Anda di halaman 1dari 10

Tribology International xx (xxxx) xxxx–xxxx

Contents lists available at ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

Fretting fatigue on thread root of premium threaded connections



Yosuke Okua, , Masaaki Suginoa, Yoshinori Andob, Taizo Makinoa, Ryosuke Komodac,d,
Daisuke Takazakic, Masanobu Kubotae
a
Steel Research Laboratories, Nippon Steel and Sumitomo Metal Corporation, 1-8 Fusocho, Hyogo 660-0891, Japan
b
Wakayama Works, Nippon Steel and Sumitomo Metal Corporation, 1850 Minato, Wakayama 640-8455, Japan
c
Graduate School of Kyushu University, 744 Motooka, Fukuoka 819-0395, Japan
d
Japan Society for the Promotion of Science, Japan
e
International Institute for Carbon-Neutral Energy Research (WPI-I2CNER), Kyushu University, 744 Motooka, Fukuoka 819-0395, Japan

A R T I C L E I N F O A BS T RAC T

Keywords: Identification of the fatigue failure mode of the premium threaded connection for Oil Country Tubular Goods
Fretting fatigue pipes was conducted via full-scale fatigue tests. A through-wall crack was found at the imperfect thread root of
Threaded connection the male embodiment, but the crack initiation site depended on the stress level. At relatively higher stress
Full-scale fatigue test amplitude region, the crack originated from the thread rounded corner by stress concentration. At relatively
Large relative slip
lower stress amplitude region, the crack originated at the middle of the thread root because of fretting fatigue.
To investigate the fretting fatigue mechanism in the threaded connection, a fundamental fretting fatigue test
was conducted. This test achieved the fretting fatigue failure at the middle of the contact surface under large
gross slip condition.

1. Introduction using a drill pipe, have been developed [2,3]. In the DwC, a drill bit is
attached to the end of the casing pipe to drill a hole. Consequently, the
Oil Country Tubular Goods (OCTG) pipes for the production of casing pipe is embedded simultaneously with the drilling. This
crude oil and natural gas are connected by threaded connections. For technology can contribute in the reduction of the total time as well as
the threaded connection, sufficient static strength and sealability are cost for the well completion. However, the threaded connections
required to achieve the stable oil and gas production. A premium receive a huge amount of cyclic rotary bending load when they go
threaded connection, which is a non-American Petroleum Institute through the well curvature. Therefore, the evaluation of the fatigue
(API) standard [1] product, is used particularly for a well with a severe performance of the premium threaded connections becomes increas-
environment. The premium threaded connection consists of metal to ingly important.
metal seal, torque shoulder and tapered trapezoidal thread portion as Several studies have been previously performed concerning the
indicated in Fig. 1. The male embodiment of a threaded connection is fatigue performance of threaded connections. For instance, full-scale
called PIN, and the female one is called BOX. When the PIN and BOX fatigue tests were conducted to evaluate the fatigue performance [3–8].
are in the make-up state, the PIN thread root and the BOX thread crest In these studies, Stress Amplification Factor (SAF) was evaluated by
are in contact with each other by interference. the ratio between the S–N curve of the threaded connection obtained
In the conventional well drilling as indicated in Fig. 2(a), a hole by the full-scale fatigue test and DNV-B1 curve [9]. The fatigue
where the casing pipe will be embedded is previously drilled by a drill performance of the connection is regarded higher, as SAF value became
pipe. Then, the casing pipe is embedded into the underground hole lower. For example, Ong et al. [3,4] reported that the sufficient fatigue
with a certain rotation. After the embedment, only static loading is performance was achieved for the premium or semi-premium threaded
applied to the threaded connection of the casing pipe. Therefore, only connection which they developed, by considering from the evaluated
static strength is required in the threaded connection. Recently, as SAF obtained by the full-scale fatigue tests. Sches et al. [5] reported the
indicated in Fig. 2(b), new drilling technologies such as Drilling with results of full-scale fatigue tests of the premium threaded connection.
Casing (DwC), which enables a casing pipe to be embedded without The fatigue performance was improved by applying the larger thread


Corresponding author.
E-mail addresses: oku.8fc.yohsuke@jp.nssmc.com (Y. Oku), sugino.77x.masaaki@jp.nssmc.com (M. Sugino), ando.8n9.yoshinori@jp.nssmc.com (Y. Ando),
makino.hb3.taizo@jp.nssmc.com (T. Makino), komoda6744@kyudai.jp (R. Komoda), 2TE16733K@s.kyushu-u.ac.jp (D. Takazaki),
kubota.masanobu.304@m.kyushu-u.ac.jp (M. Kubota).

http://dx.doi.org/10.1016/j.triboint.2016.10.021
Received 17 June 2016; Received in revised form 30 September 2016; Accepted 13 October 2016
Available online xxxx
0301-679X/ © 2016 Elsevier Ltd. All rights reserved.

