Engineering Failure Analysis 150 (2023) 107304

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

Failure Analysis of Fatigue Failed M20 Class 8.8 Galvanized

Steel Bolt
Maiada S. Abdelrahman a, Waleed Khalifa a, Mahmoud T. Abdu a, b, *
a
Metallurgical Engineering Department, Faculty of Engineering, Cairo University, Giza, Egypt
b
Department of Mechanical Engineering, College of Engineering, University of Bisha, Bisha, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Keywords: An M20 Class 8.8 galvanized steel bolt (derrick bolt) was found to fail after five years of service in
Metallurgical Engineering an offshore drilling station. A comprehensive failure analysis investigation was carried out to
Derrick bolt reveal the root cause of the detected failure. Visual inspection coupled with complete micro­
Steel
structural characterization as well as mechanical testing were performed in the investigation. The
Fasteners
Fractography
results showed that the failure was mainly caused by fatigue failure. The initiation of the fatigue
Fatigue crack was supposedly promoted by the existence of non-metallic inclusions that were present at
Manufacturing defect and maintenance error both the surface and the interior of the microstructure. Furthermore, unbalanced preload forces
Updating maintenance plans resulted in bolt loosening which caused fretting wear of the galvanized layer thickness. The
detected drop in the galvanized layer thickness by 42.20% at the primary origin of the fatigue
crack along with the presence of non-metallic inclusions were thought to be the reason for the
initiation of the fatigue crack. The crack was then found to propagate through a transitional
complete cleavage to quasi-cleavage fracture as appeared from fracture surface topography
studied by scanning electron microscopy (SEM). The final stage of the fatigue failure was found to
be caused by ductile fracture of the overloaded zone.

1. Introduction

Generally, extraction of oil and gas by drilling from oilfields present deep under the seabed requires offshore steel structures. These
structures commonly employ bolted connections for joining steel beams and columns to distribute the loads on the steel members [1].
These bolted connections require the use of high strength steel bolts. The manufacturing process of such high strength steel bolts, for
example grade 8.8, includes pickling, straightening, forging, threading, quenching and tempering, and finally coating. These steels are
medium carbon steels with a carbon content between 0.25 and 0.6 wt% and are modified with boron, chromium, and manganese to
enhance their mechanical properties especially strength and hardenability. The heat treatment of such steels is necessary, where
quenching leads to the formation of the strongest and hardest microstructural constituent (martensite) while the tempering increases
the toughness along with maintaining the high strength [2].
Several metals and alloys were employed for coating the thread of such bolts using the coating processes of zinc plating, galva­
nizing, Xylan coating, and black oxidation (blackening). The offshore structures are exposed to severe corrosive environments and so
they demand special material requirements to ensure longer life of the structure, better safety, and lesser risk of failure. Thus,
galvanized steels are considered as strong candidates for such conditions due to the protective nature of zinc coatings especially in

* Corresponding author.
E-mail address: mahmoud.t.abdu@gmail.com (M.T. Abdu).

https://doi.org/10.1016/j.engfailanal.2023.107304
Received 10 March 2023; Received in revised form 17 April 2023; Accepted 29 April 2023
Available online 12 May 2023
1350-6307/© 2023 Elsevier Ltd. All rights reserved.
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

