POLISH MARITIME RESEARCH 2 (90) 2016 Vol. 23; pp.

71-82
10.1515/pomr-2016-0023

ANALYSIS OF FATIGUE CRACK GROWTH IN SHIP STRUCTURAL

DETAILS

Heba W. Leheta
Ahmed M. H. Elhewy
Helmy A. Younes
Naval Architecture and Marine Engineering Department, Faculty of Engineering,
Alexandria University

ABSTRACT

Fatigue failure avoidance is a goal that can be achieved only if the fatigue design is an integral part of the original
design program. The purpose of fatigue design is to ensure that the structure has adequate fatigue life. Calculated
fatigue life can form the basis for meaningful and efficient inspection programs during fabrication and throughout
the life of the ship. The main objective of this paper is to develop an add-on program for the analysis of fatigue crack
growth in ship structural details. The developed program will be an add-on script in a pre-existing package. A crack
propagation in a tanker side connection is analyzed by using the developed program based on linear elastic fracture
mechanics (LEFM) and finite element method (FEM). The basic idea of the developed application is that a finite
element model of this side connection will be first analyzed by using ABAQUS and from the results of this analysis
the location of the highest stresses will be revealed. At this location, an initial crack will be introduced to the finite
element model and from the results of the new crack model the direction of the crack propagation and the values of the
stress intensity factors, will be known. By using the calculated direction of propagation a new segment will be added
to the crack and then the model is analyzed again. The last step will be repeated until the calculated stress intensity
factors reach the critical value.

Keywords: Fatigue crack, linear elastic fracture mechanics (LEFM), finite element method (FEM)

INTRODUCTION history of the ship [1, 2]

Although a ship may be designed to withstand the ultimate In the past, ship structure design often did not incorporate
stresses imposed by heavy storm waves, failure could occur explicitly an analysis of fatigue. However, as stated in the
due to apparently low stresses generated by the continuously subject-matter literature , the main reasons for interest
encountered smaller, although significant, ocean swells, as to studying fatigue cracking is, in the last few years, the
well as impact loads resulting from the routine loading and introduction of high tensile steels and the development of very
manoeuvring of a ship. When assuming that a ship has a large ships (e.g. VLCC, ULCC) which results in increasing
life of 20 years and operates at sea for 75% of her life time, the stresses in critical structural details and decreasing its
and knowing that the medium period of wave encounter is fatigue life. Although the allowable stress methods are used
approximately equal to 5 seconds, the ship will experience in the design of majority of ship structures, more and more of
nearly 100 million loading cycles throughout her life. the new designs incorporate detailed analysis methods. [3,4]
Structural failure of a ship may result in consequences that Fatigue failure avoidance is a goal that can be achieved only
range from simple leaks to severe loss of lives. The failure if the fatigue design is an integral part of the original design
consequences are functions of the ship’s age, the damaged program. An improper design may lead to an unacceptable
component, its location, and the maintenance and repair catastrophic fatigue failure resulting in loss of life and damage