Please cite this article as: Oku, Y., Tribology International (2016), http://dx.doi.org/10.1016/j.triboint.2016.10.021
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Pipe body 2. Full-scale fatigue test of premium threaded connection


Metal to metal seal
Tapered trapezoidal thread
2.1. Specimen of premium threaded connection and material

Fig. 3 displays the longitudinal-sectional view of the real premium


Contact threaded connections for OCTG pipes used for the full-scale fatigue
Oil, Gas test. The connection consists of imperfect threads, perfect threads, seal
Torque shoulder and torque shoulder from the left side of the figure to the right. For the
PIN pipe, the outside diameter (OD) was 177.8 mm, and the wall
BOX thickness (WT) was 11.5 mm. The threaded portion had 1/16 tapered
(Female embodiment) trapezoidal threads. The thread consists of thread root, thread crest,
PIN load flank and stabbing flank. Imperfect threads had less engagement
(Male embodiment) because the thread crest of the male was truncated due to the tapered
thread. When an axial load is applied to the connection, the engaged
Fig. 1. Schematic of real premium threaded connection.
portion of PIN and BOX sustain the axial load. However, at the
imperfect thread portion, only the PIN has to sustain a large part of
root radius. the axial load due to the end of the engaged portion of the PIN and
Simulation works were also performed to evaluate the stress BOX. As the result, the stress at the imperfect thread portion of the PIN
concentration of the thread root by a finite element analysis and should be increased. Therefore, the critical portion for the fatigue
predict the fatigue life of the threaded connection [10–14]. For strength of the thread connection is the imperfect thread of the PIN.
example, Cetin et al. [12] focused on the thread root radius and Machining tolerance is one of the important factors for threaded
proposed a model based on the principal stress or notch sensitivity connections. For the threaded connection used for the full-scale fatigue
methods by Peterson [15] and Neuber [16] to predict the fatigue life of tests in this study, so-called HH-PNBN configuration was applied. In
the threaded connection. Lin et al. [14] applied a multi-axial fatigue life short, the machining tolerance designated by the HH-PNBN is high for
prediction method based on the critical plane model to calculate the the thread interference (H), high for the seal interference (H), nominal
fatigue life of a standard API drill collar connection. for the thread taper of the PIN (PN) and nominal for the thread taper of
However, these anterior researches are not focused on the clarifica- the BOX (BN). Surface treatment with manganese phosphate was
tion of the fatigue failure mechanism or influential factors for the applied to the BOX thread, whereas the PIN was as machined. It is
fatigue performance intrinsic to threaded connections. Without the noted that the diameter of the PIN at the thread root was a little bit
understanding of the fatigue failure mode, it is difficult to develop the larger than that of the BOX at the thread crest. Therefore, the
effective methodologies for improving the fatigue performance and to engagement between the thread root of the PIN and thread crest of
establish a more accurate predictive model of the fatigue life. the BOX was in interference fit. This means that the thread root of the
In the present study, full-scale fatigue test of the premium threaded PIN and the thread crest of the BOX were in contact with each other
connection for the OCTG pipes was performed. Then, a failure analysis with a certain value of contact pressure. The threaded portions, seal
was performed to identify the fatigue failure modes of the threaded and torque shoulder of the specimens were covered with the lubricant
connection. As the result, it was found that the fretting fatigue failure is grease specified by API. Then, the specimens were made-up by a
one of the major failure modes of the premium threaded connection. hydraulic power tong.
Thus, a fundamental fretting fatigue test using a specimen taken out Three types of API standard materials, L80, P110 and Q125, were
from the OCTG pipes was conducted to clarify the fretting fatigue used for the full-scale fatigue tests. The mechanical properties and the
failure mechanisms of the threaded connection. chemical compositions are indicated in Tables 1 and 2. The chemical
composition is given by the maximum values specified by the API
standard. Fig. 4 displays the microstructures of these materials. The
materials exhibited tempered martensitic microstructure with approxi-
mately 30 µm of the prior austenite grain size.

