PSI - Issue 57

Yann Chevalon et al. / Procedia Structural Integrity 57 (2024) 633–641 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

639

7

Results are presented in the form of NS instead of an SN. The thinking behind that is because the number of occurrences is an input parameter driven by oceanographic data and cannot be modified (contrary to an overhead bridge crane where the number of cycles can be chosen for example). This has been done for a more intuitive reasoning and both presentations, SN or NS, are reversible because a power law is a bijective transformation. For each set of similar test parameters, a statistical analysis was conducted, and a design value was obtained by subtracting 2 standard deviations to the mean value for each set of identical test parameters. The result of these design points is represented as the dashed orange curve on figure 8. One can notice that tests on full thickness samples follow a similar trend to machined samples, but with an offset. After being checked experimentally, the SCF would be calibrated to reduce that offset (literature being based on SCFs for metallic material). One can see from the figure above that we obtain 2 trends: one for the tests on machined samples, and one for the tests on full thickness. To add conservatism during design phase, the slope of this power law is decreased. This impacts the admissible number of cycles for small curvature ranges – which represent the bulk of the fatigue damage. The SN curve would therefore have the following shape: =10 (( ∆ ) ) (1) with the Young modulus at the temperature the test is done, 10 the intercept value, the stress concentration factor, and the maximum number of cycles during design. 2.6. Experimental validation of the design curve On the figure 8, all points are test results. Though, some of the test points were used to validate the design curve. This is illustrated on the figure 9 hereunder:

Fig 9 Experimental test results circled in dashed black used to validate the design curve for HDPE

As it can be observed, test points are above the design curve. This validates the design curve for the polymer HDPE Orange. 2.7. Fatigue damage and evaluation with the SN curve. Because flexible risers are designed for a lifespan of 20 years or more, exposure to multiple temperatures must be accounted for. This is done with a pro rata temporis of the fatigue damage for the different service temperatures and shut down events. In order to get the total fatigue damage, a sum of all the fatigue damage at each temperature is done and compared to the fatigue damage obtained with the reference curve (orange curve on figure 8).