arctic to tropical environments [3]. The zinc offers threefold protection [4]. The first protection is the dense coating acts as a barrier for
the surrounding corrosive environment from reaching the parent metal. The second protection is that zinc sacrifices itself if the coating
is deteriorated and then the corrosion happens only in zinc. The third protection is the natural weathering process produces passive
layer on the surface that resist rusting.
Laurilliard [5] reported that all coatings affect the physical and chemical performance of the coated bolted connection. For
instance, coating process affects the bolts performance in terms of hydrogen embrittlement which later results in brittle fracture at
conditions of static loading and stress levels below the tensile strength of the steel due to the absorption and retention of hydrogen in
the base metal during the coating process, and thread over-tapping [5]. Most high-strength steel bolts of grade 8.8 contain zinc layer or
black appearance introduced by a coating process that deposits a thin adherent metallic layer. However, internal threads are subjected
to over-tapping by about 0.35–0.53 mm as the bolt thread in case of extra coatings become oversized [6].
Accidents in the offshore oil and gas industry are mainly caused by human factors, environmental conditions, mechanical facilities,
and technical lapses [7]. Although the rate of accidents in the offshore oil and gas industry is not frighteningly superior, material
degradation can lead to costly catastrophic failures with severe impacts on human life and environment [8,9]. The root cause of
material failure can be identified based on the different levels of severity as inoperable, loss of service life, or loss of function [10].
Commonly, failure of bolted connections has two possible failure mechanisms: thread stripping and necking of the threaded portion of
the bolt shank [11]. Thread stripping is the looser degree of fit between the internal and external threads caused by over-tapping which
in turn leads to 20% reduction in the bolt capacity [1]. As a result, the tensile resistance is reduced or even lost leading to failure [12].
Fortunately, the probability of thread stripping can be reduced by increasing the height of the nut [13].
Fatigue failure is considered a worldwide issue for steel structures where it causes 80–90% of the failures in steel structures [14].
Fatigue damage is a critical failure mode for ships and offshore structures due to the cyclic nature of the environmental loads caused by
sea waves, wind, and stress concentration present especially at bolts and holes [15,16]. Fatigue and crack growth in ships and offshore
structures are a highly uncertain and complex phenomenon. This can be related to the fact that the maximum stress at which the
fatigue starts is less than yield strength of the material which leads to difficulty in failure prediction [17]. Design against fatigue
damage, inspection, maintenance, and repair are performed to avoid catastrophic accidents. The fatigue and crack growth limit states
have been a major concern in the design, construction, and operation stages. As bolts and rods are usually the critical parts in fatigue
loaded structures, then a precise estimation of the fatigue strength is essential [18].
Fatigue failure is characterized by certain features which are crack initiation (origin), crack growth (fatigue zone) and fracture
(overloaded zone) [19,20]. The origin (crack initiation) is the place where the fracture starts. Fatigue fracture might have multiple
origins that are formed close to each other and each one develops a crack. These cracks meet with the progression and form a step
called ratchet mark [19]. However, the presence of multiple origins depends on the number of small areas with high stress concen­
trations [9]. The high stress concentration areas can be formed by surface defects caused by manual or machine processing, threading
or due to the formation of slip bands and dislocations by the previous load cycles. Also, inclusions, pits, or pores can promote fatigue
initiation as they act as sites for fatigue crack nucleation which is deleterious especially in high strength steels [21–23]. Inclusions are
formed in most of the commercial materials due to deoxidation additions, impurities or entrained exogenous material [24].
Fatigue initiation usually occurs at surface in case of low cycle fatigue and at subsurface for higher fatigue cycles or at lower stress
levels [25]. Also, the main fraction of fatigue life at low stress conditions is spent in the crack nucleation which is a defect-controlled
process [26]. Thus, the nucleation process depends on the defect size while the rate of the propagation depends on the driving force at
the crack tip and the surrounding environment. Also, fatigue zone includes progression lines known as beach marks [19]. This zone is
considered the major characteristic of the fatigue fracture because it occupies more than 50% of the cross section of the bolt. The crack
propagates during the crack growth stage under the cyclic loads until the remaining material cannot sustain the applied loads anymore
[19,20].
The threaded fastener is subjected to preload force which acts on the thread contact surface and the fastener supporting surface to
prevent the loosening of the treaded fastener and tightens the connection . Unfortunately, this force balance preloaded on fasteners can
be disrupted by external loads caused by service environment characterized with vibrations, impacts, and alternating loads (i.e.,
typical conditions in offshore structures). As a result, relative sliding of the thread surface and relative rotation of the bolt and nut
occurs which in turn causes fretting wear, reduces the matching degree in the bolted joint, decreases the clamping force, loosens the
bolt and finally causes fatigue failure of the bolt [27]. Fatigue fractures are unexpected and difficult to avoid that lead to essential need
for controlling the fatigue properties especially for high-strength bolts along with their mechanical and corrosion properties [28].
On one hand, several factors affect fatigue life of the bolts such as the axial load along the screw stem direction (external load) and
preload force. On the other hand, the fatigue strength of bolts is affected by other factors such as bolt materials, processing methods,
heat treatment procedures, stress peening at round corner of bolt, number of threads left outside for assembly, nut thickness, and stress
concentration [29]. For many years, research on the bolt fractures focused on the fatigue strength investigation due to the fatigue
failure of high strength bolts under high preload force and alternating load [30–32]. Recently, the investigations on bolt failure focused
on the bolt loosening, fatigue failure mechanisms, influencing factors, critical failure load calculation methods and fatigue life pre­
diction methods [33].
In this article we are going to present detailed failure analysis of M20 galvanized high-strength steel bolt working in offshore
drilling station. The failure analysis will focus on comprehensive study for the bolt history service, the possible root causes for failure,
the probable failure mechanisms, and the future recommendations to avoid such accident. The investigation will include exhaustive
material characterizations using destructive mechanical testing, optical microscope (OM), scanning electron microscope (SEM), and
energy dispersive x-ray spectroscopy (EDS).

2
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

2. Experimental Methods

2.1. Material

A failed M20 galvanized high-strength steel bolt of grade 8.8 was supplied. The bolt was applied in the offshore drilling rig as
a”Derrick Bolt”. The failed bolt was one of six M20 galvanized steel bolts (BOLT, HEX, M20 × 100MM) which were used to support the
rig structure in one corner. These bolts were installed and put in service since the erection of the offshore drilling station. However,
knowing the installation method and the amount of torque which was used in installing these bolts is important. Loose or overload
tightening constitutes a substantial factor in the performance of the bolted joint. Overload tightening or the uncontrolled use of lu­
bricants during installation, for example, might lead to failure of bolts [34]. Unfortunately, the installation documents which include
installation procedure and torque were not available for the authors.
Moreover, these bolts were replaced in 2012, and one of them failed in July 2017 during rig operation. The service conditions of the
failed bolt included static loading of the rig structure and dynamic loading during rig operation with no pressure or heat. Both loads
were supported by vertical steel girders fastened using bolts as shown in Fig. 1a. The tightening method was calibrated wrench. The
bolt found sheared/broken at the CMC derrick structure for the AHC supports. Fig. 1b indicates the location of the sheared bolt on the
girder.
The chemical analysis of the failed bolt was identified by optical emission analysis as shown in Table 1. The bolt chemical
composition conforms to the quenched and tempered alloy steel for property classes 8.8, 9.8 or 10.9 according to ISO 898–1/2013
[35].

2.2. Visual Examination

Visual inspection was performed using naked eye and using macroscopic examination to evaluate the fracture surface of the bolt
and the bolt as received conditions.

2.3. Microstructure Characterization

The microstructure characterization of the failed bolt involved qualitative and quantitative analysis of fracture surface topography
and possible crystal defects. The microstructure characterization of the fracture surface was performed using optical microscope,
scanning electron microscope (SEM), and semi-quantitative energy dispersive x-ray spectroscopy (EDS). The surface was prepared for
microstructural examination using griding, polishing, and etching. Nital etchant solution was prepared by mixing 100 ml ethanol and
5 ml nitric acid [36]. The sample was immersed in the solution for 30 s to reveal grain boundaries.