Fig. 2. Applied stress to the threaded connection in conventional drilling (a) and drilling with casing (b).

2
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

BOX (Coupling)
Pipe body

Pipe body
PIN

1/16 thread taper Metal to metal seal


Outer surface of pipe
BOX

Φ177.8
11.5
PIN
Torque shoulder

Thread root Thread crest Load flank


: Contact

Interference fit Thread pitch =5.08 Clearance Stabbing flank

Imperfect threads Perfect threads


Fig. 3. Configurations of the tested premium threaded connection. Longitudinal cross-sectional view is displayed. Dimensions are given in mm.

[17] as displayed in Fig. 5. This testing machine consisted of four


Table 1 supports, a motor, a loading system by screw jack and a frame. The
Mechanical properties of materials for the threaded connection. specimen, which was initially horizontally installed to the testing
machine, was rotated by the motor, and then a vertical force was
Material 0.2% Proof Tensile strength Elongation [%] Reduction of area
applied to the specimen. In this way, an alternating cyclic bending
stress [MPa] [MPa] [%]
moment was applied to the connection. The rotation speed of the
L80 570 680 28 75 specimen was 83 or 166 rpm, which corresponded to the loading
P110 850 940 18 67 frequency of 1.4 or 2.8 Hz, depending on the applied vertical force
Q125 960 1030 18 71 level. The length of the assembled PINs and BOX for the rotary bending
fatigue test was approximately 3.0 m. The ends of the specimen were
capped and water was filled inside the specimen at 0.6 MPa of internal
Table 2 pressure to detect the failure of the specimen. The failure of the
Chemical composition [mass %]. specimen was detected by the decrease in the water pressure due to the
through-wall fatigue crack. The bending stress of the specimen was
Material C Si Mn P Cr Mo
measured by two strain gauges, which were pasted on the outer surface
L80 0.43 0.45 1.90 0.03 – – of the PIN pipe at a 25 mm distance from the end of the BOX coupling
P110 – – – 0.03 – – with a 180° interval around the circumference of the PIN pipe. The
Q125 0.35 – 1.00 0.02 1.20 0.75 bending moment applied to the specimen was controlled so that the
average of the two strain values agreed with the target one.
Note that maximum values specified by API standard [1] are given.
A resonance fatigue test was performed at the relatively lower stress
amplitude region to reduce the time for the experiment. The loading
frequency was 15–30 Hz. This type of test was applied for the
2.2. Full-scale fatigue test method evaluation of the fatigue strength of full-scale railway axle or wheel
[18] and threaded connections for OCTG pipes [3,4]. Fig. 6 displays the
The full-scale fatigue test of the real premium threaded connection schematic of the resonance fatigue test for threaded connections for
was conducted by two test methods. At the relatively higher stress OCTG pipes. This testing machine consisted of two support rollers, a
amplitude region, a four points rotary bending fatigue test was applied motor, a dead weight and an eccentric mass. The cyclic bending fatigue

L80 P110 Q125

20 μm 20 μm 20 μm
Fig. 4. Microstructures of the tested materials. L80 (a), P110 (b) and Q125 (c).

3
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Φ177.8
Load Pipe (PIN) Load
Motor Bearing

Motor
BOX (Coupling)
Load Load Capped end Strain gauge Support

Bending moment
Connection
1610
2670
Fig. 5. Four points rotary bending fatigue test of the real premium threaded connection. Dimensions are given in mm.

stress was produced by the resonance vibration of the specimen with 500
L80 (TS=680 MPa)
the first mode of vibration induced by rotation of the eccentric mass. 400 P110 (TS=940 MPa)
The maximum bending moment was caused at the center of the BOX Q125 (TS=1030 MPa)

Stress amplitude, σa [MPa]