POLISH MARITIME RESEARCH, No 2/2016 71

to the environment. Non-catastrophic fatigue failures are also of cumulative damage (Miner rule). Another approach to
unacceptable due to difficulty and cost of repairs as well as the fatigue analysis is fracture mechanics which is discussed in
need to increase costly inspection and maintenance intervals. details in the following section, separately.
The aim of fatigue design is to ensure that the structure has Predicting fatigue damage for structural components
adequate fatigue life. Calculated fatigue life can also form subjected to variable loading conditions is a complex issue.
basis for meaningful and efficient inspection programs during The first, simplest, and most widely used damage model is
fabrication and throughout the life of the ship [3, 5]. the linear damage. This rule is often referred to as Miner’s
The finite element method is a numerical procedure that rule (1945). However, in many cases the linear rule often
can be applied to obtain solutions to a variety of engineering leads to non-conservative life predictions. The results from
problems e.g. stress analysis, heat transfer, fluid flow, etc. this approach do not take into account the effect of load
Such problems may be steady, linear, or nonlinear ones. sequence on the accumulation of damage due to cyclic fatigue
Yet, it was not until 1960 that Clough made the term “finite loading. Since the introduction of the linear damage rule
element” popular [6]. There are many finite element software many different fatigue damage theories have been proposed
packages available in the market however, most of them do to improve the accuracy of fatigue life prediction [3, 8, 9].
not have a built- in function for fatigue analysis. The few that Ship structure design often did not include explicit
have a built -in function for fatigue analysis can be used to treatment of fatigue by means of an analysis. However,
study fatigue crack propagation along a predefined path. For with the increasing in using higher strength steels and the
example, ABAQUS ver. 14 software package has either static increased cyclic stress ranges, fatigue analysis of structures
crack analysis or crack propagation analysis but the crack is increasingly required. Although the developed allowable
propagation path must be known before the analysis. The stress methods ( which use S-N curves) are applied in the
case of predefined path crack analysis may be practical for the design of majority of ship structures, more and more of the
analysis of lamination cracks, e.g. in fibre glass composites. new designs incorporate detailed analysis methods (which
However, this feature cannot be used for crack in steel ship use fracture mechanics) [6].
structural details where the crack path is unknown. The material resistance to fatigue failure primarily depends
The aim of this paper is to evaluate fatigue analysis by using on the characteristics of detail/joint geometry, material
linear elastic fracture mechanics ( LEFM), and to achieve chemical composition and mechanical properties, and the
this aim an add-on program is developed to analyze fatigue service environment. The material resistance is typically
crack growth in ship structural details. The program will be determined in laboratory tests by the application of constant
an add-on script in ABAQUS software. The application was amplitude stress cycle on various detail/joint geometries
developed by using python programming language which until fatigue failure occurs. By carrying out similar tests for
can be incorporated into ABAQUS to extend its features. different stress amplitudes a relationship between the stress
The fatigue analysis techniques are presented in the next amplitude, S , and the number of cycles, N , is established.
section, where a summary on cumulative damage approach The S-N curves developed for simple details (i.e., stiffener,
emphasizing linear cumulative damage approach (Miner’s cut-out, etc.) account for the peak stresses and can be directly
rule) is discussed. A detailed discussion of fracture mechanics used with the member nominal stresses [10] .
especially linear elastic fracture mechanics, LEFM, is The application of linear elastic fracture mechanics, LEFM,
presented in Sec. 4. Also crack stages and stress intensity in estimating the life of the structure assumes that cracks
factors are discussed. The application of LEFM for fatigue has already initiated in the material. The behaviour of the
analysis is given to show fatigue life calculation by using crack can be predicted under anticipated service loading.
LEFM (Paris law). The estimated behaviour is used to schedule inspection and
maintenance in order to assure that defects do not propagate
METHODS FOR FATIGUE ANALYSIS to a catastrophic size [3].

A broad view of the common methods for fatigue analysis FRACTURE MECHANICS APPROACH
is shown in Fig. 1 [7]. Fatigue analysis may be carried out by
methods based on fatigue tests (S-N data) and estimation This section provides an introduction to the important

