- Email: [email protected]
Effect of stress distribution on the tool joint failure of internal and external upset drill pipes
Materials and Design 52 (2013) 308–314
Contents lists available at SciVerse ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/matdes
Technical Report
Effect of stress distribution on the tool joint failure of internal and external upset drill pipes Shuai Luo, Sujun Wu ⇑ School of Materials Science and Engineering, Beihang University, Beijing 100191, PR China
a r t i c l e
i n f o
Article history: Received 3 December 2012 Accepted 24 May 2013 Available online 7 June 2013
a b s t r a c t Detailed investigation was carried out on the tool joint failure of the internal & external upset (IEU) GSS105 drill pipes. Stresses were analyzed using 2D finite element method (FEM) for the threaded tool joint of the drill pipe under the combined loading conditions containing preload, tensile and bending loads. The stress concentration factors in the pin and box were calculated. Results showed that the maximum stress concentration occurred at the roots of the first tooth from the pin tool joint shoulder of the drill pipe. Fractographic observation revealed that the tool joint failure of the drill pipe was caused by fatigue crack nucleated at the tooth root and propagated through the wall of the tool joint. The deterioration of the fatigue resistance of the tool joint is related to dogleg region where severe cyclic bending load exists due to the local deviation of the drill pipe from the vertical line. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Tool joints, including pin and box, are important parts joining drill pipes serially to make a drill string, which is the key tool in oil wells to transfer the applied torque to the bit at the bottom of the well. The failure of the drill pipes occurred mostly in the tool joint due to continuous changeable load of tension, bending, impact, internal pressure and a certain amount of torque [1–5]. Therefore, it is necessary and urgent to study the mechanism of fatigue failure in tool joint of the drill pipes. Threads at both ends of the tool joints play the vital role in connection. When the drill string is in service, considerable stresses are expected in the teeth of tool joint. Tafreshi and Dover [6] employed finite element method (FEM) to analyze the stress distribution around the threaded connections of tool joints, and presented the load distribution of the teeth of tool joint and stress concentration factors (SCF) at the teeth root. In their work, it was concluded that the location of maximum stress concentration at the tool joint was present. Tafreshi [7] and Bahai [8] analyzed the stresses at the drill pipe connections by using an axisymmetric 2D FEM model, which pointed out that the maximum SCF always located at the first engaged tooth of the pin. Stress concentration at the tooth root is inevitable, and should be within the design safety factor [9]. Therefore, the static stresses at the tooth root alone cannot cause fracture. Previous researches [10–12] on the failure of the drill pipes showed that the fatigue damage of drill pipes always occurs in the pipe section at the dogleg region. However, clear explanation ⇑ Corresponding author. Tel.: +86 1082316326; fax: +86 1082317108. E-mail address: [email protected] (S. Wu). 0261-3069/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.matdes.2013.05.073
is required for the reasons why the dogleg accelerates fatigue failure. In the present study, in order to understand the failure mechanism of the tool joints of the drill pipes, FEM analysis was carried out by employing Plane42 elements and linear contact elements at each tooth. Local stress distribution at each tooth of the threads and variations of SCF at the teeth roots of the pin and box were investigated under same work conditions. The effect of the alternating bending stress caused by dogleg area or bending section of the oil well on the fatigue failure was discussed.
2. Drill string configuration and axial force calculation Eight drill pipes fractured due to cracks at the tooth root of the tool joint during normal drilling in one oil well with a service life much shorter (5–15 h) than the design life (720 h). Therefore, it is of great significance to analyze the failure causes of the drill pipes. The detailed parameters of drill pipe failures are listed in Table 1. All the failed drill pipes are 5–1/200 IEU drill pipes with similar failure mode. In the present study, the failed drill pipe with the shortest service life of 302 min before failure, Case No. 3, was taken as the failure analysis specimen for consideration. The drilling direction of drill pipe always changes and deviates from the vertical direction due to the geologic characteristics and manufacture related factors. The projection drawing of the oil well, including the drilling depth, overall variation of the drill string inclination angle and the dogleg severity, is shown in Fig. 1. It was found that the depth of failure position on the drill pipe concerned is 910 m underground, where the actual maximum overall 3D angle change rate (dogleg severity) is 0.53°/25 m within the first 1000 m depth from the ground.
309
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314 Table 1 Parameters of the drill pipe failures. Failure case (No.)
Well total depth (m)
Depth of failure position (m)
Service life (Min)
Angle of inclination (b/deg)
Dogleg severity (deg/25 m)
1 2 3 4 5 6 7 8
5596.11 5598.81 6362.69 6576.20 6588.63 6604.00 6631.48 6631.63
930 930 910 1400 1350 2000 2350 2300
788 790 302 394 409 412 426 427
0.29 0.29 0.29 0.27 0.29 0.26 0.24 0.36
0.05 0.05 0.53 0.17 0.45 0.33 0.34 0.63
Fig. 1. The projection drawing of the oil well (a) vertical drawing and (b) horizontal view.
