From the title of the project, Forensic Examination of Fatigue Cracks on Welded Steel-Plates, three areas of concentration that can be surmised from it are application of forensic methods in determining of engineering failures, study of fatigue behaviour and lastly understanding how welded steel structures are affected by fatigue cracks. In essence how forensic methods can be applied in determining fatigue cracks on welded mild structural steel (A36) plates is the area of focus.

A key motivator for having chosen this project is trying to understand the application of forensic methods in solving and/or predicting failure of structures due to fatigue crack and the material of choice being mild structural steel (A36) with the weld material being ASW E6013. The reasons for having chosen this material are: Its good Weldablity properties, its wide application in the engineering industry and lastly the availability of data and studies for fatigue cracks in structural steel.

Three methods applied in this project include numerical analysis where equations are used to determine the fatigue strength and endurance strength of the mild steel, a simulation of the design specimen using the fatigue module in ANSYS Simulation software and lastly subjecting the test piece to a fatigue test in the Laboratory: the data to be collected being the load applied and the cycles to failure. By correlating the three sets of data collected and comparing with studies done by other authors an inference is then made.

1.2 Aims and Objectives

This project will focus on the examination of signs of fatigue cracks on welded steel plates. The key objectives of this project are as stated below:

To examine and identify high stressed points on structural welded steel plates by correlating numerical, simulated and experimental data collected.

To subject the test pieces to dynamic loads till failure and examine the fractured surfaces for signs of fatigue failure and interpret the results.

Identification of the early onset indicators of crack propagation on welded parts is vital when it comes to failure prevention. The test piece used will be of rectangular shape structural steel welded together. This material is chosen because of its good Weldablity, availability of data from various studies which can be used for comparison and its wide applicability within the engineering industry.

This test pieces will be subjected to cyclic dynamic loads with the aid of the Essom TM211 rotating fatigue testing equipment. Preliminary examination of the test piece to identify areas of high stress is to be carried out with the aid of an optical microscope or a magnifying glass. A simulation of the experiment will be done with the aid of ANSYS Software, to aid in identifying the highly stressed locations of the test piece.

To be able to show how the cracks progressed, an examination of the failed surfaces is to be done. Key features to look for are the direction in which the beach marks are moving too, discoloration of surface etc.


2.1 Introduction

Mechanical structures are subjected to repeated loads when in use. These loads can be alternating or cyclic in nature. The result of these loads is the change in the microstructure of the materials used to design the structures. A change in the microstructure of the material occurs when the stress loads/level applied fall beyond the stress range of the material in use. This causes the material to fatigue.

It has been estimated that 80 percent of mechanical structural failure can be attributed to these loads or fatigue failure. A study carried out for the United States Congress by the Commerce Department's National Bureau of Standards estimates the cost of fatigue failure to be $119 Billion a year, this representing roughly 4 percent of the country's Gross National Product. The study further suggests that 26 percent of this cost can be reduced by using new technologies in design and material science [6].

2.1.1 History of Fatigue Studies

The study of fatigue failure can be traced back to the early 19th Century. It's during this time that the railroad was a common means of transportation and that failure of the railroad occurred at the shoulders of the axis. Some major milestones in fatigue studies [7] are as follows:

1839: The term fatigue is introduced to describe the failure of materials due to repeated load application.

1843: Dangers of stress concentration in machine and recognition of distinctive features of fatigue failure are noted by W.J.M Rankine.

1860: August Wöhler introduces the use of S-N curves and the concept of fatigue limits after studies on railroad axles.

1910: Empirical laws are introduced by Basquin to characterize the fatigue limit of materials.

1913 and 1921 Inglis and Griffith respectively using stress analysis and energy concept provide a mathematical tool for quantitative treatments of fracture in brittle material.

1939: A method to determine the stress and displacement field ahead of sharp crack tips is developed by Westergaard.

1954: Coffin and Manson propose an empirical relationship between the number of load reversals to fatigue failure and the plastic strain amplitude. They also discovered that plastic strains are responsible for cyclic damage.