300
coupling. The length of the assembled PINs and BOX for the resonance
fatigue test was approximately 4.0 m. The failure of the specimen was
200
detected by the reduction in the water pressure in a similar way to the
rotary bending fatigue test, but the initial water pressure was 1.0 MPa.
The bending stress was measured by four strain gauges attached on the
Rotary bending
outer surface of the PIN pipe at a 135.6 mm distance from the end of 100
the BOX at a 90° interval around the circumference of the PIN pipe. 90 Resonance
80
This distance was determined by the result of the previous study as 70 Origin of through-wall crack
3 × OD × WT [19]. The bending stress was controlled by the mea- 60
Double: thread rounded corner
sured strain in a similar way to the rotary bending fatigue test. 50
Open: middle of thread root
Both full-scale fatigue tests were conducted at room temperature. 40 Half: no failure
Stress amplitude, σa, was obtained by multiplying elastic modulus, 30
which was 210 GPa, by the measured strain amplitude. The fatigue test 103 104 105 106 107
was finished when the specimen was unbroken beyond 1.00×107
Number of cycles to failure, Nf
cycles.
Fig. 7. S–N curves of premium threaded connection with three OCTG materials
obtained by the full-scale fatigue test.
2.3. Identification of fatigue failure mode of premium threaded
connection from both upper side and longitudinal section (line A-A′). Fig. 8(b)
displays an example of failure at the short life region, and Fig. 8(c)
Fig. 7 displays the S–N curves of the tested premium threaded displays an example of failure at the long life region. The stress
connections. All the tests were halted by the reduction in the internal amplitude was σa=278.8 MPa and the fatigue life was Nf=7.79×104
pressure because of the through-wall crack. Although the Q125 has the for Fig. 8(b). The stress amplitude was σa=73.4 MPa and the fatigue life
highest ultimate tensile strength, the fatigue life of the Q125 connec- was Nf=3.11×106 for Fig. 8(c). The material was P110. As displayed in
tion was shorter than that of the P110 one at relatively higher stress Fig. 8(b) and (c), the through-wall cracks were found at the rounded
amplitude region. The L80 connection also had the shortest fatigue life. corner of the thread root for the short-life region, and at the middle of
To investigate the fatigue failure mechanism, all the specimens the contact surface for the long-life region. For the crack which
were observed by using an optical microscope after the break-out of the originated from the rounded corner at the thread root, it is presumed
PIN and BOX by a hydraulic power tong. that the cause was stress concentration at the thread rounded corner. A
The results of the observations are displayed in Fig. 8. Fig. 8(a) detailed analysis of the crack at the thread rounded corner has been
displays the position of the through-wall crack in the specimen viewed
Φ177.8

Eccentric mass weight Dead weight


Motor Pipe (PIN) BOX (Coupling)
Sample

Eccentric mass weight


Capped end Strain gauge
Support roller

Bending moment

4080
Fig. 6. Resonance fatigue test of the real premium threaded connection. Dimensions are given in mm.

4
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Imperfect threads Perfect threads

PIN

A A’

Through-wall crack

30 mm
A - A’
BOX
PIN Imperfect threads Perfect threads
Through-wall crack 10 mm
Macroscopic observation of failed connection after break-out
Stress concentration Fretting
Thread crest Thread root
Thread root
Rounded corner
Through-wall
crack Through-wall
crack
0.5 mm 0.5 mm

Short life region Long life region


(σa= 277.8 MPa, Nf= 7.79 104) (σa= 73.4 MPa, Nf= 3.11 106 )
Fig. 8. Typical observation results of broken premium threaded connection made of P110. Macroscopic observation of failed connection after break-out (a), short life region
(σa=278.8 MPa, and Nf=7.79×104) (b), and long life region (σa=73.4 MPa, and Nf=3.11×106) (c).