Fig. 1 Fatigue Analysis Techniques

72 POLISH MARITIME RESEARCH, No 2/2016

aspects of linear elastic fracture mechanics (LEFM) and tip stress intensity factor is an extremely useful parameter
shows how it can be used to describe and predict fatigue crack to address crack growth behaviour as long as the bulk of the
growth rate and fatigue life. It provides a general background material is elastic and plastic deformation is limited to a small
for the concepts of fracture mechanics and numerical tools region at the crack tip [12] .
needed for fatigue design involving crack growth and fracture In order to illustrate the basic parameters of the fatigue
analysis. The basic control factors in fracture mechanics are crack growth, Fig. 3 shows a typical plot of crack propagation
the stress intensity factor K, the energy release rate G, the where it is noticed that:
crack opening displacement COD, and the J-integral J. These • Crack length increases with the increasing number
quantities along with their critical or limiting values are the of loading cycles;
foundation of any fracture analysis [7]. • Crack growth rate most often increases with increasing
crack length;
LEFM CONCEPTS • Crack growth rate increases with increasing stress
level;
Fracture mechanics is used to evaluate the strength of a • Crack becomes longer at an increasingly rapid rate;
structure or component in the presence of a crack or flaw. • Most of the loading cycles involved in the total life of
Its application to fatigue involves the crack growth process, the component are consumed during the early stages
covering the range from a detectable crack or flaw to final of crack extension [10, 13].
fracture. One of the common methods used to analyze
this process is LEFM. The method is used to determine
crack growth in materials under the basic assumption that
material conditions are predominantly linear elastic during
the fatigue process. For crack growth or fracture conditions
that violate this basic assumption, elastic-plastic fracture
mechanics approaches are used to describe the fatigue and
fracture process. It is necessary to define the basic crack
surface displacement modes by which a crack can extend
before using LEFM concepts for the crack growth analysis [11].

MODES OF CRACK EXTENSION

Fig. 2 shows three modes by which a crack can extend.

• Mode I is the opening (tensile) mode where the crack
surfaces move directly apart ;
• Mode II is the sliding (in-plane shearing) mode
where the crack surfaces slide over one another in
the direction perpendicular to the leading edge of Fig. 3 Effect of crack length and stress level on crack propagation rate
the crack;
• Mode III is the tearing (anti-plane shear) mode where In general, as shown in Fig. 4, Fig. 5 and 6 , it has been
the crack surfaces move relative to one another and observed that the fatigue process involves the following stages:
parallel to the leading edge of the crack. 1. Stage I – Initiation
a. Cyclic slip
FATIGUE CRACK GROWTH BASED ON LEFM b. Crack nucleation,
c. Short (micro) crack growth,
The main goal of the application of fracture mechanics in 2. Stage II – Stable crack growth
fatigue analysis is to estimate the total number of cycles for a. Long (macro) crack growth, and
the assumed initial crack ai to reach a final length af. The crack 3. Stage III – unstable crack growth
a. Final fracture.

FATIGUE ANALYSIS OF SHIP

STRUCTURES

Fatigue is responsible for a large amount

of cracks occurring in welded ship structural
details. For many years fatigue - related failure
Fig. 2 Modes of crack extension has become a major concern in the maintenance

POLISH MARITIME RESEARCH, No 2/2016 73

 Fig. 4 Phases of fatigue cracking

Fig. 5 Different scenarios of fatigue crack growth Fig. 6 Fatigue rate curve da/dN versus ∆K

of existing ships and the design of new ships especially after may give the basis for the structural design (steel selection,
the introduction of high tensile steel ( HTS). Numerous cracks scantlings and local details). Furthermore, they can form the
were experienced by relatively new oil carriers constructed basis for efficient inspection programs during fabrication and
of HTS materials. The cracks were discovered when the ships throughout the service life of the structure. [14]
were about 3 to 4 years old without any significant corrosion
or wastage [1, 14] . AREAS SUSCEPTIBLE TO FATIGUE DAMAGE
The main factors affecting increased number of fatigue
cracks in ships may be summarized as follows: Fatigue damages are known to occur more frequently for
1. The trend of reducing the ship scantlings based on some ship types and categories of hull structure elements.
detailed stress analysis and the increased use of HTS, The fatigue life is in particular related to the magnitude of the
has resulted in the increase of the general stress level. dynamic stress level, the corrosiveness of the environment
Note that improved detail design which would cause a and the magnitude of notch and stress concentration factors
reduction of the SCF value for ship structural details, of the structural details, which all vary depending on ship
has minimized the effect of increased stress level [4, type and structure. The importance of possible fatigue
14] . damage is related to the number of potential damage points
2. Tankers operate on trade routes (e.g., TAPS trade route of a considered type for the ship or structure in question and
from California to Alaska) having severe weather for to its consequences. [14]
the majority of their operational time, which causes In ship structures a major fraction of the total number of
very severe loading with respect to fatigue damage. fatigue damages occurs in panel stiffeners on the ship side and
3. The presence of corrosion (general, pitting, and bottom and on the boundaries of ballast and cargo tanks [14] .
grooving) in ballast and cargo tanks resulting in a Fig. 7 and 8 show a possible distribution of fatigue cracks
reduction of the fatigue life of ship structural details in a typical ship structural component. In tankers, cracks
[4] . occur mostly on the side longitudinals at the connections to
Due to the growing number of fatigue cracks a more direct transverse bulkheads or transverse webs. Other ship types
control of fatigue is needed. The aim of the fatigue control is to such as bulk carriers also suffer from fatigue cracks. Where
ensure that all parts of the hull structure subjected to fatigue in some bulk carriers, cracks were commonly found in the
(dynamic) loading have an adequate fatigue life. Calculated hard corners of the lower hopper tanks connecting to the
fatigue lives, calibrated with the relevant fatigue damage data,