1957: Irwin introduces the stress intensity factor (K).

1960 and 1962: Dugdale and Barrenblatt introduce simple crack models.

1961: Fatigue crack propagation rate per cycle is suggested to be related to the stress intensity factor, by Paris Gomez and Anderson.

1970: Elbers work on fatigue cracks creates a base for the development of crack-closure concept.

2.2 Definition of Fatigue and Fatigue Failure

In order to solve cases involving fatigue failure, forensic engineers need an in-depth understanding of fatigue failure. To answer the question "why failure occurs?" the Forensic engineer needs to know the "where", "how" and "when" fatigue failures would occur. These three questions can best be answered when an engineer knows where stress concentrators can be located, how fatigue cracks propagate and finally fatigue life predictions.

In lay-mans term the word fatigue is used to describe the "tiring" of a material due to repeated load applications. These loads can be cyclic or alternating in nature. In his book Shigley [8] explains that repeated load application does cause ultimate failure and that the load at which failure does occur is sometimes found to be less than the yield strength or ultimate strength of the material. As a result failure of a structure due to fatigue will occur long after initial material deformation has taken place.

Thus fatigue failure, which occurs as a result of fatigue cracks, is due to fatigue of the material or structure.

2.2.1 Fatigue behavior of mild structural steel

The behavior of mild steel under loading is predominantly determined by its microstructure. At the elastic limit (yield point), elasticity-ability of a material to return to its original state after a load is removed from it, is lost and plastic deformation starts. Deformation at this point onwards is accompanied by the formation of slip bands which act as stress raisers thus creating points at which cracks can be initiated [9]. It should be noted that failure as a results of these crack propagation can occur at both low-less than 105[10], and high-greater than 105 endurance cycle ranges. Fatigue cracks grow through three stages with continued load application.

2.2.1 Stage I: Crack Initiation

Initial crack initiation starts at a microscopic level when dislocation occurs at slip planes. The buildup of dislocations forms persistent slip bands-PSBs, Figure 2-1, which form extrusions or intrusions. These create areas that are highly stressed, stress raisers, and with repeated load application cracks do start to form. Observation of these can be done with the aid high magnification microscopes. See Figure 2.1 below

Picture courtesy of NDT Resource Center

Figure 2-1: Persistent Slip Bands (PSBs) and crack initiation [11].

In her article Susan Kristoff [12] gives three causes of crack initiation as geometric inconsistence in a component, mechanical or thermal fatigue and material inclusions, impurities or loss due to wear or corrosion. Other studies carried out have found out that localized plasticity at grain boundaries has been a key initiator of cracks in cubic material that have been subjected to low amplitude cycles [13]. Stage II: Crack Propagation

Continued cyclic load application causes the change from micro-cracks to macro cracks. Formation of surface features on the failed pieces are present as a result of the continuous rubbing together of the surfaces thus forming dark plateau like features. These features are usually dark in color and are visible when viewed with an optical microscope, (Figure 2-2.). A common term given to them is beach marks or clamshells. The frequency at which the cracked surfaces are rubbing each other causes the shade of the marks to vary.

Figure 2-2: Crack Propagation-Beach mark directions [14]. Stage III: Ultimate Failure

At this stage the material left will not be able to bear up the stresses being applied to the structure as such failure of the material occurs.

2.2.2 Characteristics of Fatigue Failure/Fractography

Examination of material surfaces to determine cause/s or mechanism of failure is termed as Fractography. Examination of these surfaces is done with the aid of microscopes, with different magnifications being able to show different features. Common features looked for include inclusions, voids, foreign material stress raisers etc. Figure 2-3 shows the point of origin of the failure and its direction. As with other fracture modes, proper identification of fatigue failure requires understanding of the fracture behavior of the particular material.

