Experimental and analytical studies on the pullout strength of round thread casing connections

Indian Journal of Engineering & Materials Science Vol. 20, December 2013, pp. 497-503 Experimental and analytical studies on the pullout strength of round thread casing connections Lianxin Gao a * & Jiaoqi Shi b a Key Laboratory of Pressure Systems and Safety, Ministry of Education, East China University of Science & Technology, Shanghai 200237, China b Tubular Goods Research Institute of China National Petroleum Corporation, Shaanxi, Xi'an, 710065, China Received 26 February 2013; accepted 8 July 2013 Pullout failure is one of the primary failure modes of API round thread casing connections. Make-up torque and thread tolerances are the two most important factors that can affect the pullout strength. In this paper, the pull-out failure process and mechanism are studied using the finite element analysis (FEA) method. In addition, the effects of make-up torque and thread tolerances on the pullout strength of API round thread connections are analyzed. The FEA results are validated through tension tests on a total of 34 pins and 17 couplings. The results show that pull-out failure initially occurs at the first and second teeth near the coupling end. At the same time, thread tolerances especially unmatched taper have notable effect on pullout strength. The tests show that higher make-up torque levels are beneficial to the pullout strength. Make-up torque 15~20% higher than the maximum API limit is recommended when round thread casing connections are used in oilfield. Keywords: Round thread casing, Pullout strength, Full-scale test, Finite element analysis The American Petroleum Institute (API) round thread casing connection is one of the primary consumables among Oil Country Tubular Goods (OCTG) for conventional oilfield applications. The quality of casing and connections can directly affect oil well life and safety. API round thread casing connection has many more practical features over the proprietary casing connection joints, such as its relative low cost, the standardization of tools and gauges, and substantial knowledge accumulated by the industry of its limits and characteristics. Of major concern, however, is its low reliability of round thread connections which may fail due to pullout mechanism, resulting in a potential high cost. This paper therefore focuses on an analytical study of the mechanism of pullout failure to identify possible methods to improve connection strength and reliability 1-4. It is well known that thread geometry and associated tolerances, and make-up torque are the two key factors affecting the pullout strength of round thread casing connections 5-7. In this paper, the effects of the make-up torque, and thread tolerances on the pullout strength are studied systematically using elastic-plastic finite element analysis, which was also validated with experimental results. This study has led *Corresponding author (E-mail: gaolianxin@163.com) to a few conclusions that can be useful to the manufacturers and users of API round thread casing. Theoretical Analysis API round thread connections are tapered coupling connections. Their geometric dimensions and tapers, as well as the associated tolerances are defined in API SPEC 5B 8. The round thread connections were designed for casing with OD varying from 114.3 mm to 508.0 mm, for external upset tubing with OD from 60.3 mm to 114.3 mm, and for non-upset tubing with OD from 101.6 mm to 114.3 mm. Figure 1 shows a schematic of an API round thread connection with its characteristic parameters. As shown in Fig. 2, upon make-up, there will be a normal force, F S, acting on the stab flank of the thread because of the thread interferences and a normal force, F L, acting on the load flank. When tension load is applied to the connection, F S decreases and F L increases. In addition, F S and F L can decrease slightly due to Poisson effects for a round thread joint loaded in tension. Note that in case the tension load is significantly large, the stab flank of the pin and coupling will separate and the force F S will diminish to zero. When the force induced by tension is sufficient to overcome the friction force, slide will occur and the joint will be subjected to a pullout