2.4. Mechanical Tests

Two mechanical tests were performed. The tensile test was conducted to investigate the tensile properties of the failed bolt. The
tensile test was conducted according to the guidelines of the standard ISO 6892–1:2019 [37]. The M20x100mm bolt was locally
machined to a diameter of 15 mm near to the head, for about 40 mm long. Fig. 2 shows a typical bolt specimen prepared for tensile test
and the gripping during method. During the test, the bolt head was hanged to the movable (upper) head of the testing machine, while
the threaded part of the bolt was screwed into a female threaded mandrel which was gripped to the lower head of the testing machine.

Fig. 1. Location of failure at offshore drilling rig: (a) arrangement of bolts in the girder, and (b) location of the failed bolt (circled).

3
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

Table 1
The chemical composition of the failed bolt.
Elements C Mn Si P S Cr B

Failed bolt 0.243 0.81 0.10 0.015 0.005 0.81 0.002

Standard material 0.2–0.55 – – 0.025 (max.) 0.025 (max.) 0.3 (min.) 0.003 (max.)

Fig. 2. Bolt tensile testing: (a) specimen after machining, and (b) gripping of specimen during test.

Loading was applied to the bolt until fractured. From the test results, the yield and tensile strengths were directly obtained from the
machine. The reduction in area was obtained from the fractured specimen. Elongation could not be obtained from this bolt since it is
too short to obtain the required gauge length of 75 mm and having at the same time a room for gripping the specimen at both sides.
Thus, the elongation percent could not be determined for this bolt.
The tensile testing machine measures the extension of the specimen from the distance between the upper and lower heads of the
machine during the test. Consequently, the value of the extension includes the extension of the specimen in addition to the elastic
extension of the machine frame. The obtained extension value is, thus, not proper nor accurate for calculating strain for the
stress–strain curve. In addition, it was not possible to determine the elastic properties such as the elastic modulus and the Poisson’s
ratio. Apart from this, the elastic modulus and the Poisson’s ratio are not among the standard mechanical properties that characterize
bolts according to the British Standard (BS) [35]. Thus, no effort was made to determine these values during the testing phase of the
current investigation.
The Vickers hardness test was used to measure the hardness of the failed bolt and the test was conducted according to the re­
quirements of the standard ISO 6507–1:2018 [38], so that readings were measured on a transverse section through the threaded
portion of the bolt, between the bolt axis and the half-radius position. Both the tensile test and hardness test were performed to confirm
the mechanical properties of the failed high-strength steel bolt for grade 8.8 as specified in standard [35 35].

Fig. 3. The visual inspection of the fracture surface of the failed bolt (a) marking on head MRB 8.8 (b) fracture surface at bolt head (c) and (d) the
fracture surface at the thread in the transverse direction (e) bolt failure at first thread from the head and (f) no thread stripping along the longi­
tudinal direction at the contact surface with the nut.

4
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

3. Results and Discussion

A derrick bolt of high-strength steel (grade 8.8) implanted in offshore structure failed during rig operation. A failure analysis was
acquired to identify the root cause of the failure, its mechanism, and recommendation to avoid such accident. In order to do so, visual
inspection, mechanical testing and complete microstructural characterization were performed.

3.1. Visual Inspection

A visual inspection for the fracture surface of the failed bolt has been performed. Fig. 3 shows the fracture of the failed bolt. The
initial inspection of the bolt reveals no sign of corrosion damage. This can be related to the threefold protection offered by zinc that is
introduced by the galvanization process [4]. Hence, the galvanization process of the bolt was proper and successful. Fig. 3(a) dem­
onstrates marks of “MRB 8.8” indicating that the bolt was supplied as high-strength bolt of grade 8.8. Generally, such steel grade is used
successfully for the conditions included into the offshore structures [39]. Hence, the failure due to selection of such grade (grade 8.8)
was not typical.
As mentioned earlier, failure of the bolt has two possible mechanisms which are thread stripping and necking of the thread portion
of the bolt shank. Also, the failure mechanism of thread stripping is less ductile than if the fracture caused by threaded shank of the bolt
(bolt fracture). Fig. 3(b to d) show the morphology of the fracture surface at both the bolt head and thread. Apparently, the fracture
morphology demonstrated little or even no plastic deformation indicating brittle fracture rather than ductile fracture. The failure of the
bolt shows no necking thread of the bolt shank. Concludingly, the possibility of the first failure mechanism of the bolt should be
eliminated as a root cause for the bolt failure.
Fig. 3(e) represents the failure location which occurs at the first thread root of the studs close to the nut from the bolt head in the
transverse direction. This might be related to the high stress gradient introduced by the service loads especially at this area.
On the other hand, Fig. 3(f) shows the contact surface between the bolt thread and the nut thread along longitudinal direction.
Evidently, there is no sign of thread stripping that can reduce the bolt capacity and severely deteriorate its ductility [1,13]. Thus, we
eliminate the probability of failure due to the thread stripping.
Macrostructure examination of the fracture surface in the transverse direction at both sides (the bolt head side and bolt thread side)
was obtained at magnifications less than 30X as shown in Fig. 4. The macrostructure examination is used to have a closer look at the
fracture surface at both sides. Fig. 4(a and b) reveal the fracture surface at the bolt head while Fig. 4(c and d) show the fracture surface
at bolt thread side. Clearly, the fracture at both sides is typically fatigue fracture. As mentioned previously, fatigue failure consists of

Fig. 4. Macrographs of the fracture surface of bolt in the transverse direction at two sides (a) and (b) the fracture morphology at the bolt head side,
and (c) and (d) the fracture morphology at the bolt thread.

