PSI - Issue 13

Letícia dos Santos Pereira et al. / Procedia Structural Integrity 13 (2018) 1985–1992 Author name / Structural Integrity Procedia 00 (2018) 000–000

1990

6

the discussion regarding ductile fracture triggering and control. First, in the laboratory specimens, an interaction between the crack plane and the hammer takes place and needs to be corrected; Second, the different loading modes, and the different thicknesses between Charpy and DWTT, cause a strong difference in local (near the crack tip) stress fields and triaxiality. For example, DWTT presents, in average and considering the evaluated thicknesses, 50% more stress triaxiality than Charpy for the analysed X80 steel – it means less plasticity spread ahead of the crack for the DWTT and a different energy distribution in each geometry. Consequently, the ductile fracture micromechanism behaves in a different manner for varying geometries and such effect must be taken into account. In this context, stress fields were evaluated in details, including stress traxiality. Such analyses, whose details can be found in Moço (2017) and Pereira (2017), can support the discussion that follows regarding the limitations and potential of Charpy and DWTT specimens for the quantification and description of energies involved in ductile crack Crack extension in real pipelines usually reach a steady state before arrest can take place. Consequently, to be able to quantify energies associated with crack propagation (see. Fig. 1(a) ), it is interesting to verify if Charpy and DWTT geometries are able to reach stable crack propagation before final failure. Figure 4 presents the variation of absorbed energy ( dEnergy/da ) versus crack size ( a ) for Charpy ( Fig. 4(a) ) and DWTT ( Fig. 4(b) ). After crack initiation takes place and the specimens bend, in both cases a stable propagation region can be identified. However, it can be realized that in the case of Charpy specimen the steady state crack propagation is very limited - it does not have enough remaining ligament and when the stable propagation is reached, the crack tip lies in the highly compressive region near the hammer-specimen contact. In the case of DWTT, in its turn, the steady-state propagation takes place along several millimeters of the remaining ligament, favoring the study of the energies related to the ductile fracture process related to a running crack. To support the energy analyses and favor comparability, the domains illustrated by Fig. 3(b) were evaluated during crack initiation and during steady-state propagation. extension and crack arrest. 4.2. Steady state verification

(a)

(b) Fig. 4. evolution of the absorbed energy rate as a function of a (a) for the Charpy specimen, (b) for the DWTT specimen.

4.3. Energy analysis Based on the explained domains and understanding the difference between the energy for crack initiation and energy for stable crack propagation, the two geometries could be investigated using both GTN and XFEM damage models, providing the energies absorbed for each phenomenon for the GTN damage model. All details can be found in Moço (2017) and Pereira (2017) and selected results are presented here, since trends for XFEM are similar as explained below. It is useful to recall Fig. 1(a) for a better comprehension.