A clear distinction can be made when it comes to identifying failure due to static loading and dynamic loading. Static loading failure tends to cause the elongation of the material as these causes the yield strength to be exceeded gradually, unlike dynamic or fatigue failure which does not offer any signs or warnings until failure has occurred. Figures 2-4 and 2-5 shows how failure due to fatigue on steel bolts appear under magnification and when viewed from a scanning electron microcopy (SEM)

shaft fracture surface.JPG

Figure 2-3: Fatigue failures [15]

Fatigue crack growth on a steel stub axel for a road vehicle. It can cleary be seen that the crack originated in the direction of the arrow.

Figure 2-4: Fatigue crack growth on Steel stub axel [16].

Steel bolt

This high tensile steel bolt failed under low stress high cycle conditions with a fatigue crack running from 9 o'clock as shown by the beach marks. The SEM image of the fatigued surface (shown left) is found to have no striations due to the high yield strength and high cycle conditions.

Figure 2-5: Fatigue failure of high tensile steel bolt under low stress cycle [17].

2.2.3 Cases of Fatigue Failure accidents

Throughout the years fatigue failure has been the cause of various catastrophic failures. Examples of infamous fatigue failure cases include:

Crash of 3 de Havilland Comet planes in 1954.

The failure of the fuselage of the Comet aircraft due to continued pressurization and depressurization resulted in formation of fatigue cracks.

Boston Molasses plant in 1919.

Fatigue cracks that had initiated from a manhole cover at the base of the tank are suspected to have been the cause of failure of the molasses tank that caused widespread damage within the North End neighbourhood of Boston, Massachusetts in the United States.

2.3. Welding and Weld Failures

2.3.1 Basics of Weld Joints

Welding offers a cheap quick and affordable way of joining two pieces of metal together. Designing and fabrication of proper welds calls for a specialist input as well as understanding the basis and significance of weld details [18]. Various standards for weld designs such as American Welding Society (AWS) are used to describe technical requirements for a material, process, product, system or service. Examples of weld joints available include butt, corner, and edge lap and tee joints, Figure 2-6.

Figure 2-6: Types of weld joints [19].

2.3.2 Weld Imperfections

For most structures, the area that accounts for the greatest failure is the point at which joints occur. Weld joints on closer examination have imperfections, which can be categorised as planer imperfections, volumetric imperfections and geometric imperfections [20].

Planer imperfections are those that tend to have sharp crack like features, which tend to reduce the weld strength as such cause brittle fracture initiation. They are microscopic and include hydrogen cracks, reheat cracks, solidification cracks and lamellar tears

Volumetric imperfections, although they tend to lack any effect on fatigue behaviour, they tend to reduce the load bearing area of the weld thus reducing the static strength of the weld. Examples include porosity, slug inclusions etc

Geometric imperfections tend to elevate the stress level around the weld to values over those designed for the weld. They include undercuts, weld ripples, over fills etc Figures 2-7 and 2-8.

Figure 2-7: Undercut Overlap, Under-fill and incomplete joint weld penetration [20].

Figure 2-8: Incomplete Fusion of weld [20].

2.4 Forensic Examination Methods

The objective of any forensic examination is the determination of cause or causes of failure. In order to do this forensic engineers do employ various methods, from the most basic use of sight to the most advanced techniques, which involve high-powered microscopy. The choice of method to be employed will be the one that offers the best results under the conditions being used, as such knowing the trade-offs and limitation off these methods aids the forensic engineer in determining which method to use.

2.4.1 Visual Examination

The most basic and mundane type of examination that a forensic engineer can use is the use of sight. We tend to see but not perceive when it comes to observing things. As such a good forensic engineer will see and perceive the vital information required to give an initial assessment as to the cause of failure. A disadvantage to this method is that minute vital clues can be missed as such a wrong conclusion can be reached.

2.4.2 Photography

Photography offers a way through which evidence can be captured and preserved. Key points to consider when taking photographic evidence include lighting conditions, correct lens angles, and collection of different views. The use of digital cameras is becoming common, as they offer high-resolution pictures as seen in Figure 2-9. A simple 35 mm camera with good lighting conditions will provide quality pictures that can be used to assess for trace evidence.