5
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

three features namely, origin, fatigue zone, and instantaneous zone [19,20].
Apparently, several origins formed close to each other at different locations leading to formation of several ratchet marks. The
ratchet marks that are located at the root of the first thread as pointed by the red arrows in Fig. 4(a and d). These ratchet marks merge
into one fracture that propagates across the bolt toward the core. Obviously, the ratchet marks are formed on one side of thread root
(not across the bolt circumference) which indicates that this one side is subjected to stress concentration or loading higher than the
other side. Concludingly, the fracture propagated from the side with higher stress concentration toward the lower one.
Fig. 4(b and c) reveal the beach marks through the fatigue zone on both sides of the fracture (head side and thread side). However,
the beach marks are less clear than in the ductile material due to the high strength of the steel bolt.
Fig. 4(b and c) demonstrate the overload fracture in the overloaded zone. Also, the overloaded zone seems to be small compared to
the progression zone. The combined effect of many ratchet marks and small overloaded zone implies that the load was light but there
were high stress concentrations.

3.2. Microstructure Characterization

Complete microstructure characterization was performed to study the fracture surface of the failed bolt, detect microstructural
defects, and identify root cause of the failure. The microstructure characterization includes microstructure examination by optical
microscope, SEM examination, and EDX examination.
The microstructure of the failed bolt was investigated by the optical microscope. Fig. 5(a and b) shows the microstructure of the
failed bolt through two directions, transverse direction (cross section of the bolt) and longitudinal direction (parallel to the bolt axis) at
the center of the bolt. The microstructure is typically that of tempered martensite.
Fig. 5a shows the microstructure along the transverse direction with some ferrite grains that might be related to high-tempering
temperature or prolonged tempering time. Though, fatigue crack growth was reported to be reduced by the increase in the
tempering time and temperature [40,41]. Thus, the high tempering temperature and prolonged tempering should not be considered
the reason for the fatigue crack nucleation and propagation in the transverse direction of the fractured bolt.
Fig. 5b demonstrates the microstructure in longitudinal direction with bands containing more retained austenite with tempered
martensite. This can be related to the micro-segregation of substitutional elements such as Mn and Cr that affects the Ms temperature of
regions enriched or depleted with Mn and Cr. Both Mn and Cr have sluggish diffusion in austenite which results in the formation of Mn/
Cr segregation bands that cannot be eliminated during austenitization [42,43]. During quenching, Mn and Cr lower the Ms temperature
and stabilize the austenite in the regions enriched with them while a high-volume fraction of martensite forms in the depleted regions.
Subsequent tempering, on one hand, turns the formed martensite into tempered martensite which becomes more depleted in carbon.
On the other hand, the regions enriched with Mn and Cr (retained austenite regions) become more enriched in carbon. During cooling
from tempering, part of the retained austenite transforms into fresh martensite because the partitioning of carbon to the surrounding
austenite is insufficient to stabilize all the austenite.
Zhao et al. [44] reported that film-like and rod-like martensite/austenite (M/A) constituents in ferrite/bainite microstructure act as
a barrier for fatigue crack propagation. Also, Zhan et al. [45] investigated the effect of banded structure in high-strength steel on the
fatigue crack growth in both directions longitudinal and transverse. The results revealed that the banded structure in the longitudinal
direction lowered the fatigue crack growth rate compared to that of the transverse direction. Thus, the longitudinal banded structure
had a little effect on arresting the transverse fatigue crack during its propagation.
Fig. 6 reveals the microstructure in the transverse direction at different locations, at the center of the bolt and near the thread root.
The microstructure was investigated with etching (Fig. 6(a and b)) and without etching (Fig. 6c) conditions. The micrographs show
typical tempered martensite microstructure at both the center and near the thread root. They also demonstrate widespread of non-
metallic inclusions, especially oxides and MnS inclusions at both the core and near the thread root. The formation of such in­
clusions, their size and distribution are related to the manufacturing process [46,47]. For example, nodular inclusions such as oxide
inclusions are formed at the austenite grain boundaries during the partial deoxidation of the steel while large-faced inclusions are
formed due to the excessive addition of deoxidizers. The content of such inclusions can be controlled by implementing the proper

Fig. 5. Microstructure of failed bolt showing tempered martensite structure along two directions at the center of the bolt (a) transverse direction
near the fracture surface, and (b) longitudinal direction with micro-segregation bands containing more retained austenite with tempered
martensite (100×).

6
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

Fig. 6. Microstructure at different locations showing wide spreading of non-metallic inclusions (oxide and MnS inclusions) along transverse di­
rection of the failed bolt: (a) at the center of the bolt, (b) near the thread, and (c) near the thread root without etching (100×).