The depth of the oil well is 6362.69 m for Case No. 3. The drilling string of the well is divided into four segments as illustrated in Fig. 2 and the detailed parameters of each segment are listed in Table 2. The failure position of the drill pipe is in the first segment which is the IEU GSS105 drill pipe with a diameter of 139 mm. The axial force of the drill pipe at failure position which is required for the FEM stress analysis can be divided into two parts. One is from the gravity of the drilling string and another is from the torque. The force from the gravity can be calculated based on the following formula [13]:
Fa ¼
n X DHk qkm Ha ðAa0 c0 Aa1 c1 Þ
ð1Þ
K¼1
where qkm ¼ qK ðAK0 c0 AK1 c1 Þ, Fa is the axial force at failure position under wellhead, n is the segment of drilling string under failure position, DH is the length of segment, Ha is the distance from ground to failure position, Aa0 is the outer circle area of the failed segment, Aa1 is the inner circle area of the failed segment, AK0 is the outer circle area of the kth segment, AK1 is the inner circle area of the kth segment, c0 is pipe mud weight rate, c1 is tube mud weight rate, qk is gravity unit length pipe in air, qkm is gravity unit length pipe in mud. Based on the above theoretical calculation, an axial force from gravity was obtained as 1.38 106 N.The axial force resulting from the applied torque can be calculated based on the following formula:
Fig. 2. The sketch of the drilling string.
310
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Table 2 Parameters of the drilling string. Segment k
Inner diameter (m)
Outer diameter (m)
H (m)
q (N/m)
c0 (N/m3)
c1 (N/m3)
First Second Third Fourth
0.1186 0.108 0.0702 0.0508
0.1397 0.127 0.0889 0.1206
2659.22 2543.28 960.35 200.54
397.684 261.758 179.732 681.002
11.564e+3
11.466e+3
Table 3 Parameters of the drilling pipe at the failure position. Torque
Preloading torque
Hole curvature (Cw)
Angle of inclination (b)
Moment of inertia (I)
Rotary speed
2537 N m
32276 N m
0.0215 rad/m
0.29°
9.09 105 m4
60 rpm
Fig. 3a. Macroscopic images of (a) cracked drill pipe. Fig. 3d. Macroscopic images of (d) crack on inner surface of the tool joint of the drill pipe.
T¼
FT d2 tanðw qv Þ 2
ð2Þ
Fig. 3b. Macroscopic images of (b) numbering order of the pin tool joint.
where T is the applied torque, FT is the axial force resulting from the applied torque, d2 is the effective diameter of thread, w is the lead angle and qv is the equivalent friction angle. According to the detailed drilling parameters at failure position in Table 3, an axial force from applied torque can be calculated as 1.53 106 N.
Fig. 3c. Macroscopic images of (c) crack at the first tooth root of the tool joint.
Fig. 4a. The images of the fracture surface. (a) The crack initiation sites at the threaded root.
311
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
The crack was nucleated at the first tooth root from the tool joint shoulder of the pin (Fig. 3c) and propagated along the thread root to the inner surface of the drill pipe, as shown in Fig. 3d which is the piece cut from Fig. 3c along the two white lines. 3.2. Fractographic examination Scanning electron microscope (SEM) was applied to observe the fracture surface of the cracked tool joint. Figs. 4a and 4b reveals that multiple fatigue cracks were initiated along the tooth root of the tool joint and propagated towards the inner surface of the drill pipe which implies that the tooth root area was subjected to highly concentrated stress. The initial cracks joined together after some amount of inward extension to form a major crack. Further observation shows that striations are present on the fracture surface (Fig. 4c) which confirms that the failure was caused by fatigue. Fig. 4b. The images of the fracture surface. (b) The magnification of the crack initiation sites.
4. Finite element modeling In the previous study, 2D finite element modeling has been used to calculate the stress distribution under axial and bending loads for the drill string threaded connections [6–8]. Based on the experimental verification, it was reported that the stress analysis of the experiment study were the same as the FEM calculation. Moreover, 3D FEM analysis has been used to investigate the loads distribution under the compressive loading, tensile loading and preloading [9]. The result is that the stress distribution obtained from 3D analysis is similar to that of 2D analysis thanks to the symmetrical geometry of the object concerned. In the present study, the loading condition used in this static stress calculation only consists of the simple axisymmetric axial force. The thickness of the tool joint is small. The pipe model and boundary condition are also axisymmetric. Therefore, 2D FEM was employed to calculate the static stress distribution. 4.1. Modeling
Fig. 4c. The images of the fracture surface. (c) Fractography of the crack propagation area.
Based upon the above calculation, the total axial force subjected to the drill pipe at failure position can then be determined as 2.91 106 N which will be applied to the finite element model. 3. Failure mode analysis The analysis of the chemical composition and mechanical performance of the drill pipe showed that the material is in accordance with the API SPEC 5D standard [14]. 3.1. Visual examination The pin tool joint of the drill pipe is shown in Fig. 3a. The teeth of the tool joint were numbered as T1 to T16 from the shoulder of the pin, while the number of the teeth root was R1 to R16 (Fig. 3b).
In this study, friction coefficient of the contact surfaces was taken as 0.02. Moreover, it was assumed that all 16 pairs of teeth on the tool joint are engaged together after tightening the tool joint. According to API standard, the drill specification parameters are listed in Table 4. The geometry of the model was created using a software ANSYS (Fig. 5). Plane42 elements were used for mesh generation. Besides, linear-based contact elements have been used for modeling the contact between engaging tooth pairs on both sides of the thread. Fig. 5 shows the finite element model of the problem and one of the most critical parts of the drill pipe, the tool joint threads. In this model, radial direction and axial direction were defined as X and Y direction, respectively. The left section of the model is the pin, while the right is the box. The boundary condition used in this model consists of the displacement constraint at the end of the pin and an axial load of 2.91 106 N applied to the opposite end of the box. Thirty-two thread contact elements and two shoulder contact elements were established in this model. The threads of the pin were numbered as 1–16 from the shoulder of the pin (Fig. 3b), while the number of the box teeth was in an ascending order from the mouth to the shoulder.