Figure 2-9: Photograph taken to show skid marks formed by faulty ladder [21].

2.4.3 Microscopy

Microscopy involves examination of test piece surfaces with the aid of microscopes. Optical, Scanning probe and Electron microscopy are the three branches of microscopy. Microscopy offers a choice in which the micro and macro structures of the material can be observed in detail. As such minute features that would be critical in identifying failure can be observed. Features such as persistent slip bands are only discernable with the aid of a microscope at high magnifications. Macrofractography, or the examination of fracture surfaces at low magnifications (≤50), is the cornerstone of failure analysis. In his article George F Vander [22] talks about the importance of Macrofractography as it gives us an overall understanding of the fracture, aids in determining sequence of fracture, origins of fracture in as well as identifying relevant features that can be key in determining fracture initiation and propagation

2.4.4 Simulation Analysis

Simulation analysis has been used by engineers to improve products before production is done. Computerized analysis is a recent development and this offers a fast easy and inexpensive way of examining a product for faults as well as rapid prototyping. Various examples of software's are currently being used in the market to assist with simulation analysis. ANSYS Software is one such example, Figures 2-10 and 2-11, that develops Finite Element Analysis as well as Computational Fluid Dynamics analysis.

Figure 2-10: ANSYS Mechanical running on Windows XP [23].

Image courtesy of ANSYS Inc

Figure2-11: Contour plot of fatigue life [24].

Forensic Engineers apply simulation analysis as a post-mortem method in identification of probable causes of failure. Examples of cases in which forensic engineers have applied simulation analysis software include [25]:

The Destruction of the space shuttle challenger- January 28, 1986

The Crush of TWA Flight 800- July 19, 1996

The Collapse of the World Trade Centre- September 11, 2001

2.5 Engineering Standards and Forensic Studies

Standards are set of regulations through which an industry governs or tries to maintain uniformity when it comes to product development. These standards do cover a wide range of issues-from material composition, material strength, joining processes, design requirements such as size, shape or geometry of the material, and various technical aspects of engineering. These standards are established either within local regions or as international standards.


3.1 Introduction

Material failure due to dynamic loads has been shown to starts at microscopic levels where the material bonds are misaligned due to excess load application. This misalignment creates slip bands that tend to offer areas of high stress-stress raisers. Various studies have been carried out and data accumulated on how micro cracks can be initiated and propagated till ultimate failure of a material [26-28].Continuous load application-which can be of variable or constant amplitude, tends to elongate this micro- cracks at the slip band. Studies carried out have shown that rapid micro crack growths occur even at loads less than the yield strength and high cycle fatigue of the material [29].

The study of fatigue failure on welded joints is made more complicated by the fact that the weld is a "mix" of metals-base material and welding rod material. As such, the analysis of this sort of failure has been mainly through experimental data that carried out over the years. Various studies carried out have produced data that has been of significance when it comes to design of weld joints.

Experimental studies for welded joints have mainly focused on two methods:

Varying the maximum stress from test to test while maintaining a constant stress ratio (R) - preferred by vehicle and machine designers

Varying the stress range, (Sr) while maintaining a constant minimum stress (S)- preferred by structural engineers

Design and Analysis

Three common methods used in the design and analysis of structure failure due to cyclic loading are: stress life, strain life and linear elastic fracture (LEFM) methods.

Of these three methods the stress life method is the most common method used. Significant support data is available from this method and its implementation is simple. It is based on stress levels and has the disadvantage of being inaccurate at low cycle application.

Strain life method tends to focus on plastic deformation at areas of high stress levels-stress raisers. It is best used at low cycle applications. The third method linear fracture method assumes the presence of a crack as such the rate of crack growth rate with respect to stress intensity is determined.

Experiment Work

In order to better understand fatigue failure in mild structural steel, a Laboratory experiment is carried out on a rectangular shaped butt welded mild structural steel plates. The tests involved subjecting the test pieces to variable loads till failure and recording the number of cycles to failure, then examining the surface of the failed piece with the aid of a magnifying glass and optical microscope.

