PSI - Issue 13

Hikaru Yamaguchi et al. / Procedia Structural Integrity 13 (2018) 1183–1188 Hikaru Yamaguchi/ Structural Integrity Procedia 00 (2018) 000 – 000

1184

2

history accurately due to its simplicity. Actual phenomenon of unstable ductile fracture is very complicating. To incorporate this complex behavior of unstable ductile fracture, many attempts have been made. Donoghue developed a numerical model, in which pipe wall deformation and crack propagation was analyzed by finite-element method, gas decompression was analyzed by finite-difference method and the both were coupled [2]. Researchers at CSM developed a similar numerical model [3]. There are some other numerical models to analyze unstable ductile fracture, e.g. [4, 5]. On the other hand, part of the authors have developed a rigorous numerical model [6, 7]. Key point in these models is a criterion of ductile crack propagation. Except for TCM, other models are based on crack-tip opening angle (CTOA). The CTOA criterion was originally applied to analyze quasi-static ductile crack extension [8, 9]. Fairly constant CTOA was observed during crack extension. Therefore, the critical CTOA or CTOA resistance value was regarded as material parameter characterizing ductile crack propagation. Attempts were made to measure CTOA resistance values for a dynamically propagating crack using laboratory-scale test pieces, including drop-weight tear test (DWTT) [10, 11]. While fairly constant CTOA resistance value was obtained, its dependency on crack velocity was recognized. If a constant CTOA resistance value is observed, it is meaningful to use the CTOA criterion in the numerical analysis models. But, a benefit of using the CTOA criterion might be reduced if CTOA resistance value depends on crack velocity. Most of the measurements of CTOA resistance value were made using laboratory-scale test pieces, although special loading system was utilized to realize high crack velocity in such small test pieces [10, 11]. Even if a constant CTOA value was obtained in such small-scale test piece, crack length is short and crack velocity might not be high enough to reproduce dynamic crack propagation in actual high-pressure gas pipelines. A. Kobayashi, et al, conducted small scale pipe burst test and measured CTOA using a high-speed camera [12]. They observed a crack velocity dependency of CTOA resistance value. However, they machined a groove on the pipe wall. It was presumed that the CTOA resistance value might be changed whether the crack propagates along the groove or not. To the authors’ knowledge, this is the only published data of measuring CTOA resistance value for a dynamically propagating crack in pressurized pipe. From this standpoint, the authors conducted a burst test of a pressurized pipe and measured CTOA, together with DWTT. 2. Experimental Procedures

2.1. Test pipe

Table1 shows grade and configuration of the pipe tested in the present study. Table 2 shows tensile, impact and DWTT properties of the pipe. Table 1 Pipe configuration Grade Outer diameter[mm] Wall thickness[mm] JIS STPG370 355.6 9.5

Table 2 Material properties

Tensile test

Charpy impact test

Press notch DWTT Absorbed energy [J/ mm 2 ] 3.5

Tensile strength [MPa]

Upper shelf absorbed energy [J]

FATT [ ℃ ]

Yield stress [MPa]

368

440

180

-25

2.2. Drop-weight tear test

Drop-weight tear test was conducted. Fig.1 shows experimental set up and configuration of press-notched DWTT specimen . DWTT machine’s capacity was 6000J, with 200kg mass and 2.8m drop height. As shown in Fig.1, high speed camera, PHOTORON SA-1.1, with 13,500 frames per second was equipped to measure deformation of the specimens and crack propagation dynamically. Two tests were conducted at ambient temperature.