published in Ref. [17]. On the other hand, for the cracks at the the fretting fatigue crack originating at the boundary between stick and
relatively long-life region, the cause might be fretting fatigue because slip region [22,23]. Another is the fretting fatigue crack that originates
the crack was initiated at the contact surface between the thread root at the middle of severe fretting wear region with relatively large slip
and thread crest. It is noted that fretting fatigue failure occurred in amount because of low contact pressure [24–26]. For the latter case,
both four points rotary bending fatigue test and resonance fatigue test, the mechanism why fretting fatigue cracks are initiated at the middle of
regardless of the difference in loading frequency or loading type. The the contact surface is not fully understood.
rotation speed of the threaded connection in actual DwC is lower than As displayed in Fig. 9, more detailed observation was performed for
50 rpm, which corresponds to the loading frequency of 0.83 Hz [20]. In the specimen with the fretting fatigue failure. Fig. 9(a) displays the
fact, applied loading frequency of both four points rotary bending contact surface on the thread root viewed from the upper side. Fig. 9(b)
fatigue test and resonant fatigue test was higher than that of DwC. displays the longitudinal cross-sectional view for the line B-B′ of the
However, the influence by loading frequency would be little in the contact surface. In Fig. 9(a), the through-wall crack originated at the
present study because there was no big difference in the morphology of middle of the contact surface as described in Fig. 8(c). In addition,
the fretting on the thread root obtained by both full-scale fatigue tests. several micro cracks were found near the through-wall crack at the
Further study is needed to clarify this influence because there seems to contact surface by the observation after buffing of the contact surface.
be few anterior researches regarding the influence of loading frequency In Fig. 9(b), several micro cracks were also found around the through-
on the fretting fatigue [21]. wall crack. Although the angle was different depending on the positions
In general, for the fretting fatigue crack originating at the middle of of the micro cracks, these cracks were also tilted.
the contact surface, it seems that there are two different cases. One is These features of the multiple cracks and tilted cracks are coincided

5
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Thread root Longitudinal section (B - B’)

direction
Slip
Slip direction
Thread Thread
crest crest
Thread root

B B’
View of the failed thread root

Through-wall crack

wall crack
Rounded

Through-
PIN
corner
Thread root 1.0 mm
0.5 mm

Through- Thread root


wall crack Micro crack

Micro crack

Micro crack

0.1 mm 0.1 mm

Fig. 9. Detailed observation of the fretting fatigue failure in the threaded connection made of P110 (σa=73.4 MPa, and Nf=3.11×106), thread root (a), longitudinal section (b).

with the features of the fretting fatigue cracks [27,28]. Many predictive connection. A fundamental fretting fatigue test with a bridge pad was
models are established for the fretting fatigue crack at the contact edge conducted to clarify the mechanism of the fretting fatigue failure on the
[29–32]. Contrary to this situation, further research is required to thread root.
clarify the mechanism for the fretting fatigue crack at the middle of the
contact surface. Therefore, as a first step, a fundamental test using
3.1. Fretting fatigue test method
small specimen was conducted to reproduce the fretting fatigue failure
at the middle of the contact surface of the threaded connection.
Fig. 10 displays the schematic of the fundamental fretting fatigue
test. A bending fretting fatigue test with a bridge pad was conducted for
the specimen taken out from the OCTG pipes. The contact pressure, pc,
3. Fundamental fretting fatigue test using a small specimen
was applied by tightening of the bar springs with the tightening bolts.
According to the previous studies [24–26], the fretting fatigue crack
Two fatigue failure modes were found in the premium threaded
was initiated at the middle of the contact surface with the relatively low
Bending moment contact pressure. Thus, to reproduce the fretting fatigue crack in the

Strain gauge
(nominal stress)
Specimen
Slip sensor
Contact pad

Contact force Contact force

Bar spring

Tightening bolt

Bending moment
Fig. 11. Loading condition for the longitudinal section of threaded connection in the
Fig. 10. Method for fundamental fretting fatigue test. full-scale fatigue test.

6
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Specimen
140
100.5

0.5
39.5

14.5
14.5

12
39.1 61.8 39.1
Thickness = 8
Contact pad
25

0.5
14

3 3
Thickness = 10
Fig. 12. Configurations of specimen (a) and contact pad (b). Dimensions are given in mm.

full-scale fatigue test, which was initiated at the middle of the contact
surface, 25 MPa of contact pressure was applied in the fundamental
fretting fatigue test. In addition, 100 MPa of contact pressure was also
Stress amplitude, σa [MPa]

500
applied in the test to evaluate the effect of contact pressure.
Fig. 11 shows the loading condition for the longitudinal section of σw= 480 MPa
the threaded connection in the full-scale fatigue test. It seems that
longitudinal section of the threaded connection receives tensile and
compressive load, while the whole connection receives the rotary
bending load. During the cyclic rotary bending, relative slip occurs
on the contact surface between PIN thread root and BOX thread crest.
Thus, the PIN thread root corresponds to the specimen in the
fundamental test, and the BOX thread crest corresponds to the foot 100 σw= 134 MPa
Axial fatigue test
of contact pad in the fundamental test. Fretting fatigue test
As displayed in Fig. 10, a strain gauge was pasted at the nominal
part of the specimen. Nominal stress amplitude during the test was 50
obtained by multiplying elastic modulus, which was 210 GPa, by the 105 106 107
measured strain amplitude. To simplify the failure analysis by restrict- Number of cycles to failure, Nf
ing crack initiation sites, a thin polyamide film was inserted into one
side of the contact part. The fretting fatigue failure was caused at the Fig. 13. S–N curves of fretting fatigue test and plain fatigue test for P110.