498 INDIAN J ENG MATER SCI., DECEMBER 2013 Fig. 1 Dimensions of casing round thread Fig. 2 Progressive schemes of API round thread connection under tension failure. The progression of the relative magnitudes of F L and F S are shown in Fig. 2. The pullout strength may be increased by increasing the resulting engaged thread length or increasing the contact force F L on the load flank. A higher make-up torque can increase the engaged thread length, while a larger contact force can increase the pullout strength. However, the effective stress in a joint can also increase with higher make-up torque, resulting in a decrease in the pullout strength. In general, the thread standoff tolerance directly affects the torque-position relationship and therefore the resulting engaged thread length. Other parameters, such as taper, thread height, or lead also affect the pullout strength, but these effects are usually considered as second order of importance when compared to the standoff tolerance, and as a result, little research has been done on their effects. However, from Fig. 2, it was apparent that the taper mismatch and the thread height could affect the contact area of the load flank, as well as the friction force. This paper therefore, considered the effects of the make-up torque, engaged thread length, taper mismatch, and thread height tolerance on pullout strength. A comprehensive analysis was performed on a 244.5 10.03 mm, J55, LTC casing. Elastic-plastic finite element analysis was conducted, along with physical tension tests on a total of 34 pins and 17 couplings. The physical tests were specifically designed for this study in order to validate the FEA findings. All tests were performed in the full-scale laboratory of Tubular Goods Research Centre of China National Petroleum Corporation. Numerical Analysis FEA is an ideal analytical tool for the study of threaded connections because its numerical formulation is capable of the analysis of complex geometry and materials. As the helical angle of the casing thread is very small, the connection was modeled as an axissymmetric entity to reduce computational time and cost. Geometric nonlinearities, material nonlinearities, and nonlinear boundary conditions were all considered in the model. The Coulomb friction model was used to simulate the contact of the threads. The friction coefficient is related to the grease of the connection, with typical values in the range of 0.015~0.025. This paper considered a friction value of 0.02. The commercial software MSC/MARC was used to run the nonlinear FEA. Three-node axi-symmetric elements were used in the model with the mesh shown in Fig. 3. In order to allow direct comparison with experimental results, the material model used in the FEA was taken from mechanical test results performed on the same parent material with the key parameters as:

GAO & SHI: PULLOUT STRENGTH OF ROUND THREAD CASING CONNECTIONS 499 E=2.06 10 5 MPa, µ=0.3, Y p =482 MPa, U p =715 MPa, δ=32%. The FEA results for the thread geometric parameters set at the nominal values, are show in Table 1. As can be seen, the FEA results are in good agreement with the test results, suggesting that the FEA model adequately simulated the physical testing system. With the FEA model shown in Fig. 3, we calculated the pullout loads under a range of make-up torques, taper, and thread heights. The effect of make-up torque was studied whilst setting the thread geometric parameters at their nominal values. Similarly, the effect of taper and thread height was studied whilst considering a make-up torque of a nominal value of 7.02 kn m. The FEA results shown in Tables 2 and 3 indicate that the make-up torque has a notable effect on pullout strength. In general, the greater the make-up torque, the larger the pullout loads. However, the rate of increase in pullout load diminishes for make-up torque values beyond the standard values. The effect of taper mismatch is particularly critical to the pullout strength, especially when the pin taper is at the maximum (67.7 mm/m) and the coupling taper is at the minimum (59.9 mm/m). The pullout load decreased by 10.5% for the Pin max +Coupling min situation when compared to the joint with nominal thread dimensions. When other thread parameters are constant, a positive thread height tolerance can increase the Make-up torque 7.02 kn m Table 1 Calculation results of pullout loads Pullout loads (kn) FEA Test result API value 8 2745 2597, 2703 2314 pullout strength. The pullout load increased by 6.6% for the Pin max (1.81+0.051 mm)+coupling max (1.81+0.051 mm) situation as compared to the joint with nominal thread dimensions. However, the pullout load decreased by 7.9% for the Pin min (1.81-0.102 mm)+coupling min (1.81-0.102 mm) situation. Experimental Study Sample description Casing connections with 244.5 10.03 mm, J55 material and LTC were selected for the experiment. The purpose of the experiment was to study the effects of the make-up torque, engaged thread length, taper, and thread height on the pullout strength. The test results were also used to validate the FEA conclusions. All the samples were machined from the same material to the selected thread dimensions. A total of 34 pins and 17 couplings were manufactured and these pins and couplings were combined to define the specimens for tension tests. The details of these samples are shown in Table 4. Results and Discussion Deformation of thread By observing and measuring the specimens after pullout failure, it was found that there were significant damages in the end of engaged threads (about 25-27 teeth from the pipe end face). Tiny metal wire was observed that had separated from the thread crest and thread bearing surface. The other threads did not show any obvious changes. Meanwhile, the thread taper of Table 2 Effect of make-up torque on connection pullout loads Make-up torque (kn m) 5.3 7.02 11.9 Pullout loads (kn) 2536 2745 2780 Table 3 Effect of thread parameters on connection pullout loads Parameters Thread height Pin max +Coupling min Pin min +Coupling max Pin max +Coupling max Pin min +Coupling min Pullout loads (kn) 2457 2620 2925 2528 Fig. 3 FEA model of API 8-rd connection