Table 4 Parameters of the drill pipe. Drill pipe specifications
Tool joint specification
Class
Size and style
Nominal weight
SS105
5½0 0 IEU
40.58 kg/m
DS55
Material properties E
V
rb
ry
P
200 GPa
0.3
1100 MPa
930 MPa
7800 kg/m3
312
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Fig. 5. Finite element model.
4.2. Static stress distribution of the tool joint The FEM analysis showed that the first and second tooth of the pin undertook the majority of the axial load, which was 16.85% and 10.34% of the total load, respectively (Fig. 6). The load distribution of the intermediate teeth is approximately the same. It means that the first tooth is the key location sustained load of the drill pipe, so this location usually fails firstly at abnormal work condition. Fig. 7a shows the static stress distribution along the tool joint. It can be seen that the maximum Von-Mises stress occurred at the first tooth root of the pin tool joint. The calculation result for this case showed that the maximum stress value is 578 MPa which is far below the yielding strength of the pipe (930 MPa as shown in Table 3), implying that the static stress concentration at the tooth root cannot cause failure by it alone. In this study, SCF is defined as the ratio of the maximum Von-Mises stress at the root of the tooth to the nominal stress (145 MPa) at the tool joint far from the thread [8]. Thus, for the specific tool joint and a known applied stress, Equivalent static
stress and SCF can be easily obtained from the results of FEM model. Results of the stress concentration factor at the root of the threads of the pin and box calculated from Fig. 7a are plotted in Fig. 7b. It shows the variation of SCF at the root of the teeth of the pin and box under working condition. It is observed that the maximum value of SCF is located at the root of the first tooth of the pin. The value of SCF at the first tooth of the pin, which is 3.99, is much higher than that of other teeth. If a fatigue load is superimposed on the static load, fatigue cracks may nucleate firstly at the root of the first tooth. 5. Discussion According to the static stress analysis using FEM, the tool joint sustained major part of the load compared with other region, and the value of Von-Mises stress in the first tooth of the pin is larger than other teeth. It means that the first tooth is the key location sustained major load of the drill pipe, this location usually fails first at abnormal work conditions. The FEM results showed that the maximum stress working on the tool joint is 578 MPa, which is far below the yielding strength of the pipe material, thus it should be safe at work condition. However, failure of the tool joints still occurs and fractographic examination revealed evidence of the fatigue crack propagation on the fracture surface of the tool joint. This indicates that there is other dynamic stress worked on the drill pipe. 5.1. Fatigue fracture mode As shown in Fig. 1, the drill pipe is always bearing the alternating bending stress in dogleg area due to rotating operation during drilling process [15]. The teeth root of the tool joint of rotating drill pipe is repeatedly subjected to tensile and compressive stresses resulting from the rotating and the bending load in the dogleg area. The bending stress ðrbending Þ is calculated based on the following formula [16]:
rbending ¼ Fig. 6. Load distribution on sixteen teeth of the pin.
DP E ðC W q sin b=F a Þg þ q sin b=F a 2 tanh g
ð3Þ
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
313
Fig. 7a. Static stress distribution of tool joint.
mum static axial stress at the root. The actual stress of the first tooth root of the pin is illustrated in Fig. 8. Fig. 8 shows that the maximum fatigue stress is higher than the yield strength of the material (listed in Table 4). Normally, under such a high fatigue load, multiple fatigue cracks would nucleate at the first tooth root and propagate to cause low cycle fatigue failure of the tool joint. This is in good agreement with the fractographic observation as shown in Fig. 3 and explains the reason for the short service life of the drilling pipes. 5.2. Evaluation based on fatigue test
Fig. 7b. SCF on teeth roots of pin and box.
pffiffiffiffiffiffiffiffiffiffiffiffi where g ¼ 0:5L F a =EI, Cw is the hole curvature, Dp is the outside diameter of drill pipe, E is the elastic modulus, I is the moment of inertia, L is the length of drill pipe, b is the average angle of inclination, q is gravity unit length pipe. Based on the parameters listed in Tables 1 and 2, the value of bending stress at the failure position as shown in Fig. 1a can be calculated as 398 MPa on the tensile side and 398 MPa on the opposite compressive side. It indicates that the drill pipe undertook fatigue stress at the failure position during operation. Considering the most dangerous position, the drill pipe worked at the constant axial stress along the axial direction and the alternating bending load which results in an additional axial stress. The constant axial stress can be considered as the mean value and the alternating bending stress as the amplitude of the fatigue stress spectrum. The fatigue fracture of the drill pipe can therefore be simply considered as the tension–tension fatigue fracture. The loading spectrum at the high stress concentrated first tooth root can be taken as the superimposition of the bending fatigue stress on the maxi-
Fatigue tests have been carried out under various fatigue loading modes [17–19] for a 42CrMo steel which is widely used for the tool joints of drill pipes [20]. The chemical composition (Table 5) and the mechanical properties (yielding strength 970 MPa and ultimate strength 1125 MPa) of the 42CrMo steel are almost same as those of the drill pipe steel concerned in the present study. Based on the S–N curve results of the 42CrMo steel, it can be found that the fatigue life (Nf) is in the range of 5 103–5 104 at a fatigue load of 350 MPa stress amplitude and 900 MPa peak stress which is very close to the loading spectrum applied to the drill pipe. This fatigue cycle number Nf can be converted into working time based
Fig. 8. The actual stress of the first tooth root of the pin.
314
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Table 5 Chemical composition of material of drill pipe (mass fraction, %).