opposite contact part. The relative slip range was measured by small
slip sensors at both contact edges of one contact pad. The fretting
fatigue test was performed at a loading frequency of 15 Hz with a stress with a #400 emery paper.
ratio of −1 in air at ambient temperature.
Fig. 12 displays the configurations of the specimen and contact pad. 3.2. Results of the fundamental fretting fatigue test of OCTG material
Both specimen and contact pad were taken out from the same OCTG
pipes made of P110 and L80. As displayed in Table 3, the mechanical 3.2.1. Fretting fatigue strength of OCTG material
properties of the material for the fundamental fretting fatigue test are The fretting fatigue S–N curve of the specimen made of the P110
slightly different from that displayed in Table 1 because the production with pc=25 MPa is displayed in Fig. 13. As a reference, the S–N curve
lot of the pipes was different between the full-scale test and the of a smooth specimen obtained by an axial fatigue test is also displayed.
fundamental test. The contact surfaces of both the specimen and The fretting fatigue limit was 134 MPa for pc=25 MPa, while the plain
contact pads were finished by polishing along the relative slip direction fatigue limit was 480 MPa. Although loading type influences the fatigue
limit, the fretting reduced the fatigue limit of OCTG pipe material to
Table 3 less than one-third.
Mechanical properties of materials for the fretting fatigue test.

Material 0.2% Proof Tensile strength Elongation [%] Reduction of area 3.2.2. Observation of the fretted surfaces
stress [MPa] [MPa] [%] After the fretting fatigue test, the contact surfaces of the specimen
were observed by an optical microscope. Fig. 14 (a) and (b) display the
L80 620 760 24 78
P110 850 940 11 73 fretted surfaces at pc=25 MPa with the material of P110 and
pc=100 MPa with the material of L80, respectively. Each contact

7
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Fig. 14. Fretted surface on the specimen in the fundamental fretting fatigue test, pc=25 MPa, with material of P110 (a), and pc=100 MPa, with material of L80 (b).

surface was fully covered with fretting wear particles. Although the surfaces in each condition. The trend of the difference of the fretting
aspect of the fretting wear near the contact edge was different from that wear aspect between the vicinity of the contact edge and the middle of
at the middle of the contact surface, no stick region within the fretted the contact surface was more significant on the fretted surface at
area existed. Therefore, a gross slip occurred between the contact pc=100 MPa than that at pc=25 MPa. The main crack was found at the

8
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

Fig. 15. Relation between the relative slip range and crack initiation site of L80.