74 POLISH MARITIME RESEARCH, No 2/2016

side frames, and the lower stools connecting to the double Also, the following assumptions are used for the analysis of
bottom [1] . the finite element model:
DNV [14] and other classification societies has published • Material is isotropic;
some tables showing the critical areas for different ship types. • Linear fracture mechanics applies;
As an example for tankers Tab. 1 is listed hereunder. • the plastic zone in the crack front is minimal.
Tab.1 Critical details for a tanker ABAQUS built- in features can only be used to analyze a
static crack (i.e. in-site crack) or a propagating
crack through a predefined path. The case of
Structure member Structural detail Load type
crack propagation through a predefined path
Butt joints, deck openings Hull girder bending, stiffen- is practical for the analysis of laminations (i.e.
and attachment to trans- er lateral pressure load and fibre glass reinforced composites). However,
Side, bottom and verse webs, transverse support deformation
deck plating and
in case of fatigue crack propagation through
bulkheads, hopper knuckles
longitudinals and intermediate longitudi-
ship structural details the path of the crack
nal girders is unknown. Hence, because of limitation of
this software a special script is developed with
Bracket toes, girder flange Sea pressure load combined the purpose of analyzing a propagating crack
butt joints, curved girder with cargo or ballast pres-
flanges, knuckle of inner sure load
without having predefined path.
bottom and sloped hopper The simplest way to explain this idea is that a
side and other panel knuck- static crack will be first analyzed and from the
Transverse girder and
stringer structures
les including intersection analysis the crack propagation direction is made
with transverse girder webs. known and then the crack is propagated in this
Single lug
slots for panel stiffeners,
direction. Another static crack will be analyzed
access and lightening holes in the new location until the crack is terminated.

Bracket terminations of Hull girder bending, and MODELLING THE ORIGINAL , NON-
Longitudinal girders butting transverse members bending / deformation CRACKED ELEMENT MODEL
of deck and bottom (girders, stiffeners) of longitudinal girder
structure and considered abutting
member The original model is analyzed by using the
finite element method to indicate the critical
point, i.e. the point where maximum stresses
PROPOSED PROCEDURES APPLICABLE TO occur and where the crack is suspected to start. The modelling
CUT-OUTS technique, meshing and loading of the original model is
discussed below.
A calculation method applied to a cut-out model in a side
connection of a crude oil carrier, is proposed. The model MODELLING THE GEOMETRY
encompasses a longitudinal stiffener passing through a
side transverse. Due to the cut-out introduced in the side The crack analysis can be carried out by using either 2D
transverse this connection is suspected of fatigue failure [15]. or 3D models. For each model either shell or solid elements