42CrMo API standard
C
Si
Mn
P
S
Cr
Mo
Ni
Cu
0.38 0.25–0.35
0.27 0.15–0.35
0.8 0.9–1.1
0.014 60.015
0.004 60.015
1.2 1.2–1.5
0.17 0.75–0.85
0.05 60.25
0.03 60.25
on the working rotation speed of the drill pipe (60 rpm), and the calculation showed that the Nf range is equivalent to the drill pipe service time of 1.4–14 h. Comparing with the service time at failure of the failed drill pipes (5–15 h), it clearly demonstrates that the fatigue load resulting from bending at the dogleg region is the main reason of the drill pipe failure which is in agreement with the work of Ref. [21]. From Table 1 it can be seen that the case No. 3 has the worst combination of the gravity force and the dogleg deviation. This is the main reason why No. 3 failed drill pipe has the shortest working life. Therefore, to increase the service life of the drill pipes, severe doglegs must be avoided during design and the drilling of the oil wells, especially in the upper first segment of the drill string. 6. Conclusions Fractographic observation and FEM analysis of the stress distribution were carried out for the failed drill pipes. The following conclusions can be drawn from the investigation of the failure cases. (1) The maximum stress concentration occurred at the first tooth root from the pin tool joint shoulder of the drill pipe. (2) Bending moment in the dogleg region caused fatigue load to the tool joint during drilling operation which is superimposed on the axial stress resulting from gravity of the drill pipes. (3) The service stress at the first tooth root of the pin tool joint is the fatigue stress with mean stress (578 MPa) and stress amplitude (398 MPa). (4) Fatigue cracks were nucleated at the first tooth root due to stress concentration, propagated through the wall of the tool joint, and led to the final failure. Acknowledgement The authors would like to thank the China Petroleum & Chemical Corporation (Sinopec Group) for financial support and provision of test material and related information. Thanks are also due to School of Materials Science and Engineering of Beihang University for the provision of research facilities.
References [1] Miscow GF, de Miranda PEV, Netto TA, Placdio JCR. Techniques to characterize fatigue behaviour of full size drill pipes and small scale samples. Int J Fatigue 2004;26:575–84. [2] Netto TA, Lourenco MI, Botto A. Fatigue performance of pre-strained pipes with girth weld defects: full-scale experiments and analyses. Int J Fatigue 2008;30:767–78. [3] Pearson TL. Tubular failure cost/frequency analysis. In: Proc tubular goods symp. Houston (USA): Exxon Co., November 18–20; 1981. [4] Gensmer RP. Field correlation between internal taper length and tube failures in 4.5-in, 16.60E, IEU drill pipes. SPE Drilling Engineering; 1990 March. p. 58– 62. [5] Placido JCR et al. Fatigue analysis of aluminum drill pipes. J Mater Res 2005;8:409–15. [6] Tafreshi A, Dover WD. Stress analysis of drill string threaded connection using finite element method. Int J Fatigue 1993;15:429–38. [7] Tafreshi A. SIF evaluation and stress analysis of drill string threaded joint. Int J Press Ves Pip 1999;76:91–103. [8] Bahai H. A parametric model for axial and bending stress concentration factors in API drill string threaded connectors. Int J Press Vess Pip 2001;78:495–505. [9] Shahani AR, Sharifi SMH. Contact stress and calculation of stress concentration factors at the tool joint of a drill pipe. Mater Des 2009;30:3615–21. [10] Rahman MK, Hossain MM, Rahman SS. Stress concentration incorporated fatigue analysis of die-marked drill pipes. Int J Fatigue 1999;21:799–811. [11] Lv Shuanlu, Feng Yaorong, Luo Faqing, Qin Changyi, Wang Xinhu. Faliure analysis of IEU drill pipe wash out. Int J Fatigue 2005;27:1360–5. [12] Lv Shuan-lu, Li Bing-yin, Kang Yan-jun, Zhang Hong, Qin Terry, Chi Jun, et al. Cause analysis on drill pipe bending and fracture. Eng Fail Anal 2010;17:1483–9. [13] Zhiyong Han. Study on axial force calculating and strength testing for drilling in vertical holes. Petrol Drill Tech 1995;23:8–13. [14] American Petroleum Institute. API SPEC 5DP-2009. Washington, DC: Specification for drill pipe; 2009. [15] Baryshnikov A, Baragetti S. Rotary shouldered thread connections: working limits under combined static loading. J Mech Des 2001;123:456–63. [16] Shanzhou Ma, Zhiyong Han. Calculation of maximum bending stress in drilling rod under action of axial load and pipe weight. J Univ Petrol 2001;25:6–8. [17] Kang Guozheng, Liu Yujie. Uniaxial ratcheting and low-cycle fatigue failure of the steel with cyclic stabilizing or softening feature. Mater Sci Eng, A 2008;472:258–68. [18] Kang Guozheng, Liu Yujie, Ding Jun. Multiaxial ratchetting–fatigue interactions of annealed and tempered 42CrMo steel: experimental observations. Int J Fatigue 2008;1:2104–18. [19] Zhao Li, Liu Yujie, Kang Guzheng. Experimental study on uniaxial cyclic plastic behavior of annealed and tempered 42CrMo alloy steel. Acta Mech Solida Sin 2007;28:77–82. [20] Lin YC, Chen Ming-song, Zhong Jue. Effect of temperature and strain rate on the compressive deformation behavior of 42CrMo steel. J Mater Process Technol 2008;205:308–15. [21] Joosten MW. New study shows how to predict accumulated drill pipe fatigue. World Oil, October; 1985.