3.3.1 Objectives of the Experiment

To examine and identify high stressed points on structural welded steel plates by correlating numerical, simulated and experimental data collected

To subject the test pieces to dynamic loads till failure and examine the fractured surfaces for signs of fatigue failure and interpret the results

3.3.2 Equipment Used

The following are the equipments used for carrying out the experiment

Optical microscope: Used to examine the surface of the failed specimen.

Ordinary Magnifying glass: Used to examine and identify flaws on the surface of the weld joint as well the specimens length

Essom TM211 Fatigue testing machine: This machine as shown in the sketch in Figure 3-1, is used to perform the fatigue test.

ANSYS v11 Mechanical Workbench:

fig 6-18

Figure 3: Rotating Cantilever beam fatigue test [30].

3.3.3 Specimen Preparation

The choice of structural steel as the material for this project is based on three key reasons:-

Good Weldablity.

Availability of data for comparison.

Wide applicability in engineering industry.

Material Composition

Table 3-1 shows the mechanical and chemical composition of the structural steel used for this experiment.

Material Type

Ultimate Strength


Yield Strength



(for Flat plates)









Mild Structural Steel









≤ 0.09

Table 3-1: Test piece properties

The Table 3-2 shows the standard and mechanical properties of the welding rods used.

Weld Material Standard

Weld Material strength

Yield Strength

AWSE6013 or

JIS D4313

520 MPa

470 MPa

Table 3-2: Weld Material Properties

Specimen Design

In order to be able to fit the test pieces to the machine, a few modifications to the standard fatigue test pieces is required. Figure 3-2 shows the final assembly of the test piece used whereas Figure 3-3 gives the views and dimensions.

Weld MaterialC:\Documents and Settings\user\Desktop\FYP Test Piece Catia Drwn.jpg

Figure 3-2: Final CATIA Design of test piece

Figure 3-3: Orthographic view of specimen

The final test piece when fabricated and showing the location of the weld and the weld material standard used is as shown in Figure 3-4.

Location of Weld joint

[AWSE6013 Weld Material standard]


Figure3-4: Final design of fatigue test specimen

3.3.4 Experimental Procedure

For this study, the experiment involved three stages through which various data was collected. They include:

Initial Non-destructive testing (NDT) stage

Initial examination of the test pieces was done so as to identify any signs of fatigue crack initiators. This involved examination of the, weld penetration, signs of cracks on the weld and any other visual inconsistency.

Destructive Testing (DT) Stage

The second stage involved performing a fatigue test using the Essom TM211fatigue testing machine. Care was taken to ensure that the test pieces were correctly placed.

An initial load of 900 N was used with decrements of 100 N per test. Each piece was subjected to variable loads till failure. The numbers of cycles to failure for each piece are recorded.

Examination of surface of failed piece.

Stage three involved examination of the surface of the failed pieces. Here directions of beach marks, signs of brittle failure are observed and indicated on a schematic of the test piece.

The data collected is then used to plot an S-N curve.

ANSYS Simulation Procedure

Simulation of the design using the Fatigue simulation module in the ANSYS Simulation software is performed and the following data collected: contour plot of fatigue damage, contour plot of fatigue life and S-N curve for the design. Figures 3-5 and 3-6(a) and (b) depict the steps taken for the setup of the simulation workbench to perform a fatigue simulation analysis.

Figure 3-5: ANSYS Simulation workbench setup for Fatigue simulation flowchart.

Step 5

Step 3vlcsnap-2010-05-01-10h35m50s255.png

Figure 3-6(a): ANSYS Simulation workbench setup for Fatigue simulation Steps

Step 11

Step 8

Step 9

Step 10

Step 7

Step 6

Step 4vlcsnap-2010-04-25-21h30m24s196.png

Figure 3-6(b): ANSYS Simulation workbench setup for Fatigue simulation Steps

3.3.6 Data Collected

Three sets of data will be collected for analysis, simulation data, experimental data and visual data as well as numerical data. These data will then be analyzed by showing a relationship between them and then drawing of conclusion.