metallurgy technique. Decreasing the contents of harmful elements (e.g., O and S) to less than 10 ppm effectively decreases the sizes of
inclusions even below 1 µm in diameter [48].
Microscopically, these non-metallic inclusions form stress concentration sites that dramatically affect the stress distribution and
become preferred sites for fatigue crack nucleation [49]. On the other hand, the geometry, size, location, and distribution of non-
metallic inclusions significantly affect the fatigue life. Lambrighs et al. [50] reported that fatigue life increases when the inclusion
geometry approaches a perfect circle. As revealed in Fig. 6a most of the center of the bolt inclusions geometry is closer to perfect circle
while those near the thread root as shown in Fig. 6(b and c) had geometry ranging from angular to sharp-edged. Thus, the fatigue life
can be dramatically decreased due to the near thread root inclusions. Lipiński [51] reported that the distribution of non-metallic
inclusions affects the fatigue life where both individual and clustered inclusions increase the fatigue damage. Clearly, Fig. 6 dem­
onstrates both types of distributions, individual and clustered inclusions. Also, the location of the non-metallic inclusions significantly
affects the fatigue life where short fatigue lives are controlled by surface inclusions while long fatigue lives are controlled by interior
inclusions. In our case, there are both interior and near surface inclusions and so the fatigue life can be easily decreased. As for the non-
metallic inclusions size, there is critical size below which the fatigue crack cannot initiate at the inclusion. The reported critical size
was 50–100 µm [52] but fatigue failure due to inclusions just below the surface with size smaller than 45 µm was also reported [53].
The size of the near surface inclusions ranged from 41 to 83 µm (see Fig. 6a) while that at the core ranged from 38 to 59 µm (see Fig. 6(b
and c)). Thus, the inclusions with larger size were located near the surface indicating that the origin of the fatigue fracture initiated at
the thread root.
The fractography of the failed bolt in the transverse direction was further investigated using scanning electron microscope (SEM)
and the energy dispersive X-ray spectroscopy (EDS). Fig. 7 shows the SEM micrographs of the fracture surface and reveals the fatigue
failure stages. Fig. 7(a and b) demonstrate the initiation stage of the fatigue at the primary origin. The circled area in Fig. 7b shows the
outer surface of the thread in the longitudinal direction at the primary origin that reveals clear deterioration with small pits or in­
clusions. Fig. 7c shows the SEM fractography of the progression zone near to the initiation zone. Fig. 7d reveals the transition zone
between the progression zone and overload zone.
The fracture surface in the initiation stage reveals relatively smooth facets around ratchet marks indicating cleavage fracture that is
brittle separation across well-defined habit crystallographic planes [54]. This fracture occurs due to the small intensity factor at the
crack tip [55]. Through this stage the crack progression under the cyclic loading is relatively slow, stable, and obeys the crack
propagation law (steady-state stage).
As the loading cycles increase and the crack propagates more through the material, both the stress concentration factor at the crack
tip and crack propagation rate increase. Hence, the crack propagation depth under each load cycle is higher than that of the steady-
state region (initiation zone) and then the progression zone is verified. However, the fracture through progression zone shows tran­
sition from completely cleavage fracture with cleavage facets near to the initiation zone (see Fig. 7c) and changed to quasi-cleavage
fracture near the overload zone (see Fig. 7d). The quasi-cleavage fracture shows the cleavage facets mixed with little dimples [55].
Also, the striation marks in the progression zone with quasi-cleavage fracture are clear and parallel to the beach marks.
With further crack propagation under cyclic loading, the crack depth reaches a certain value at which fracture toughness of the

7
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

Fig. 7. SEM fractography of the failed bolt showing the fatigue failure stages (a) the initiation at the origin and formation of ratchet marks, (b) the
primary origin near the ratchet marks, (c) the progression zone near to the initiation zone and (d) the transition zone between the progression zone
and overload zone.

material cannot resist the stress concentration at the crack tip. Thus, the crack propagation rate becomes unsteady, and overloading
occurs leading to instantaneous fracture [55]. The fractography of the overloaded zone induced more dimples than the progression
zone indicating ductile fracture (see Fig. 7d). The ductile fracture occurs due to the material sliding along dimples walls under
combined effect of the tensile and shear stresses. Moreover, there won’t be enough material or low fatigue strength at the overload
zone to resist the deformation and then the material fractured with ductile features.
However, the shape of ratchet marks edges can indicate the type of forces that led to the fracture. Tension or bending forces lead to
ratchet marks with perpendicular sides to the fracture surface while tortional forces cause ratchet marks with tapered edges [20]. The
edges of the ratchet marks were tapered and with perpendicular sides that indicates the combination effect of the torsion, bending and
tension forces (see Fig. 7a). Also, the appearance of the fracture can tell one whether the stress concentration is a root cause or not. The
fracture surface with one origin and no ratchet marks means low loading stress and insignificant stress concentration while multiple
ratchet marks around origins indicates low loading stress with high stress concentrations [20]. The fracture surface as shown in Fig. 7a
seems to have multiple ratchet marks around the primary origin revealing that high stress concentrations play an important role in
causing the fracture to initiate.
Fig. 8 shows the presence and distribution of the non-metallic inclusions (dark spots) on the fractured surface. Fig. 8(a and b)
clearly reveals the presence of non-metallic inclusions near the primary origin along with pits and around the ratchet marks indicating
the presence of high stress concentrations that might lead to the initiation of the fatigue fracture. Fig. 8c demonstrates the intense
presence of non-metallic inclusions in the progression zone just near the initiation zone (i.e., near the primary origin). Hence, the high
stress concentrations induced by non-metallic inclusions can strongly promote the crack initiation and propagation across the bolt
material.
Further investigation for the element’s distribution at the initiation zone especially near the primary origin is introduced by the
EDX elemental mapping. Fig. 9 shows the EDX mapping for the zinc, chromium, carbon, and iron. The distribution of chromium,

8
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

Fig. 8. SEM of non-metallic inclusions (Mn (Cr) sulfide (dark spots) in (a) and (b) the initiation zone and (c) progression zone.