Contents lists available at SciVerse ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/matdes
Technical Report
Effect of stress distribution on the tool joint failure of internal and external upset drill pipes Shuai Luo, Sujun Wu ⇑ School of Materials Science and Engineering, Beihang University, Beijing 100191, PR China
a r t i c l e
i n f o
Article history: Received 3 December 2012 Accepted 24 May 2013 Available online 7 June 2013
a b s t r a c t Detailed investigation was carried out on the tool joint failure of the internal & external upset (IEU) GSS105 drill pipes. Stresses were analyzed using 2D finite element method (FEM) for the threaded tool joint of the drill pipe under the combined loading conditions containing preload, tensile and bending loads. The stress concentration factors in the pin and box were calculated. Results showed that the maximum stress concentration occurred at the roots of the first tooth from the pin tool joint shoulder of the drill pipe. Fractographic observation revealed that the tool joint failure of the drill pipe was caused by fatigue crack nucleated at the tooth root and propagated through the wall of the tool joint. The deterioration of the fatigue resistance of the tool joint is related to dogleg region where severe cyclic bending load exists due to the local deviation of the drill pipe from the vertical line. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Tool joints, including pin and box, are important parts joining drill pipes serially to make a drill string, which is the key tool in oil wells to transfer the applied torque to the bit at the bottom of the well. The failure of the drill pipes occurred mostly in the tool joint due to continuous changeable load of tension, bending, impact, internal pressure and a certain amount of torque [1–5]. Therefore, it is necessary and urgent to study the mechanism of fatigue failure in tool joint of the drill pipes. Threads at both ends of the tool joints play the vital role in connection. When the drill string is in service, considerable stresses are expected in the teeth of tool joint. Tafreshi and Dover [6] employed finite element method (FEM) to analyze the stress distribution around the threaded connections of tool joints, and presented the load distribution of the teeth of tool joint and stress concentration factors (SCF) at the teeth root. In their work, it was concluded that the location of maximum stress concentration at the tool joint was present. Tafreshi [7] and Bahai [8] analyzed the stresses at the drill pipe connections by using an axisymmetric 2D FEM model, which pointed out that the maximum SCF always located at the first engaged tooth of the pin. Stress concentration at the tooth root is inevitable, and should be within the design safety factor [9]. Therefore, the static stresses at the tooth root alone cannot cause fracture. Previous researches [10–12] on the failure of the drill pipes showed that the fatigue damage of drill pipes always occurs in the pipe section at the dogleg region. However, clear explanation ⇑ Corresponding author. Tel.: +86 1082316326; fax: +86 1082317108. E-mail address: [email protected] (S. Wu). 0261-3069/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.matdes.2013.05.073
is required for the reasons why the dogleg accelerates fatigue failure. In the present study, in order to understand the failure mechanism of the tool joints of the drill pipes, FEM analysis was carried out by employing Plane42 elements and linear contact elements at each tooth. Local stress distribution at each tooth of the threads and variations of SCF at the teeth roots of the pin and box were investigated under same work conditions. The effect of the alternating bending stress caused by dogleg area or bending section of the oil well on the fatigue failure was discussed.
2. Drill string configuration and axial force calculation Eight drill pipes fractured due to cracks at the tooth root of the tool joint during normal drilling in one oil well with a service life much shorter (5–15 h) than the design life (720 h). Therefore, it is of great significance to analyze the failure causes of the drill pipes. The detailed parameters of drill pipe failures are listed in Table 1. All the failed drill pipes are 5–1/200 IEU drill pipes with similar failure mode. In the present study, the failed drill pipe with the shortest service life of 302 min before failure, Case No. 3, was taken as the failure analysis specimen for consideration. The drilling direction of drill pipe always changes and deviates from the vertical direction due to the geologic characteristics and manufacture related factors. The projection drawing of the oil well, including the drilling depth, overall variation of the drill string inclination angle and the dogleg severity, is shown in Fig. 1. It was found that the depth of failure position on the drill pipe concerned is 910 m underground, where the actual maximum overall 3D angle change rate (dogleg severity) is 0.53°/25 m within the first 1000 m depth from the ground.
309
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314 Table 1 Parameters of the drill pipe failures. Failure case (No.)
Well total depth (m)
Depth of failure position (m)
Service life (Min)
Angle of inclination (b/deg)
Dogleg severity (deg/25 m)
1 2 3 4 5 6 7 8
5596.11 5598.81 6362.69 6576.20 6588.63 6604.00 6631.48 6631.63
930 930 910 1400 1350 2000 2350 2300
788 790 302 394 409 412 426 427
0.29 0.29 0.29 0.27 0.29 0.26 0.24 0.36
0.05 0.05 0.53 0.17 0.45 0.33 0.34 0.63
Fig. 1. The projection drawing of the oil well (a) vertical drawing and (b) horizontal view.