Visual Data:

Direction of beach marks on failed piece

Stress raisers location

Simulation Data:

Location of high stress

Stress distribution on the weld

Experimental data:

Number of cycles to failure

Load applied


4.1 Introduction

Point to note is that, a butt weld is used: its analysis will be similar to that of the base material. The region that will need attention will be the heat affected zone-HAZ as this is the region that the base metal will tend to mix with the weld material as such a non-homogenous composition mix of metal is formed. Analysis of this section would best be done by applying the Scanning Electron Microscopy-SEM to examine the microstructure of the material formed. This part of the analysis is left as future works.

Results collected have been grouped into three categories:

4.1.1 Numerical Results

Numerical results are calculated to determine the yield stress and endurance strength of the structural steel test piece. The results are then used to plot an S-N curve that is compared with that obtained from the simulated and experimental results. Particular attention is paid to the endurance limit of the joint, as crack initiation starts here.

4.1.2 Simulated results

Simulated results are obtained by using ANSYS Inc Software. The ANSYS Fatigue Module of this software is used to come up with the results. This module has the capability of simulating performance of designs under anticipated cyclic loading conditions over a product's anticipated life span. The module integrates both stress life and strain life analysis with a variety of mean stress corrections. It does provide a variety of results such as contour of fatigue life, damage, stress biaxiality, fatigue sensitivity and factors of safety.

To best interpret the result, contour plot of fatigue life, contour plot of fatigue damage and S-N curve will be collected. Additional plots that will be used are contour plot of factor of safety, Equivalent stress contour plots and maximum shear- stress plots. The contour plot of fatigue life shows the critical locations of the test piece at various life cycles.

4.1.3 Experimental Results

The experimental data collected mainly involved the number of cycles to failure of the test pieces. This was then used to plot the experimental S-N curve. A visual observation of the test pieces failed surface is done to ascertain the direction or type of failure.

4.2 Summary of Results

4.2.1 Numerical Results

Studies carried out by Mischkle [31] have shown that endurance limit can be related to tensile strength. These studies have shown the following relation for steel, Eq. (i):

Since the material used is structural steel with an ultimate strength (Sut) of 550 MPa, and the weld material had strength of 520MPa and the analysis of the weld joint can be done the same way as that of the base metal, the equation(ii) is used for the numerical calculation of the endurance strength (Se) of the weld is:

This gives an endurance limit of 262.34 MPa. To calculate the fatigue strength, equation (iii) is applied at three key cycle points of 100, 103 and 106.

To calculate the values of "a" and "b" the equations (iv) and (v) below are used and the corresponding values are obtained, where "a" and "b" represents the points defined by 103 and 106 cycles respectively.

Applying the constants 'a' and 'b' to equation (iii) and the number of cycles 'N', Table 4-1 is obtained.

Cycle, N

Fatigue Strength, (MPa)















Table 4: Numerical Fatigue Strength Calculations.

These values are used to plot the S-N curve for numerical data as shown in Figure 4-1 below.

Figure 4-1: S-N curve for numerical data

4.2.2 Simulated Results

The chart below, Figure 4-1 shows the conditions applied and the results needed in order to perform the simulation.

Figure 4-2: ANSYS Fatigue Simulation analysis flow chart

With the parameters above kept at a constant my only variable factor was the load applied for each simulation carried out. Two key results that collected are:

Fatigue Life: This plot represents the maximum and minimum cycles to failure for the specimen at a particular loading, Figure 4-3.

Fatigue Damage: This plot represents the ratio of design life over available life for the specimen at a particular loading, Figure 4-4.

These two results are represented by the contour plots as shown in Figures 4-3 and 4-4 below.


Figure 4-3: Contour Plot for Fatigue Life

The plot above (Figure 4-3) shows the fatigue life of the test piece at a loading of 100 N and shows the piece has a minimum cycle of 634,730.