carbon, and iron seems to be normal. The distribution of chromium and carbon is uniform through the analyzed area while the iron
distribution shows depletion through the galvanized layer (see Fig. 9(c-e)) and enrichment in the bolt body. However, element of zinc
shows varied distribution through the analyzed area which indicates the change in the thickness of the galvanized layer (see Fig. 9b).
The thickness of galvanized layer subjected to deteriorations at the primary origin where the thickness dropped from 37.2 to 21.5 µm
(i.e., the thickness dropped by 42.20%). Thus, the protection of the galvanized layer was not optimum at the primary origin that might
have promoted the presence of pits and the initiation and progression of the crack. The reason for the drop in the galvanized layer
thickness could probably result from two possibilities. The first possibility might be a manufacturing defect during the bolt coating
resulting in uneven thickness of the galvanization layer at the crack origin. The second possibility might have resulted from unbalanced
preload forces that caused bolt loosening leading to fretting wear of the galvanized layer at the crack origin. Bolts in dynamic services
such as offshore structures should be tightened or pretensioned to 90% of their yield strengths. This reduces the cyclic load range
experienced by bolts and greatly improves their fatigue life [56]. Loose tightening, on the other hand, results in increasing the cyclic
load range experienced by bolt and hence lower the fatigue resistance and life of bolt. Not only this, but this leads to a decrease of
contact pressure between the joining surfaces, leading to possible sliding between surfaces, and changing the way the joint works. The
bolted joint needs, therefore, to be inspected regularly, so that the tightening torque is maintained during the service life of bolts.
Concludingly, at 42.20% reduction in the thickness of the galvanized layer along with the presence of the non-metallic inclusions
especially near the thread root, the possibility of the fatigue crack initiation was more pronounced.

3.3. Mechanical Properties

The mechanical properties of the failed bolt were investigated to ensure that the failed bolt had the satisfied properties to carry out
the applied loads. Both tensile test and Vickers hardness test were conducted on the bolt material (see Table 2). The hardness mea­
surements were collected in the transverse direction through the fractured surface of the bolt. Both the tensile and hardness mea­
surements were compared to the standard values for property classes 8.8, 9.8, and 10.9 according to the EN ISO 898-1-2013 [35] . The
tensile properties of the failed bolt satisfy the requirements for all the classes. The hardness measurements of the failed bolt were found
inconsistent to class 8.8 but satisfy that of the class 10.9 although the failed bolt was supplied as class 8.8. Thus, the bolt material

9
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

Fig. 9. EDX mapping at the primary origin in the initiation zone. The SEM micrograph shown at (a) demonstrates the area analyzed with EDX
elemental mapping for (b) Zinc element, (c) Chromium element, (d) Carbon element, and (e) Iron element. EDX spectrum with major peaks is shown
in (f).

Table 2
The mechanical properties of the failed bolt along with that of the property classes 8.8, 9.8, and 10.9 according to [35].
Property 0.2% offset yield strength, MPa Tensile strength, MPa Reduction in area, % Hardness, HV

Bolt 947 ± 10 1103 ± 10 69 ± 5 352 ± 4

(d ¼ 20 mm)
Class 8.8 660 (min.) 830 (min.) 52 (min.) 255–335
(d > 16 mm)
Class 9.8 720 (min.) 900 (min.) 48 (min.) ————
Class 10.9 900 (min.) 1000 (min.) 48 (min.) 320–380

satisfies the requirements for class 10.9 not class 8.8. The bolt material retains high tensile strength along with high reduction in area
that was predictable due to the nature of the microstructure as tempered martensite. On the other hand, the hardness measured from
the bolt material was higher than that accepted for class 8.8. This might be related to the qualitatively medium content of non-metallic
inclusions. These non-metallic inclusions play an important role as stress raisers which in turn increases the hardness and promotes the

10
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

crack initiation and propagation.

It might be interesting if the stress fields could be calculated for the location of fatigue crack initiation, which were the first and
second thread roots after the bolt head. Unfortunately, a major part of variables that are necessary for calculating the stress fields are
not available. However, it was pointed out earlier that a significant high stress concentration was found at the first thread root of the
studs close to the nut [57]. Our findings go well with these results.

4. Conclusions and Preventive Actions

A detailed failure analysis of M20 galvanized high-strength steel bolt constructed in offshore drilling station was conducted. The
investigation found that the main mechanism of bolt failure is fatigue cracking. The initiation of the fatigue crack was supposedly
promoted by the existence of non-metallic inclusions that were present at both the bolt center and near the thread root microstructures.
Moreover, unbalanced preload forces resulted in bolt loosening which caused fretting wear of the galvanized layer thickness. The
detected drop in the galvanized layer thickness by 42.20% at the primary origin of the fatigue crack along with the presence of non-
metallic inclusions were thought to be the reason for the initiation of the fatigue crack. The crack was then found to propagate through
a transitional complete cleavage to quasi-cleavage fracture as appeared from fracture surface topography studied by scanning electron
microscopy (SEM). The final stage of the fatigue failure was found to be caused by ductile fracture of the overloaded zone.
Finally, it is recommended to impose more control during the stages of secondary steelmaking which involves deoxidation,
desulphurization, vacuum degassing, and argon stirring or during subsequent casting. Such control will significantly control the ex­
istence of non-metallic inclusions and reduce their risk. Also, more sophisticated calculation for the preloads and external loads are
required to ensure the accurate balance and reduce the risk of bolt loosening.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

Data will be made available on request.