middle of the contact surface except in a case that pc=100 MPa, and 4. Conclusion
σa=143 MPa. The fundamental fretting fatigue test was successfully
achieved to reproduce the fretting fatigue crack originating at the In the present study, the full-scale fatigue test of the premium
middle of the contact surface of the threaded connection. Furthermore, threaded connection for the OCTG pipes was conducted to identify the
it was found that the position of the fretting fatigue crack changed fatigue failure mode. Then, the fundamental fatigue test with a small
depending on the contact pressure and stress amplitude. specimen was also conducted to discuss the fretting fatigue mechanism
Unfortunately, the area where fretting wear occurred in the fretting in the threaded connection. The obtained results are as follows.
fatigue test with pc=25 MPa was small compared with the nominal
contact area. Therefore, the effect of the contact pressure cannot be (1) Two fatigue failure modes in the premium threaded connection
correctly evaluated in this study. However, the important point is that were found. In the case that a through-wall crack originated at the
the fretting fatigue crack at the middle of the contact surface is thread rounded corner, the stress concentration at the corner was
reproduced. the root cause of the failure. In the case that a through-wall crack
originated at the middle of the thread root, fretting fatigue was the
3.2.3. Relative slip range on the fretted surfaces cause.
Fig. 15 displays the relative slip range and the crack initiation site (2) The fretting fatigue failure in the threaded connection occurred at
obtained by the fretting fatigue test with the material of L80. The solid the middle of the contact surface when the stress amplitude was
symbol represents the relative slip range measured at the contact edge relatively low.
of pad foot where the main crack was initiated, and open symbol (3) In the fundamental fretting fatigue test, the fretting reduced the
represents the relative slip range measured at the opposite side pad fatigue limit of the OCTG pipe material to less than one-third.
foot. The relative slip range for the fretting fatigue test with pc=25 MPa (4) The reproduction of the fretting fatigue failure found in the
was significantly larger than that with pc=100 MPa. The relative slip threaded connection was successfully achieved in the fundamental
range at pc=100 MPa increased with an increase in the stress ampli- fretting fatigue test.
tude. When considering the relation between the crack position and the (5) It is proposed that the fretting fatigue failure at the middle of the
relative slip range, the crack was initiated at the middle of the contact contact surface that was observed in this study is related to large
surface except for the case that the relative slip range was the relative slip.
minimum. It is implied that the crack position moved depending on
the relative slip range level. This may indicate that the stress conditions Acknowledgements
at the contact surface were changed by the relative slip range. In
addition, with a large slip, fretting wear becomes severe. Therefore, the The present study was performed under collaborative research with
change in the stress condition at the contact surface with the develop- the International Institute for Carbon-Neutral Energy Research (WPI-
ment of fretting wear should be considered. I2CNER) at Kyushu University. The International Institute for Carbon-
When the morphology of the fretting fatigue failure is compared Neutral Energy Research (WPI-I2CNER) is supported by the World
between the full-scale test (Fig. 9) and fundamental test (Fig. 14), it is Premier International Research Center Initiative (WPI), MEXT, Japan.
found that the position of the fretting fatigue crack in the fundamental
fretting fatigue test at larger relative slip region is similar to the fretting References
fatigue crack in the real threaded connection. This means that large
relative slip is one of the keys to understand the mechanism of the [1] American Petroleum Institute. Specification for casing and tubing. API specification
5CT (SPEC 5CT), third edition; 1990
fretting fatigue failure in the threaded connection.
[2] A. Mohammed, C.J. Okeke, I. Abolle-Okoyeagu, Current trends and future
Regarding the relative slip range, the threaded portions have an development in casing drilling, Int J Sci Technol 2 (8) (2012) 567–582.
approximately 100 µm clearance at the stabbing flank between the PIN [3] Ong G, Yamaguchi S, Imai R. Successful casing drilling experience with premium
and BOX threads along axial direction of the pipe (Fig. 3). Thus, large connection for production casing application on a subsea well. In: Proceedings of
the SPE Asia Pacific oil & gas conference and exhibition. 2013
relative slip range would be possible in the contact surface of real [4] Ong G, Ramli MN, Ahmad H, Ando Y, Yamamoto T. Evaluation of fatigue
threaded connection. However, sufficient evidences indicating large performance on semi premium connection for casing drilling application to prevent
relative slip on the thread root in the threaded connection are still not connection fatigue failure. In: Proceedings of the offshore technology conference
Asia. 2016
obtained. Moreover, the effect of lubricant grease on the fretting wear [5] Sches C, Hill JW, Peter T, Massaglia J, Lepine E. Fatigue resistant threaded and
as well as stress condition at the contact surface should be investigated. coupled connectors: the new standard for deep water riser applications. In:
Further experiment is needed to clarify the mechanism that fretting Proceedings of the 26th international conference on OMAE. 2007
[6] Sches C, Shilling R. Design and qualification of fatigue resistant heavy wall
fatigue failure occurred at the middle of the contact surface.