Figure 4-4: Contour Plot for Fatigue Damage

The data collected from the simulation is presented below in Table 4-2.



Fatigue Life


Safety Factor














































































Table 4-2: Results from ANSYS Simulation Report for the test piece

From the data tabulated above (Figure 4-2), Table 4-3 is extracted from it by collecting the minimum Fatigue life cycles and the corresponding loading. The minimum life is taken because it is the least number of cycles required to cause the spacemen to fail.

Loading (N)






















Table 4-3: Data for S-N Curve from simulated results

Figure 4-5: S-N Curve of Simulated data.

Other contour plots collected included:

Equivalent Stress Contour plots, Figure 4-6 :

Simulation Picture 2

Figure 4-6: Maximum Equivalent (von-Mess) Stress contour plot.

Maximum Shear Stress Plots, Figure 4-7:

Figure 4-7: Maximum Shear Stress contour plot.

Fatigue Safety Factor Contour plot, Figure 4-8:


Figure 4-8: Safety Factor contour plot.

4.2.3 Experimental Results

Data collected during the experiments at the Material lab is presented in Table 4-4 and this was used to generate the S-N curve shown in Graph 4-3.

Test Piece Specimen



Cycles to failure

























Table 4-4: Fatigue Test Experimental results

Figure 4-9: Experimental S-N curve plot.

FYP Test Piece

Figure 4-10: Surface of failed test piece.

Figure 4-11: Comparison of numerical, simulated and experimental S-N curves.

4.3 Discussion on Findings

Based on the three areas of which results were collected, the following can be observed:

The weld material did not fail

Failure occurred at the base of the test piece where a change in cross section was

Examination of a failed experimental test pieces showed a ductile failure that was rapid

From the simulated results, the contour plot of fatigues life shows us that the test piece had a life of 634,730 cycles at a loading of 100 N which when compared to that collected from the numerical result of 106, shows a slight deviation. Further observation of the simulated results show a decrease in the maximum shear stress as well as the equivalent stress as the loading is increased.

One key reason that can account for the source in error of the number of cycles produced for the simulation and experimental results is that in real life cyclic loads are usually random in that we do not have constant amplitude whereas when performing of the simulation an assumption is made where by the loads are considered to have a constant amplitude. Producing of real life cyclic loading during experiments is impossible.

Observation of the fatigue life showed that as the load was decreased, the minimum life of the test piece at the point of fracture increased. Its minimum safety factor as well as damage of the test piece at this point increased.

The location of the fatigue failure can be observed to have occurred at the base of the test pieces. This can be seen both on the contour plot for maximum stress as well as that of fatigue life. This is further reinforced by the failure of the test pieces at the same location during the laboratory experiment carried out.

From these results we note that the weld did not failure. This can be attributed to its mechanical properties. From Table 3-2, we note that the yield strength of the weld material is 470 MPa as compared to that of the base material of 250 MPa. From our theoretical studies it shows that the base material should be the first to yield/fail should the test piece be subjected to repeated loading, this is due to the fact that the weld material has higher yield strength as compared to that of the base material. This is verified by both the simulation and experimental results.

On observing the failed test piece surface I noticed a failure that was similar to that of a rapid ductile failure. Absence of smooth beach marks can be observed on the failed surface; instead we notice a rough inward directional failure. This failure is common with surfaces that have had a rapid brittle failure. An observation that was also noted was the formation of elliptical patterns on the surface with different colors, Figure 4-10.

4.4 Challenges of the project

While carrying out this experiment, I did encounter some challenges that required me to "think out-of-the-box". Some of these challenges required long periods of contemplation through which various ideal solution I had to …

Some of the challenges I encountered are as summarized below

Load Cell Weights

My initial calculations showed that the weld would fail at a stressed level of 2385 N and the initial designed work piece had an area of 10 mm by 9 mm. The maximum load available at the Material lab was 100 Kg (1000 N). This fell short of what was required to cause the material to fail.