The depth of the oil well is 6362.69 m for Case No. 3. The drilling string of the well is divided into four segments as illustrated in Fig. 2 and the detailed parameters of each segment are listed in Table 2. The failure position of the drill pipe is in the first segment which is the IEU GSS105 drill pipe with a diameter of 139 mm. The axial force of the drill pipe at failure position which is required for the FEM stress analysis can be divided into two parts. One is from the gravity of the drilling string and another is from the torque. The force from the gravity can be calculated based on the following formula [13]:
Fa ¼
n X DHk qkm Ha ðAa0 c0 Aa1 c1 Þ
ð1Þ
K¼1
where qkm ¼ qK ðAK0 c0 AK1 c1 Þ, Fa is the axial force at failure position under wellhead, n is the segment of drilling string under failure position, DH is the length of segment, Ha is the distance from ground to failure position, Aa0 is the outer circle area of the failed segment, Aa1 is the inner circle area of the failed segment, AK0 is the outer circle area of the kth segment, AK1 is the inner circle area of the kth segment, c0 is pipe mud weight rate, c1 is tube mud weight rate, qk is gravity unit length pipe in air, qkm is gravity unit length pipe in mud. Based on the above theoretical calculation, an axial force from gravity was obtained as 1.38 106 N.The axial force resulting from the applied torque can be calculated based on the following formula:
Fig. 2. The sketch of the drilling string.
310
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Table 2 Parameters of the drilling string. Segment k
Inner diameter (m)
Outer diameter (m)
H (m)
q (N/m)
c0 (N/m3)
c1 (N/m3)
First Second Third Fourth
0.1186 0.108 0.0702 0.0508
0.1397 0.127 0.0889 0.1206
2659.22 2543.28 960.35 200.54
397.684 261.758 179.732 681.002
11.564e+3
11.466e+3
Table 3 Parameters of the drilling pipe at the failure position. Torque
Preloading torque
Hole curvature (Cw)
Angle of inclination (b)
Moment of inertia (I)
Rotary speed
2537 N m
32276 N m
0.0215 rad/m
0.29°
9.09 105 m4
60 rpm
Fig. 3a. Macroscopic images of (a) cracked drill pipe. Fig. 3d. Macroscopic images of (d) crack on inner surface of the tool joint of the drill pipe.
T¼
FT d2 tanðw qv Þ 2
ð2Þ
Fig. 3b. Macroscopic images of (b) numbering order of the pin tool joint.
where T is the applied torque, FT is the axial force resulting from the applied torque, d2 is the effective diameter of thread, w is the lead angle and qv is the equivalent friction angle. According to the detailed drilling parameters at failure position in Table 3, an axial force from applied torque can be calculated as 1.53 106 N.
Fig. 3c. Macroscopic images of (c) crack at the first tooth root of the tool joint.
Fig. 4a. The images of the fracture surface. (a) The crack initiation sites at the threaded root.
311
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
The crack was nucleated at the first tooth root from the tool joint shoulder of the pin (Fig. 3c) and propagated along the thread root to the inner surface of the drill pipe, as shown in Fig. 3d which is the piece cut from Fig. 3c along the two white lines. 3.2. Fractographic examination Scanning electron microscope (SEM) was applied to observe the fracture surface of the cracked tool joint. Figs. 4a and 4b reveals that multiple fatigue cracks were initiated along the tooth root of the tool joint and propagated towards the inner surface of the drill pipe which implies that the tooth root area was subjected to highly concentrated stress. The initial cracks joined together after some amount of inward extension to form a major crack. Further observation shows that striations are present on the fracture surface (Fig. 4c) which confirms that the failure was caused by fatigue. Fig. 4b. The images of the fracture surface. (b) The magnification of the crack initiation sites.
4. Finite element modeling In the previous study, 2D finite element modeling has been used to calculate the stress distribution under axial and bending loads for the drill string threaded connections [6–8]. Based on the experimental verification, it was reported that the stress analysis of the experiment study were the same as the FEM calculation. Moreover, 3D FEM analysis has been used to investigate the loads distribution under the compressive loading, tensile loading and preloading [9]. The result is that the stress distribution obtained from 3D analysis is similar to that of 2D analysis thanks to the symmetrical geometry of the object concerned. In the present study, the loading condition used in this static stress calculation only consists of the simple axisymmetric axial force. The thickness of the tool joint is small. The pipe model and boundary condition are also axisymmetric. Therefore, 2D FEM was employed to calculate the static stress distribution. 4.1. Modeling
Fig. 4c. The images of the fracture surface. (c) Fractography of the crack propagation area.
Based upon the above calculation, the total axial force subjected to the drill pipe at failure position can then be determined as 2.91 106 N which will be applied to the finite element model. 3. Failure mode analysis The analysis of the chemical composition and mechanical performance of the drill pipe showed that the material is in accordance with the API SPEC 5D standard [14]. 3.1. Visual examination The pin tool joint of the drill pipe is shown in Fig. 3a. The teeth of the tool joint were numbered as T1 to T16 from the shoulder of the pin, while the number of the teeth root was R1 to R16 (Fig. 3b).
In this study, friction coefficient of the contact surfaces was taken as 0.02. Moreover, it was assumed that all 16 pairs of teeth on the tool joint are engaged together after tightening the tool joint. According to API standard, the drill specification parameters are listed in Table 4. The geometry of the model was created using a software ANSYS (Fig. 5). Plane42 elements were used for mesh generation. Besides, linear-based contact elements have been used for modeling the contact between engaging tooth pairs on both sides of the thread. Fig. 5 shows the finite element model of the problem and one of the most critical parts of the drill pipe, the tool joint threads. In this model, radial direction and axial direction were defined as X and Y direction, respectively. The left section of the model is the pin, while the right is the box. The boundary condition used in this model consists of the displacement constraint at the end of the pin and an axial load of 2.91 106 N applied to the opposite end of the box. Thirty-two thread contact elements and two shoulder contact elements were established in this model. The threads of the pin were numbered as 1–16 from the shoulder of the pin (Fig. 3b), while the number of the box teeth was in an ascending order from the mouth to the shoulder.
