PSI - Issue 7

Ivo Černý et al. / Procedia Structural Integrity 7 (2017) 431 – 437 Ivo Černý / Structural Integrity Procedia 00 ( 2017) 000–000

436

6

from the mean values represented by the regression lines in Fig. 7. Schematic graphs of optimised probability density functions are shown in Fig. 8. Note that unlike X70 and X65 steels, where probability density corresponded to normal distribution, in case of X60 steel, the optimised function corresponded to Weibull 3-parametric distribution – Fig. 8 c) and d).

1.2

1

0.8

0.6

X70 X65 X60

0.4

Probabi l ity of fai lure (0.8 t)

0.2

0

20000 30000 40000 50000 60000 70000 80000 90000 100000 Number of cycles

Fig. 9. Results of simulations of failure probability of evaluated pipes made of different pipeline steels with the specific initial defect

It follows from Fig. 9 that significantly higher care and attention has to be focused on the pipe made of the steel of highest strength parameters – X70, where 10% failure probability occurs already after less than 25000 pressure cycles. Somewhat better situation concerns the X65 steel, where the 10% failure probability corresponds to almost 30000 cycles. On the contrary, the largest inspection intervals are possible for the X60 steel with the 10% failure probability after 43000 cycles. The tendency is consistent with general knowledge from the literature, where high strength high grade steels have a good resistance to fatigue crack initiation process from the smooth surface, but lower resistance to FCG or also to stress corrosion cracking, e.g. Luke et al. (2009), Černý and Linhart (2004). Jiang and Chen (2012) published results of FCG measurement in X60 and X70 pipeline steels, which are quite consistent with FCG results described above in this article. The authors also observed higher FCG rates in case when X60 steel was strengthened. Similar dependence for a structural carbon steel was described in Černý et. al. (1993) – higher FCG rates in case of higher static strength. Different crack closure values connected with different proof stress or strength and also material microstructure could be the factors affecting FCG rates. Eventually, similar phenomenon occurs in other steels or even non-ferrous alloys, when fatigue endurance limit of smooth material not containing defects is directly proportional to static strength – Zerbst et. al. (2013) or Fajkoš et. al. (2014), but concerning resistance to FCG, the dependence is contrary. So this is likely the reason, why the X60 steel is the most favourable from the three types of steels, if crack-like defects, which could grow, are detected and inspections intervals are to be specified. 4. Conclusions On the basis of recent experimental programmes aimed at an evaluation of crack-type defects indicated during high pressure pipeline internal inspection and at evaluation of fatigue crack growth (FCG) rates in high pressure pipeline steels X60, X65 and X70, probabilistic simulations of failure of pipes made of different steels and containing a model crack of 2 mm depth and 20 mm surface length was performed using ALIAS HIDA software developed during HIDA Applicability project within European framework programme. The main results can be summarised as follows: − Though most of the cracks likely occurred already during manufacture of the pipe, fatigue pressure test