Two possible solutions were available, one involved redesigning the whole piece and having it fabricated a fresh and the second involved reducing the flat area of the work piece.

I opted for the second solution as it involved reduced time and cost. The uniformity of the weld joint was not affected with this reduction in area. The weld was done from top to bottom of the work piece as such reducing the area had little effect on the welded work piece. The final test piece has an area of 10 mm by 4 mm.

Work Piece Fabrication

Initially I was to fabricate the work piece at the Mechanical Workshop lab and the CAM Lab. The steel rods I had purchased for the experiment were of length 6 meter by 25 mm diameter. This had to be reduced into smaller pieces of length 200 mm. this was done at the mechanical workshop lab. See Figures 4-12 and 4-13.


Figure 4-12: Base Section of the test piece.


Figure 4-13: Front Section of the test piece

In order to produce the above shapes the code shown in Figure 4-13 is used with the CNC machines located at the CAM labs. The initial challenge to this process was which of the two processes to start with first, turning or milling.

BILLET X155 Y25 Z25



G21 G95

G91 G28 Z0

G28 X0 Y0

M06 T1

M03 S1500


G00 X40 Y12.5 Z25

G01 Z-0.5 F60

G01 X132
























































































M06 T2

G00 X138.5 Y24 Z-6

G01 Z-25


G01 Z5

G00 X33.5 Y1

G01 Z-25


G00 25

G91 G28 Z0

G28 X0 Y0


Figure 4-14: CNC Code used for test piece fabrication


5.1 Conclusion

Overall the project objectives were met. From the data collected it can be inferred that subjecting a welded mild structural steel will cause fatigue cracks to originate from locations that would exhibit a reduced fatigue life and factor of safety as the load is increased. The significance of this study is mainly in the practical application of structures made of mild structural steel and which are subjected to cyclic loads for example ships, cranes, bridges, off-shore structures etc.

5.2 Recommendation for future works

It's been said that prevention is better than cure; the same principle applies to fatigue failure. Being able to identify or recognize early signs of fatigue failure would prevent catastrophes from happening. As such future efforts should be placed on pre-emptive methods of rather than after the fact analysis of failures. One such area would be the advances in the development of intelligent materials. This would be materials that would counter effects of fatigue failure such as crack initiation. Advances in Nano-technology would aid in the development of such material.

Another method that would be of interest would be the advances made in fluid mechanics when it comes to understanding speed of sound. Sonification involves the representation of data in the form of sound. With the help of computers and information obtained from sensor transducers, faint sounds can be detected. The method can be used to "hear" flaws in materials, thus a way of identify early signs of crack initiation and propagation.

Two topics that I would propose for future studies are:

Application of simulation as a method in forensic studies within the automotive industry

The automotive industry is one of the leading industry in which product failure results are catastrophic. As such any failure of a product is viewed as a challenge to the quality of the brand. In 2009 and early 2010, the automotive industry experienced one of its largest recalls in automotives as a result of fault designs. The role of a forensic engineer is thus put to test to identify, recommend and in some cases provide technical or scientific evidence during court cases. Simulation provides a way through which cases of design failure or inadequacy can be identified and rectified before failure. This study should involves a student applying one of the simulation software's available in identify, recommending and presentation of probable causes of failure of a design within the automotive industry

Sonification as a forensic method in predicting signs of fatigue failure in Aircraft turbine blades

Vibration poses one of the greatest risks during aircraft engine operation. One of the key parts in an aircraft engine that experiences the greatest vibration is the turbine blades. With improved design in terms of material this problem can be reduced significantly. One way of further improving on the design of the table blade in terms of fatigue resistance due to vibrations is to perform simulations which predict onset of failure or by having systems that monitor the vibration levels as the engine operates.

Sonification is the representation of data in the form of sound and vibrations produce sound. Thus the representation of the vibration levels can be used to detect early signs or stages of crack initiation and propagation. The project should involve the design of a system that would represent these vibrations as sound.