Table 4 Parameters of the drill pipe. Drill pipe specifications
Tool joint specification
Class
Size and style
Nominal weight
SS105
5½0 0 IEU
40.58 kg/m
DS55
Material properties E
V
rb
ry
P
200 GPa
0.3
1100 MPa
930 MPa
7800 kg/m3
312
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Fig. 5. Finite element model.
4.2. Static stress distribution of the tool joint The FEM analysis showed that the first and second tooth of the pin undertook the majority of the axial load, which was 16.85% and 10.34% of the total load, respectively (Fig. 6). The load distribution of the intermediate teeth is approximately the same. It means that the first tooth is the key location sustained load of the drill pipe, so this location usually fails firstly at abnormal work condition. Fig. 7a shows the static stress distribution along the tool joint. It can be seen that the maximum Von-Mises stress occurred at the first tooth root of the pin tool joint. The calculation result for this case showed that the maximum stress value is 578 MPa which is far below the yielding strength of the pipe (930 MPa as shown in Table 3), implying that the static stress concentration at the tooth root cannot cause failure by it alone. In this study, SCF is defined as the ratio of the maximum Von-Mises stress at the root of the tooth to the nominal stress (145 MPa) at the tool joint far from the thread [8]. Thus, for the specific tool joint and a known applied stress, Equivalent static
stress and SCF can be easily obtained from the results of FEM model. Results of the stress concentration factor at the root of the threads of the pin and box calculated from Fig. 7a are plotted in Fig. 7b. It shows the variation of SCF at the root of the teeth of the pin and box under working condition. It is observed that the maximum value of SCF is located at the root of the first tooth of the pin. The value of SCF at the first tooth of the pin, which is 3.99, is much higher than that of other teeth. If a fatigue load is superimposed on the static load, fatigue cracks may nucleate firstly at the root of the first tooth. 5. Discussion According to the static stress analysis using FEM, the tool joint sustained major part of the load compared with other region, and the value of Von-Mises stress in the first tooth of the pin is larger than other teeth. It means that the first tooth is the key location sustained major load of the drill pipe, this location usually fails first at abnormal work conditions. The FEM results showed that the maximum stress working on the tool joint is 578 MPa, which is far below the yielding strength of the pipe material, thus it should be safe at work condition. However, failure of the tool joints still occurs and fractographic examination revealed evidence of the fatigue crack propagation on the fracture surface of the tool joint. This indicates that there is other dynamic stress worked on the drill pipe. 5.1. Fatigue fracture mode As shown in Fig. 1, the drill pipe is always bearing the alternating bending stress in dogleg area due to rotating operation during drilling process [15]. The teeth root of the tool joint of rotating drill pipe is repeatedly subjected to tensile and compressive stresses resulting from the rotating and the bending load in the dogleg area. The bending stress ðrbending Þ is calculated based on the following formula [16]:
rbending ¼ Fig. 6. Load distribution on sixteen teeth of the pin.
DP E ðC W q sin b=F a Þg þ q sin b=F a 2 tanh g
ð3Þ
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
313
Fig. 7a. Static stress distribution of tool joint.
mum static axial stress at the root. The actual stress of the first tooth root of the pin is illustrated in Fig. 8. Fig. 8 shows that the maximum fatigue stress is higher than the yield strength of the material (listed in Table 4). Normally, under such a high fatigue load, multiple fatigue cracks would nucleate at the first tooth root and propagate to cause low cycle fatigue failure of the tool joint. This is in good agreement with the fractographic observation as shown in Fig. 3 and explains the reason for the short service life of the drilling pipes. 5.2. Evaluation based on fatigue test
Fig. 7b. SCF on teeth roots of pin and box.
pffiffiffiffiffiffiffiffiffiffiffiffi where g ¼ 0:5L F a =EI, Cw is the hole curvature, Dp is the outside diameter of drill pipe, E is the elastic modulus, I is the moment of inertia, L is the length of drill pipe, b is the average angle of inclination, q is gravity unit length pipe. Based on the parameters listed in Tables 1 and 2, the value of bending stress at the failure position as shown in Fig. 1a can be calculated as 398 MPa on the tensile side and 398 MPa on the opposite compressive side. It indicates that the drill pipe undertook fatigue stress at the failure position during operation. Considering the most dangerous position, the drill pipe worked at the constant axial stress along the axial direction and the alternating bending load which results in an additional axial stress. The constant axial stress can be considered as the mean value and the alternating bending stress as the amplitude of the fatigue stress spectrum. The fatigue fracture of the drill pipe can therefore be simply considered as the tension–tension fatigue fracture. The loading spectrum at the high stress concentrated first tooth root can be taken as the superimposition of the bending fatigue stress on the maxi-
Fatigue tests have been carried out under various fatigue loading modes [17–19] for a 42CrMo steel which is widely used for the tool joints of drill pipes [20]. The chemical composition (Table 5) and the mechanical properties (yielding strength 970 MPa and ultimate strength 1125 MPa) of the 42CrMo steel are almost same as those of the drill pipe steel concerned in the present study. Based on the S–N curve results of the 42CrMo steel, it can be found that the fatigue life (Nf) is in the range of 5 103–5 104 at a fatigue load of 350 MPa stress amplitude and 900 MPa peak stress which is very close to the loading spectrum applied to the drill pipe. This fatigue cycle number Nf can be converted into working time based
Fig. 8. The actual stress of the first tooth root of the pin.
314
S. Luo, S. Wu / Materials and Design 52 (2013) 308–314
Table 5 Chemical composition of material of drill pipe (mass fraction, %).