References

[1] Y. Hu, L. Shen, S. Nie, B. Yang, W. Sha, FE simulation and experimental tests of high-strength structural bolts under tension, J. Cons. St. Res. 126 (2016)
174–186.
[2] G. Krauss, Martensite in steel: strength and structure, Mater. Sci. Eng. A273–275 (1999) 40–57.
[3] A. Kalendová, Effects of particle sizes and shapes of zinc metal on the properties of anticorrosive coatings, Prog. Org. Coat. 46 (2003) 324–332.
[4] P. Sahoo, S.K. Das, J. Paulo Davim, 3.3 Surface Finish Coatings. Comprehensive Materials Finishing, vol. 3, 2017, pp. 38–55.
[5] J. Laurilliard, Fastener coatings, in: J.H. Bickford, S. Nassar (Eds.), Handbook of Bolts and Bolted Joints, Marcel Dekker, New York, 1998, pp. 75–106.
[6] W. Wallace, Bolt coating thickness & nut over-tapping – a lesson in practical necessity@ www.appliedbolting.com/pdf/Coated_Bolts_and_Overtapping.pdf 2015.
[7] ISO, Iso 965 General Purpose Metric Screw Threads – Tolerances – Part 5: Limits of Sizes for Internal Screw Threads to Mate with Hot-Dip Galvanized External
Screw Threads with Maximum Size of Tolerance Position H Before Galvanizing, International Organization for Standardization, Switzerland, 1998.
[8] U.S. Bureau of Labor Statistics. Employer-reported workplace injuries and illnesses–2015. Report No. USDL-16-2056, (Washington, D.C., 2016).
[9] U.S. Chemical Safety and Hazard Investigation Board. Investigation report volume 2-Explosion and fire at the Macondo well. Report No. 2010-10-I-OS,
Washington, D.C., 2014.
[10] ASM Handbook, Failure Analysis and Prevention, vol.11, ASM International, Materials Park, OH, 2002.
[11] Y.C. Wang, J.B. Davison, I.W. Burgess, H.X. Yu, R.J. Plank, C.G. Bailey, The safety of common steel beam/column connections in fire, Struct. Eng. 88 (21) (2010)
26–35.
[12] Y. Hu, J.B. Davison, I.W. Burgess, R.J. Plank, Fire perfromance of grade 8.8 structural bolts, Eurosteel 2011, Budapest, Hungary, 2011.
[13] E.L. Grimsmo, A. Aalberg, M. Langseth, A.H. Clausen, Failure modes of bolt and nut assemblies under tensile loading, J. Constr. Steel Res. 126 (2016) 15–25.
[14] Y. Ni, X.W. Ye, J.M. Ko, Monitoring-based fatigue reliability assessment of steel bridges: analytical model and application, J. Struct. Eng. 136 (2010) 1563–1573.
[15] W. Lin, T. Yoda, Chapter Fourteen - Repair, strengthening, and replacement, in: Bridge Engineering, 2017, pp. 245–271.
[16] R. Brandi, P. Rossetto, Fatigue design of offshore structures, Weld. Int. 1 (12) (1987) 1155–1161.
[17] G. Antaki, R. Gilada, Chapter 2 - Design Basis Loads and Qualification, Nuclear Power Plant Safety and Mechanical Integrity, 2015, pp. 27–102.
[18] H. Bartsch, M. Feldmann, Reassessment of fatigue detail categories of bolts and rods according to EC 3-1- 9, J. Constr. Steel Res. 180 (2021), 106588.
[19] J.M. Djoković, R.R. Nikolić, J. Bujnák, B. Hadzima, Estimate of the steel bridges fatigue life by application of the fracture mechanics, in: IOP Conference Series:
Materials Science and Engineering, vol. 419, 2018, p. 012010.
[20] N.W. Sachs, Understanding the surface features of fatigue fractures: How they describe the failure cause and the failure history, J. Failure Anal. Prevention 5 (2)
(2005) 11–15.
[21] Q.Y. Wang, C. Bathias, N. Kawagoishi, Q. Chen, Effect of inclusion on subsurface crack initiation and gigacycle fatigue strength, 24 (12) (2002) 1269–1274.
[22] S. Nishida, Failure Analysis in Engineering Applications, Butterworth Heinemann, Oxford, UK, 1992.
[23] F. Casanova, J.J. Coronado, Investigació n de la falla de un eje piñó n conectado a una extrusora de doble tornillo, DYNA 77 (2010) 88–97.
[24] C. Kaynak, A. Ankara, T.J. Baker, Initiation and early growth of short fatigue cracks at inclusions, Mater. Sci. Technol. I12 (1996) 421.
[25] C. Bathias, Relation Between Endurance Limits and Thresholds in the Field of Gigacycle Fatigue, American Society for Testing and Materials, West
Conshohocken, PA, 2000, STP 1372, p. 135.
[26] I. Serrano-Munoz, J.-Y. Buffiere, R. Mokso, C. Verdu, Y. Nadot, Location, location & size: defects close to surfaces dominate fatigue crack initiation, Sci. Rep. 7
(2017) 45239.
[27] J. Liu, H. Ouyang, J. Peng, C. Zhang, P. Zhou, L. Ma, M. Zhu, Experimental and numerical studies of bolted joints subjected to axial excitation, Wear 346–347
(2016) 66–77.
[28] Z.L. Zeng, X.J. Feng, Research and discussion on fatigue performance evaluation method of high strength bolts, Internal Combustion Engine Parts 16 (2019)
35–36.