Fig. 7 Distribution of ship’s fatigue cracks Fig. 8 Critical structural details

POLISH MARITIME RESEARCH, No 2/2016 75

 Fig. 9 Model dimensions Fig. 10 Model dimensions

can be used. The modelling by using shell and continuum After several trials by using both medial axis and advancing
elements can be applied to analyzing fully through cracks, front meshing algorithms, the advancing front meshing
e.g. a cracked plate. However continuum elements cannot be algorithm was selected for this analysis. The medial axis
used for 3D modelling. The solid modelling has the advantage meshing algorithm is much faster than the advancing front
of analyzing both fully through cracks and partially through meshing algorithm , especially with the repeated re-meshing.
cracks, e.g. penny cracks in shafts [16]. Fig. 9 and 10 show However, due to sensitivity of the mesh around the crack tip,
model dimensions in two views. especially in the initiation phase, the mesh has to follow an
Since most cracks in the ship structures will be fully exact seeding pattern which can only be achieved by using
through ones, hence the using of shell elements will be much the advancing front meshing technique [17].
easier than solid elements with no reduction in the quality
of the analysis. This also helps reducing the programming BOUNDARY CONDITIONS
used in propagating cracks and reduces the time required for
the analysis. The proposed element is that of four nodes, S4 , Some studies were made to show the effect of boundary
which can be degenerated into a triangle to allow the account conditions on the hot-spot stress. The results showed the
of crack tip singularity. The area around the cut-out in the insensitivity of the fixation boundary conditions on the hot-
web frame, with estimated high stress concentration, will spot stress away from the boundary. However, boundary
be meshed by using a finer mesh as shown in Fig. 11 and 12. conditions are based on symmetry, continuity and engineering
judgment. The boundary conditions applied to this analysis
are shown in Fig.13, where, UX, UY and UZ are the translation

Fig. 11 Different mesh sizes Fig. 12 Different mesh sizes

76 POLISH MARITIME RESEARCH, No 2/2016

in X, Y and Z directions , respectively , and ROTX, ROTX and as a function of an effective stress intensity factor:
ROTX are the rotation about X, Y and Z axis , respectively,
= C ( ∆K eff )
da m
[15, 18] .
dN

where the effective stress intensity factor range ∆K eff for

combined mode -I –and- II loadings is expressed by:
1
∆K eff = ( ∆K I4 + 8∆K II4 ) 4

INTRODUCTION OF A NEW CRACK SEGMENT

After the analysis of the initial crack and determining the

value of the crack extension angle θˆ , a new crack extension
segment with the length da is introduced. The da - length is
not an essential value for the analysis, however a smaller da
value will lead to better numerical integration for the fatigue
Fig. 13 FEM model boundary conditions life. And of course the smaller the value the more calculation
processes are needed and hence the time to complete a full
APPLIED LOADS analysis.
One major problem that occurs when selecting a relatively
To analyze the crack propagation by using fracture large length da is that the meshing becomes impossible
mechanics the fatigue loading will be simplified as much as sometimes. This is due to the big difference between the
possible. This could be done by choosing a detail placed as seed values of the lines representing the start of the crack,
close to the ship’s neutral axis as possible to eliminate hull see Fig.14 below.
girder loads. In addition, the cargo tank will be assumed
empty to eliminate internal loads. Hence, for the considered
model of oil tanker’s side connection, the fatigue loads playing
the major role in this analysis, are those resulting from the
varying sea water pressure on the outer shell.

INITIAL CRACK

LOCATION OF INITIAL CRACK

The welded structure in question will be assumed to have

only one initial crack. After the analysis of the non-cracked
element model the location of the maximum stresses (von
Mises stress) is recorded and will represent the location for
the start point of the initial crack.
Fig. 14 Different values of the crack extension da
SIZE OF INITIAL CRACK