42CrMo API standard
C
Si
Mn
P
S
Cr
Mo
Ni
Cu
0.38 0.25–0.35
0.27 0.15–0.35
0.8 0.9–1.1
0.014 60.015
0.004 60.015
1.2 1.2–1.5
0.17 0.75–0.85
0.05 60.25
0.03 60.25
on the working rotation speed of the drill pipe (60 rpm), and the calculation showed that the Nf range is equivalent to the drill pipe service time of 1.4–14 h. Comparing with the service time at failure of the failed drill pipes (5–15 h), it clearly demonstrates that the fatigue load resulting from bending at the dogleg region is the main reason of the drill pipe failure which is in agreement with the work of Ref. [21]. From Table 1 it can be seen that the case No. 3 has the worst combination of the gravity force and the dogleg deviation. This is the main reason why No. 3 failed drill pipe has the shortest working life. Therefore, to increase the service life of the drill pipes, severe doglegs must be avoided during design and the drilling of the oil wells, especially in the upper first segment of the drill string. 6. Conclusions Fractographic observation and FEM analysis of the stress distribution were carried out for the failed drill pipes. The following conclusions can be drawn from the investigation of the failure cases. (1) The maximum stress concentration occurred at the first tooth root from the pin tool joint shoulder of the drill pipe. (2) Bending moment in the dogleg region caused fatigue load to the tool joint during drilling operation which is superimposed on the axial stress resulting from gravity of the drill pipes. (3) The service stress at the first tooth root of the pin tool joint is the fatigue stress with mean stress (578 MPa) and stress amplitude (398 MPa). (4) Fatigue cracks were nucleated at the first tooth root due to stress concentration, propagated through the wall of the tool joint, and led to the final failure. Acknowledgement The authors would like to thank the China Petroleum & Chemical Corporation (Sinopec Group) for financial support and provision of test material and related information. Thanks are also due to School of Materials Science and Engineering of Beihang University for the provision of research facilities.
References [1] Miscow GF, de Miranda PEV, Netto TA, Placdio JCR. Techniques to characterize fatigue behaviour of full size drill pipes and small scale samples. Int J Fatigue 2004;26:575–84. [2] Netto TA, Lourenco MI, Botto A. Fatigue performance of pre-strained pipes with girth weld defects: full-scale experiments and analyses. Int J Fatigue 2008;30:767–78. [3] Pearson TL. Tubular failure cost/frequency analysis. In: Proc tubular goods symp. Houston (USA): Exxon Co., November 18–20; 1981. [4] Gensmer RP. Field correlation between internal taper length and tube failures in 4.5-in, 16.60E, IEU drill pipes. SPE Drilling Engineering; 1990 March. p. 58– 62. [5] Placido JCR et al. Fatigue analysis of aluminum drill pipes. J Mater Res 2005;8:409–15. [6] Tafreshi A, Dover WD. Stress analysis of drill string threaded connection using finite element method. Int J Fatigue 1993;15:429–38. [7] Tafreshi A. SIF evaluation and stress analysis of drill string threaded joint. Int J Press Ves Pip 1999;76:91–103. [8] Bahai H. A parametric model for axial and bending stress concentration factors in API drill string threaded connectors. Int J Press Vess Pip 2001;78:495–505. [9] Shahani AR, Sharifi SMH. Contact stress and calculation of stress concentration factors at the tool joint of a drill pipe. Mater Des 2009;30:3615–21. [10] Rahman MK, Hossain MM, Rahman SS. Stress concentration incorporated fatigue analysis of die-marked drill pipes. Int J Fatigue 1999;21:799–811. [11] Lv Shuanlu, Feng Yaorong, Luo Faqing, Qin Changyi, Wang Xinhu. Faliure analysis of IEU drill pipe wash out. Int J Fatigue 2005;27:1360–5. [12] Lv Shuan-lu, Li Bing-yin, Kang Yan-jun, Zhang Hong, Qin Terry, Chi Jun, et al. Cause analysis on drill pipe bending and fracture. Eng Fail Anal 2010;17:1483–9. [13] Zhiyong Han. Study on axial force calculating and strength testing for drilling in vertical holes. Petrol Drill Tech 1995;23:8–13. [14] American Petroleum Institute. API SPEC 5DP-2009. Washington, DC: Specification for drill pipe; 2009. [15] Baryshnikov A, Baragetti S. Rotary shouldered thread connections: working limits under combined static loading. J Mech Des 2001;123:456–63. [16] Shanzhou Ma, Zhiyong Han. Calculation of maximum bending stress in drilling rod under action of axial load and pipe weight. J Univ Petrol 2001;25:6–8. [17] Kang Guozheng, Liu Yujie. Uniaxial ratcheting and low-cycle fatigue failure of the steel with cyclic stabilizing or softening feature. Mater Sci Eng, A 2008;472:258–68. [18] Kang Guozheng, Liu Yujie, Ding Jun. Multiaxial ratchetting–fatigue interactions of annealed and tempered 42CrMo steel: experimental observations. Int J Fatigue 2008;1:2104–18. [19] Zhao Li, Liu Yujie, Kang Guzheng. Experimental study on uniaxial cyclic plastic behavior of annealed and tempered 42CrMo alloy steel. Acta Mech Solida Sin 2007;28:77–82. [20] Lin YC, Chen Ming-song, Zhong Jue. Effect of temperature and strain rate on the compressive deformation behavior of 42CrMo steel. J Mater Process Technol 2008;205:308–15. [21] Joosten MW. New study shows how to predict accumulated drill pipe fatigue. World Oil, October; 1985.