500 INDIAN J ENG MATER SCI., DECEMBER 2013 the pin showed obvious changes after the test, as seen in Fig. 4. The box showed no obvious changes. The measurement found that the pin shrunk notably in the first to second engaged threads and this was in Sample Pin Table 4 Objective values of the samples Thread height agreement with the findings from the finite element analysis. The deformation of the pin threads after pullout failure can be seen in Fig. 5. There was apparent Coupling Thread height Make-up torque(kn-m) 1 68 0 60 0 7.02 2 68 0 60 0 7.02 3 60 0 68 0 7.02 4 60 0 68 0 7.02 5 63 0 63 0 0 6 63 0 63 0 4.2 7 63 0 63 0 5.3 8 63 0 63 0 8.9 9 63 0 63 0 10.9 10 63 0 63 0 14.4 11 63 0 63 0 17.5 12 63 0 63 0 7.02 13 63 0 63 0 7.02 14 63-0.102 63-0.102 7.02 15 63 0.051 63 0.051 7.02 16 60 0 60 0 7.02 17 68 0 68 0 7.02 Fig. 4 Photographs of the pin before and after pullout failure

GAO & SHI: PULLOUT STRENGTH OF ROUND THREAD CASING CONNECTIONS 501 damage occurring in the end of the engaged pin threads. The bearing surface bends visible toward the leading surface. No obvious deformation could be found on the remaining threads. For coupling threads, the first three threads from end plane showed apparent damage. Tiny metal wire can be found separated from the thread crest and thread bearing surface. The deformation on the remaining threads was less apparent. A closer look shows that the damaged coupling threads matched with the damaged pin threads. From the above observation, it can be said that the first two threads in the end might carry the maximum force when the connection was under tensile load. Pullout failure might have initiated in the first two threads. Furthermore, when the threads in this position become ineffective, the rest of the threads will slip rapidly due to insufficient bearing capacity. The experimental results agreed well with the findings from finite element analysis. Pullout load of connections Table 5 shows the actual thread parameters (include taper and thread height), actual make-up torque, J value (see Fig. 1) and pullout loads (P is actual pullout load, P j is theoretical value which is the pullout load calculated by the formula from API BUL 5C3 10, 2314 kn). As can be noted, the taper of samples 1-4 are not within the API limits (according Sample Pin Thread height Table 5 Test results Coupling Thread height Make-up torque (kn-m) J (mm) Pullout load(kn) 1 70 0.025 60 0 7.1 11.55 2502 1.081 2 77-0.025 60-0.013 7.0 16.5 2228 0.963 3 59 0 70 0 7.0 14.6 2222 0.960 4 56 0 70 0 7.1 18.55 2104 0.909 5 63 0.02 63-0.025 0 28.4 2137 0.923 6 62 0 63 0 4.3 17.5 2413 1.043 7 63 0 63 0 5.3 19.2 2413 1.043 8 63 0 63 0 7.1 15.6 2597 1.122 9 64 0 62 0 8.9 14.5 2582 1.116 10 63-0.013 63 0.013 10.9 13.5 2648 1.144 11 62 0 63 0 14.4 12.5 2693 1.164 12 63 0 63 0 17.9 9.7 2777 1.200 13 62 0 64 0 7.1 15.3 2703 1.168 14 63-0.076 63-0.051 7.1 16.65 2485 1.074 15 63 0.051 63 0.051 7.0 10.25 2749 1.188 16 60 0 61 0 7.2 13.95 2547 1.101 17 66 0 67 0 7.1 16.5 2685 1.160 P/P j Fig. 5 Photographs of thread teeth before and after pullout failure