As discussed before, the initial crack length can be

estimated by using different methods, and to keep the ANALYSIS OF THE CRACKED ELEMENT MODEL
estimation practical, the crack length is assumed as that of AFTER EXTENSION
the smallest crack detectable by using X-ray NDT method;
this length is equal to 3.81mm. [7] The analysis of the model after each crack extension is
continued and all relevant values are stored for each step of
RESULTS OF THE INITIAL CRACK ANALYSIS the crack extension. The most important parameters are :
a, ∆KI, ∆KII, ∆Keff, θˆ .The event of ∆Keff > KC means that the
After the analysis of the initially cracked element model, crack will continue to propagate under Paris’ law. Hence a
the values of ∆KI and ∆KII are obtained directly by applying new crack segment is introduced and analyzed , as stated
the finite element model. The fatigue crack growth rate has previously in 6.4.
been expressed by Tanaka [20] who used a Paris type equation This analysis continues until the effective stress intensity

POLISH MARITIME RESEARCH, No 2/2016 77

factor ∆Keff reaches its maximum value which is equal to the upon the analysis being undertaken, this approximation may
value of the critical stress intensity factor KC . The event of not be adequate. Finally, the Paris law does not consider the
∆Keff > KC means that the stage II of stable crack growth have effect of stress ratio and it depends upon the used material.
ended and Paris law is no more applicable. The crack size a For steels tested at various stress ratios, a family of straight
at this stage , can be considered the critical crack size, ac. lines parallel to each other is produced. This means that the
As discussed before, after reaching this value the crack value of m is the same for all stress ratios but the value of C
propagation enters stage III where the crack propagates at is specific for a particular stress ratio.
very high speed and will not affect the fatigue life calculation.
Hence, at this stage the analysis can be ended without any FATIGUE LIFE
compromise in the calculation of the fatigue life of the model.
Another important check which must be programmed is to As indicated in Bureau Veritas rules [19] the number of
check that each time a new crack segment is introduced and cycles for the expected ship’s life Nt can be estimated by using
the end of this crack segment still remains within the domain the following equation:
of the model. Unless this quite simple check is programmed
the program may never terminate if the variable load is small
and the model is completely cracked without reaching the
condition of ∆Keff > KC .
where: T is the design life in seconds, L is the length of
FATIGUE LIFE CALCULATION the ship, αo is the sailing factor which takes into account the
time needed for loading/unloading operations, repairs, etc.
The fatigue life calculation can be accomplished in two As a rule, αo may be taken equal to 0.85.
stages. First, the number of fatigue load cycles required Calculating the number of cycles corresponding to a crack
to propagate the crack from the initial size ai to the final length or final failure, knowing the number of cycles by using
size af. This can be done by integrating Paris formula , as crack propagation analysis as stated before, the previous
discussed below. The second step is to estimate the number relation could be used accordingly, to estimate the time (e.g.
of fatigue load cycles that occur each year for a ship. This number of years) corresponding to each stage of the crack
way the calculated number of cycles from the first step can growth.
be translated into number of years .

NUMBER OF CYCLES TO FAILURE

At the end of the calculations and after the value of the

range of effective stress intensity factor ∆Keff exceeds the Finally, by using the assumptions of number of load
value of the critical stress intensity factor KC , the fatigue life, cycles per year, the life of the ship can be calculated. Also
as number of cycles to failure , can be calculated by using the following graphs can be plotted:
numerical integration of Paris law, as discussed in 6.3. • Crack profile;
• ∆Keq, ∆KI and ∆KII versus crack length a;
• ∆Keq, ∆KI and ∆KII versus time;
• Crack length a versus time;

RESULTS

After the successful running of the developed program

In the discussed analysis method the step at which crack shown in the flow chart of Fig. 15, the results of the analysis
propagates , da , is held constant all over the analysis. Also could be presented as follows:
the material constants C and m are kept constant. Hence the A crack will initiate at the scallop of the cut-out in the web
above mentioned integration can be simplified into a very frame due to increased stresses at this area. This crack will
simple summation as follows: start propagating towards the side shell plating due to fatigue
loading. The total life of the model was calculated to be 7.8
years. The illustration of the crack propagation during the
fatigue life of the model is illustrated in the form of graphs
and crack profiles for different stages of the crack propagation.
Tab. 2 shows the summary of the different phases.
The limitation of the Paris law is that it is only capable of
describing data in Stage II (see Fig. 6). If the data exhibits a
threshold (Stage I) or an accelerated growth (Stage III) Paris
law cannot adequately describe these regions. Depending

78 POLISH MARITIME RESEARCH, No 2/2016

 the material.