9
Y. Oku et al. Tribology International xx (xxxx) xxxx–xxxx

threaded & coupled premium connectors for drilling and production riser [19] ISO/FDIS 13679:2011. Petroleum and natural gas industries – procedures for
applications in deepwater hpht dry tree systems. In: Proceeedings of the of the testing casing and tubing connections. 2011
ASME 29th international conference on OMAE. 2010 [20] Graves KS, Herrera DC. Casing while drilling utilizing rotary steerable technology
[7] Baily G, Strickler RD, Hannahs D, Kamruzzuman S. Evaluation of a casing drilling in the stag field-offshore Australia. In: Proceedings of the SPE annual technical
connection subjected to fatigue and combined load testing. In: Proceedings of the conference and exhibition. 2013
offshore technology conference. 2006 [21] K. Endo, H. Goto, Initiation and propagation of fretting fatigue cracks, Wear 38
[8] Lu Q, Hannahs D, Wu J, Langford S. Connection performance evaluation for (1976) 311–324.
casing-drilling application. In: Proceedings of the offshore technology conference. [22] R.B. Waterhouse, D.E. Tayler, The initiation of fatigue cracks in a 0.7% carbon steel
2007 by fretting, Wear 17 (1971) 139–147.
[9] DET NORSKE VERITAS. Fatigue design of offshore steel structures. Recommended [23] Nakazawa K, Sumita M, Maruyama N. Effect of contact pressure on fretting fatigue
practice, DNV-RP-C203. 2010 of high strength steel and titanium alloy. ASTM STP; 1159; 1992. pp. 139–147
[10] Collins T, Vaghi F. Analysis of the fatigue resistance of rotary shouldered [24] Kubota M, Odanaka H, Sakae C, Ohkomori Y, Kondo Y. The analysis of fretting
connections. In: Proceedings of the IDAC/SPE drilling conference. 2002 fatigue failure in backup roll and its prevention. ASTM STP; 1425; 2003. pp. 434–
[11] Carcagno GE, Bufalini A, Conde L, Toscano R. Low SCF integral premium 445
connections for use in highly demanding casing and tubing drilling operations. In: [25] K. Nakazawa, N. Maruyama, T. Hanawa, Effect of contact pressure on fretting
Proceedings of the offshore technology conference. 2004 fatigue of austenitic stainless steel, Tribol Int 36 (2003) 79–85.
[12] A. Cetin, G. Härkegård, Fatigue life prediction for large threaded components, [26] J. Dobromirski, I.O. Smith, Metallographic aspects of surface damage, surface
Procedia Eng 2 (2010) 1225–1233. temperature and crack initiation in fretting fatigue, Wear 117 (1987) 347–357.
[13] J.V. Wittenberghea, J.D. Pauwa, P.D. Baetsa, W.D. Waelea, M.A. Wahaba, [27] R.B. Waterhouse, Fretting fatigue, Mater Sci Eng 25 (1976) 201–206.
G.D. Roec, Experimental determination of the fatigue life of modified threaded pipe [28] Y. Mutoh, J.Q. Xu, Fracture mechanics approach to fretting fatigue and problems to
couplings, Procedia Eng 2 (2010) 1849–1858. be solved, Tribol Int 36 (2003) 99–107.
[14] T. Lin, Q. Zhang, Z. Lian, Y. Liu, Y. Zhang, Y. Chen, Multi-axial fatigue life [29] M. Kubota, S. Kataoka, Y. Kondo, Effect of stress relief groove on fretting fatigue
prediction of drill collar thread in gas drilling, Eng Fail Anal 59 (2016) 151–160. strength and index for the selection of optimal groove shape, Int J Fatigue 31
[15] R.E. Peterson, Notch sensitivity, in: G. Sines, J.L. Waisman (Eds.), Metal fatigue, (2009) 439–446.
McGraw-Hill, New York, 1959. [30] T. Hattori, T. Watanabe, Fretting fatigue strength estimation considering the
[16] Neuber H. Theory of notch stresses: principles for exact stress calculation. Raven fretting wear process, Tribol Int 39 (2006) 1100–1105.
FA, Brock JS, editors. Michigan, Ann Arbor. J. W: Edwards; 1946 [31] T. Hattori, V.T. Kien, M. Yamashita, Fretting fatigue life estimations based on
[17] Oku Y, Sugino M, Tsujimura T, Makino T. Life prediction of rotary bending fatigue fretting mechanisms, Tribol Int 44 (2011) 1389–1399.
on premium threaded connection for OCTG. Trans JSME Japan. 2012 [32] D.A. Hills, R.J.H. Paynter, D. Nowell, The effect of wear on nucleation of cracks at
[18] Y. Okagata, K. Kirimiya, T. Kato, Fatigue strength evaluation of the Japanese the edge of an almost complete contact, Wear 268 (2010) 900–904.
railway wheel, Fatigue Fract Eng Mater Struct 30 (2007) 356–371.

10

Anda mungkin juga menyukai