502 INDIAN J ENG MATER SCI., DECEMBER 2013 to API standard, the value is 59.9-67.7 mm/m) and are somewhat different from the desired values. These four samples were used to show the trend of taper mismatch effects. Other parameters are all in the standard and well controlled to meet the objective. Make-up torque and pullout strength Figure 6 shows that the make-up torque has a notable effect on pullout strength. When the torque is 0.6A (A is the optimal torque according to API BUL 5C2 9 ), the pullout strength is still greater than the value given in API BUL 5C2 (2413 kn-m). However, when make-up torque is less than 0.6A, the pullout strength decreases dramatically. Especially when the torque is 0 (hand tight), the pullout strength is only 2137 kn-m; a decrease of 18% compared to the sample whose make-up torque is A. The pullout strength is higher for the samples whose make-up torque is 1.25A-2.5A. Higher make-up torque is beneficial to pullout strength. However, when the torque is too high, the increase becomes less significant (when the torque is 2.5A, the pullout strength only increased by 7% compared to the sample whose torque is A); this is consistent with the FEA results. Fig. 6 Make-up torque versus pullout strength relationship Engaged thread length and pullout strength The following equation was established to calculate the engaged thread length: L = N L / 2 M J where, L is engaged thread length (mm), N L is coupling length (mm), M is the length between coupling end and hand-tight face (see Fig. 1) and J is the distance between pipe end and coupling center after make-up (see Fig. 1) Figure 7 shows the relationship between engaged thread length and pullout strength. As can be seen, in general, the pullout strength increases with engaged thread length. For example, the engaged thread length of sample 4 is the shortest, its pullout strength is also near the lowest observed. Meanwhile, the engaged thread length of sample 12 is the longest, with the largest pullout strength. The above conclusion can apply to the samples with similar thread parameters too, as shown in Fig. 7b. However, for samples with unmatched thread parameters, this conclusion is not valid, as shown in Fig. 7a. This indicates that the thread parameters (taper, thread height etc) also have notable effects on pullout strength. In general, however, longer engaged thread length results in higher pullout strength, regardless the thread parameter values. Thread taper and pullout strength Six samples (1-4 and 16-17) were machined to study the taper versus pullout strength relationship. Test results showed that unmatched taper (1-4 samples) had notable effect on pullout strength. Three among the four samples showed the pullout strength lower than the prediction based on API BUL 5C2. The average value is only about 82% of the matched taper. However, although the tapers for the Fig. 7 Engaged thread length versus pullout strength relationship

GAO & SHI: PULLOUT STRENGTH OF ROUND THREAD CASING CONNECTIONS 503 four samples were beyond API limits, the results indicate the significant influence of unmatched taper on pullout strength. Meanwhile, although the tapers for the samples 16 and 17 were also near the API limits, because their tapers are suited to each other, they have little effect on pullout strength. Thread height and pullout strength As can be noted from the test results, the pullout strength of sample 14 (Pin min +Coupling min ) decreased by about 9.6% when compared to sample 15 (Pin max +Coupling max ), suggesting that thread height also has some effect on pullout strength and is also consistent with the FEA results. Conclusions The following conclusions can be drawn from this study: Pullout failure initially occurs at the first or second teeth near the coupling end. Increasing the load capability of these two teeth can notably improve the joint strength. Because both make-up torque and engaged thread length have a notable effect on pullout strength, make-up procedure based only on position control or torque control is not sufficient. If torque control procedure is adopted, the J value should also be controlled in order to get a proper make-up. Higher make-up torque levels are beneficial to the pullout strength. Make-up torque 15-20% higher than the maximum API limit is recommended as long as the joint does not experience galling under this torque. The pullout strength will decrease by about 10-20% if the tapers of the pin and coupling are unmatched. Joint manufacturers should control this thread parameter carefully. Acknowledgments This study was supported by Tubular Goods Research Centre of CNPC. The authors would like to thank Dr. Junren Xie, Senior Engineer, C-FER Technologies Canada, for improving the paper. References 1 Hainey M F, Make-up torques for API type round thread casing connections with non-api weights, grade, and coupling diameters. SPE 15517. 2 Yuan Guangjie & Yao Zhengqiang, Eng Fail Anal, 13(8) (2006) 1275-1284 3 Chen Shoujun, Gao Lianxin & An Qi, Frontiers Mech Eng China, 6(2) (2011) 241-248 4 Duan W & Joshi S, Eng Fail Anal, 18(8) (2011) 2008-2018 5 Solazzi L & La Vecchia G M, J Fail Anal Preven, 12 (2012) 41-549 6 Pan Zhiyong & Song Shengyin, Adv Mater Res, 2012,479-481:2028-2032 7 Mallenhalli & Nandakumar K, Oil Gas J, 90(3) (1992) 51-53 8 Specification for threading, gauging, and thread inspection of casing, tubing, and line pipe thread, API SPEC 5B, 4 th ed, (American Petroleum Institute), 1996. 9 Bulletin on performance properties of casing, tubing, and drill pipe, API Bulletin 5C2, 20 th ed, (American Petroleum Institute), 1987. 10 Bulletin on formulas and calculations for casing, tubing, drill pipe, and line pipe properties, API Bulletin 5C3, 6 th ed, (American Petroleum Institute), 1994.