Issue 46

A. Deliou et alii, Frattura ed Integrità Strutturale, 46 (2018) 306-318; DOI: 10.3221/IGF-ESIS.46.28 310 Hardness measurement Vickers hardness filaments were carried out in the thickness on a cross section of the weld joint. The measurements are taken under 10 kgf of load using a ZWICK ZHV10 type Durometer, the step is 0.5 mm in the HAZ and 1 mm in the base metal and the weld metal. The hardness profile measured across the average diameter changes to a letter M as shown in Fig. 6. In addition, Tab. 5 shows the average values of the Vickers hardness for each region of the welded joint. The large values of hardness of HAZ zone related to the existence of acicular ferrite in its microstructure (presents the greatest value of the mechanical properties by virtue of the fine-grained and interlocked microstructures) [42-44], on the other hand, the low values are in the BM respect to welded joint can be related by the reverse cold deformation caused by the rolling (Bauschinger effect). The hardness results representing the weld seam showed a uniformity of values between base metal and seam weld sign a good property. Figure 6: Hardness profile of the welded joint. Region Average value BM 207 HAZ 230 WM 227 Table 5: Results of HV10 hardness along the weld joint F ATIGUE TESTS atigue cracking tests were performed on CT50 specimen’s thickness of 7 mm, in accordance with ASTME-E647 whose dimensions given by the Fig. 7. These tests were carried out in laboratory of materials and reactive systems (LMSR), Mechanical Department of University of Sidi Belabbes. These tests were conducted in ambient air for the same load ration value R at a nominal frequency of 20Hz in the three zones. Welded structures present a gradient of microstructure and mechanical behaviours from the weld metal to base metal. The study of fatigue crack growth in welded joints of pipeline steels, particularly in the heat affected zone, is difficult research field. In order to more clearly compare the evolution of the cracking rate in the three zones, we chose to represent them by their respective linear regression lines obtained from the experimental points (seven point method) on the linear parts of the curves. F