CRACK LENGTH VERSUS TIME

Fig. 25 shows that the crack starts propagating at

a relatively very low rate. As an example , 25% of the
fatigue life is consumed to propagate the crack for
only 6% of its final length. However the rate increases
rapidly after reaching about 50% of the fatigue life.
Also, it shows that the last 50% of the crack length
is consumed rapidly in less than 7% of the total life.

EQUIVALENT STRESS INTENSITY FACTOR

VERSUS CRACK LENGTH AND TIME

Fig. 26 shows that the equivalent stress intensity

factor ∆Keq increases steadily during the crack
propagation phase while increases rapidly at the final
stages. This shows that the ∆Keq follows Paris law
during stage II of the propagation phase .
In combination with Fig. 26, Fig. 27 shows that
during stage II of the propagation phase , where
the ∆Keq follows Paris law, the ∆Keq increases by the
increasing of the crack size. However, at the final
stages it decreases with the increase of crack size.
This shows that the analysis is only valid for the
propagation period because the developed programs
is only applicable to the area following Paris law and
does not include formulas for stage III of the crack
propagation.

Fig. 15 Flow chart of the developed program

Tab. 2 Summary of crack propagation

Crack length Time

Figure
a ( mm ) %a f T (years) % Life

Fig. 16 3.81 0% 0 0%
Fig. 17 39.81 6% 1.9 25%
Fig.18 99.81 15% 3.9 50% Fig. 16 Initial crack profile
Fig. 19 163.81 25% 5.5 70%
Fig. 20 183.81 28% 5.8 75%
Fig.21 323.81 50% 7.3 93%
Fig. 22 487.81 75% 7.7 98%
Fig. 23 647.81 100% 7.8 100%

CRACK PROFILE AND STRESS FIELD

The stress field shown in Fig. 24 illustrates the distribution

of the stresses around the crack tip. The red area in the region Fig. 17 Crack profile
indicates where the stress is higher than the yield stress of a=39.81 mm=6% af t = 1.9Years = 25% Life

POLISH MARITIME RESEARCH, No 2/2016 79

80 POLISH MARITIME RESEARCH, No 2/2016
 CONCLUSIONS very important because designing a ship that would
not suffer fatigue, i.e. in the case of ∆K ≤ ∆Kth , is highly
1. The fatigue durability of ship structural details is impractical. The lightweight of the ship will be then
achieved not only by careful initial design but also increased exponentially, reducing her deadweight. In
by a proper construction and an effective inspection, addition, the added material will increase the initial
maintenance and repair program. cost of building the ship.
2. FEM is a very cost- effective tool for structural analysis 6. All of the above mentioned side effects could be
which saves a lot compared to physical model testing. avoided by carefully designing the ship structural
However, FEM cannot substitute physical testing since details vulnerable to fatigue damage.
the results from FEA should be verified by using 7. Immediate repair and docking delay the delivery of
physical models. However application of FEM reduces cargo and decrease the profit from this ship.
the number of physical models to be tested hence also 8. The ships that usually encounter fatigue problem
the cost of the analysis. are large tankers and bulk carriers. Hence, in the
3. The developed program is used for the fatigue most extreme cases where the cracks are left without
analysis of a side connection of a tanker. It uses basic inspection or repair, the situation may lead to
programming and features of ABAQUS software leakage of oil or hazardous fluids into the sea causing
package. A more complicated and generic programs environmental problems.
can be derived by applying the same methodology for 9. Finally, the most effective way is to conduct an overall
the analysis of other models. fatigue analysis for any new design of tankers or bulk
4. Crack propagation analysis for ship structural details carriers. As a result of such analysis the schedule of
helps scheduling the inspection programs (locations inspection of critical details can be determined. In
and period) for such parts of the ship, e.g. close-up this case, the high initial cost and loss of deadweight
survey for tankers. can be avoided, moreover the critical parts where
5. Fatigue analysis of the ship connections and details is fatigue cracks may occur are in advance determined