11
M.S. Abdelrahman et al. Engineering Failure Analysis 150 (2023) 107304

[29] X.K. Zhai, Discussion and research of the calculation method for the fatigue life of the bolt, Manuf. Automation 38 (01) (2016) 89–91.
[30] N. Liu, X.D. Cui, J. Xiao, J. Lua, N. Phan, A simplified continuum damage mechanics based modeling strategy for cumulative fatigue damage assessment of
metallic bolted joints, Int. J. Fatigue 131 (2020), 105302.
[31] J. Wang, B. Uy, D.X. Li, Y.C. Song, Fatigue behaviour of stainless-steel bolts in tension and shear under constant-amplitude loading, Int. J. Fatigue 133 (2020),
105401.
[32] S.M. Hosseini, M.S. Mamun, O. Mirza, F. Mashiri, Behaviour of blind bolt shear connectors subjected to static and fatigue loading, Eng. Struct. 214 (2020),
110584.
[33] L. Yang, B. Yang, G. Yang, Y. Xu, S. Xiao, S. Jiang, J. Chen, Analysis of competitive failure life of bolt loosening and fatigue, Eng. Failure Anal. 129 (2021),
105697.
[34] M.S. Mohamed, W. Khalifa, M.G. Mahmoud, Failure analysis of alloy boron steel bolts in steel structure assembly, J. Failure Anal. Prevention 23 (1) (2023)
88–98.
[35] EN ISO 898-1:2013, Mechanical properties of fasteners made of carbon steel and alloy steel Part 1: Bolts, screws and studs with specified property classes -
Coarse thread and fine pitch thread.
[36] D.A. Scott, R. Schwab, Metallography in Archaeology and Art, Cultural Heritage Science, Springer Nature Switzerland AG, 2019.
[37] EN ISO 6892-1:2019, Metallic materials — Tensile testing — Part 1: Method of test at room temperature.
[38] EN ISO 6507-1:2018, Metallic materials — Vickers hardness test — Part 1: Test method.
[39] SCI and BCSA, Joints in Steel Construction - Simple Connections. 2002.
[40] M. Liang, Y. Zhao, M. Zheng, X. Mao, Effect of tempering time on fatigue crack growth behavior of CLAM steel, J. Nucl. Mater. 510 (2018) 437–445.
[41] S.S. Samant, I.V. Singh, R.N. Singh, Effect of tempering and rolling on fatigue crack growth behavior of modified 9Cr-1Mo steel, J. Mater. Eng. Performance 27
(2018) 5898–5912.
[42] F. Forouzan, L. Borasi, E. Vuorinen, F. Mücklich, Optimization of quenching temperature to minimize the micro segregation induced banding phenomena in
quenching and partitioning (Q&P) steels, Steel Res. Int. (2018), 1800281.
[43] S. Yan, X. Liu, T. Liang, J. Chen, Y. Zhao, Effect of micro-alloying elements on microstructure and mechanical properties in C-Mn-Si quenching and partitioning
(Q&P) steels, Steel Res. Int. (2018), 1800257.
[44] Z.-P. Zhao, G.-Y. Qiao, G.-P. Li, W.-W. Yang, B. Liao, F.-R. Xiao, Fatigue properties of ferrite/bainite dual-phase X80 pipeline steel welded joints, Sci. Technol.
Weld. Joining 22 (3) (2016) 217–226.
[45] N. Zhan, Z. Hu, X. Zhang, Experimental Investigation of Fatigue Crack Growth Behavior in Banded Structure of Pipeline Steel, Metals 10 (2020) 1193.
[46] T. Lipiński, A. Wach, Dimensional structure of non-metallic inclusions in high-grade medium carbon steel melted in an electric furnace and subjected to
desulfurization, Solid State Phenom. 223 (2015) 46–53.
[47] T. Lipiński, A. Wach, Influence of outside furnace treatment on purity medium carbon steel, in: Proceedings of the METAL 2014—23rd International Conference
on Metallurgy and Materials, Brno, Czech Republic, 21–23 May 2014, pp. 738–743.
[48] Z.G. Yang, S.X. Li, J.M. Zhang, J.F. Zhang, G.Y. Li, Z.B. Li, W.J. Hui, Y.Q. Weng, Acta Mater. 52 (2004) 5235–5241.
[49] U. Zerbst, M. Madia, C. Klinger, D. Bettge, Y. Murakami, Defects as a root cause of fatigue failure of metallic components. I: Basic aspects, Eng. Failure Anal. 97
(2019) 777–792.
[50] K. Lambrighs, I. Verpoest, B. Verlinden, M. Wevers, Influence of non-metallic inclusions on the fatigue properties of heavily cold drawn steel wires, Procedia
Eng. 2 (1) (2010) 173–181.
[51] T. Lipiński, The effect of the diameter and spacing between impurities on the fatigue strength coefficient of structural steel, Arch. Metall. Mater. 63 (1) (2018)
519–524.
[52] R. Kiessling, Non-metallic Inclusions and their Effects on the Properties of Ferrous Alloys, in: Encyclopedia of Materials: Science and Technology, 2001, pp.
6278–6283.
[53] P. Juvonen, Effects of Non-metallic Inclusions on Fatigue Properties of Calcium Treated Steels, Helsinki University of Technology, Espoo, Finland, 2004.
[54] J.H. Chen, R. Cao, Introduction, in: Micromechanism of Cleavage Fracture of Metals, 2015, pp. 1–54.
[55] Y. Wang, W. Zhang, X. Pan, Y. Zheng, Experimental study on fatigue crack propagation of high-strength steel wire with initial defects for bridge cables, Appl.
Sci. 10 (2020) 4065.
[56] R.W. Hertzberg, Deformation and Fracture Mechanics or Engineering Materials, fourth ed., John Wiley & Sons, 1995, pp. 582–586.
[57] M. Fonte, L. Reis, V. Infante, M. Freitas, Failure analysis of cylinder head studs of a four-stroke marine diesel engine, Eng. Failure Anal. 101 (2019) 298–308.

12