POLISH MARITIME RESEARCH, No 2/2016 81

 before their growing into catastrophic cracks. On the 15. Dassault Systèmes Simulia Corp.: ABAQUS Analysis User’s
contrary, if fatigue analysis was not conducted, this Manual. Dassault Systèmes Simulia Corp., 2008.
may lead to cracks in different high - stressed spots. If
the cracks are not repaired in a satisfactory manner, 16. K. Ma and R. G. Bea: A Repair Management System for
they may lead to damage of ship members Fatigue Cracks in Ships. SNAME Transactions, vol. 103,
pp. 343-369, 1995.
BIBLIOGRAPHY
17. M. R. Andersen: Fatigue crack initiation and growth in ship
1. B. Ayyub, et al. : Risk-informed inspection of marine structures. Ph. D Thesis, Department of Naval Architecture
vessels. Ship Structures Committee, SSC-421, 2002. and Offshore Engineering, Technical University of
Denmark, Lyngby, 1998.
2. R. Bea: Maintenance of marine structures: A state of the
art summary. DTIC Document1993. 18. Dassault Systèmes Simulia Corp.: ABAQUS Theory Manual.
Dassault Systèmes Simulia Corp., 2008.
3. C. C. Capanoglu: Fatigue Technology Assessment and
Strategies for Fatigue Avoidance in Marine Structures. 19. Registro Italiano Navale : Rules for the checking of the
Ship Structure Committee (SSC), 1993. fatigue strength of ship hull structures. 1995.

4. T. Xu : Fatigue of ship structural details – Technical 20. A. Miranda, et al.: Path and life predictions under mixed
Development and Problems. Journal of Ship Research, vol. mode I-Mode II complex loading. Paper presented at the
41, pp. 318-331, 1997. International Symposium on Solid Mechanics, University
of São Paulo, São Paulo, Brazil, 2007.
5. S. Moaveni: Finite element analysis: theory and application
with ANSYS. Upper Saddle River, N.J.: Prentice Hall, 1999.

6. Y.-L. Lee: Fatigue testing and analysis: theory and practice. CONTACT WITH THE AUTHOR
Amsterdam; Boston: Elsevier Butterworth-Heinemann,
2005. Ahmed M H Elhewy

7. J. Schijve: Fatigue of structures and materials. New York: Naval Architecture and Marine Engineering Department
Springer, 2008. Faculty of Engineering
Alexandria University
8. International Association of Classification Societies ltd.
(IACS) : Standard Wave Data . Recommendation 34, 2001. e-mail: ahmed.elhewy@gmail.com

9. Det Norske Veritas : Fatigue Strength Analysis of Offshore egypt

Steel Structures. (DNV) RP-C203 Recommended
Practice-C203, 2001.

10. B. Farahmand, et al., Fatigue and fracture mechanics of

high risk parts: application of LEFM & FMDM theory.
New York: Chapman & Hall, 1997.

11. B. Farahmand : Fracture mechanics of metals, composites,

welds, and bolted joints: application of LEFM, EPFM, and
FMDM theory. Boston: Kluwer Academic Publishers, 2001.

12. S. Beden, et al.: Review of Fatigue Crack Propagation

Models for Metallic Components. European Journal of
Scientific Research, vol. 28, pp. 364-397, 2009.

13. H. A. Rothbart and T. H. Brown: Mechanical design

handbook: measurement, analysis, and control of dynamic
systems. 2nd Ed. New York: McGraw-Hill, 2006.

14. Det Norske Veritas : Fatigue Assessment of Ship Structures.

(DNV) CN 30.7, 2010.

82 POLISH MARITIME RESEARCH, No 2/2016