PSI - Issue 38

5

Yuichi Shiraishi et al. / Procedia Structural Integrity 38 (2022) 588–595 Author name / Structural Integrity Procedia 00 (2021) 000 – 000

592

3.2. Fatigue test In the existing structure, strain which was measured by gauge No.11 and 12 was increased after 80 thousand cycles loading. For these results, it is assumed that fatigue cracks were initiated from welded joints between the deck plate and the trough rib then. Furthermore, strain which was measured by gauge No.11 was decreased sharply, and gauge No.12 was broken after 0.5Mcycles loading. For these results, it is assumed that these fatigue cracks fractured a deck plate then. Fig. 7 shows fatigue cracks on the surface of the deck plate when fatigue test was ended (after 1Mcycles loading). These cracks propagated close to strain gauges pasted on the surface of the deck plate. Fig. 8 shows relationship between loading cycles and length of fatigue cracks which were initiated from welded joints between a trough rib and a cross rib. At first, fatigue test was conducted with angles on the existing structure as preventive countermeasure. Any fatigue cracks initiated from welded joints between a trough rib and a cross rib had not been detected still the specimen was loaded over 2.0M times. Thereafter, fatigue test was conducted without angles on the existing structure. When the specimen loaded 0.5M times after angles were removed from the existing structure, a fatigue crack was detected. The crack was initiated from a welded joint where was near gauge No.6 and propagated about 10mm to a trough rib then (Fig. 9). Angles were retrofitted to the existing structure again when the length of a crack was about 20mm. After retrofitting, fatigue cracks had not propagated. After angles were removed from it again, all fatigue cracks propagated. For these results, the post countermeasure effect of retrofit using angles was verified. From Fig. 10 to 13 shows fatigue cracks when fatigue test was ended. These fatigue cracks propagated to a trough rib from welded joints between the trough rib and a cross rib. Fig. 14 shows locations of core holes 40mm across for confirming presence of fatigue cracks which are initiated from welded joints between a deck plate and a cross rib. From Fig. 15 to 17 show inside of core holes. Fig. 18 and 19 show the surface of a deck plate around TRS when magnetic particle test was conducted for detecting cracks. Around gauge No.12 which were pasted on the existing structure, the fatigue crack was initiated from weld root and propagated to a deck plate along weld metal. Finally, the crack went through the deck plate. On the other hand, in a deck plate on the new structure, any fatigue cracks weren’t detected. The new structure was loaded 4.6Mcycles while fatigue test of 3 steps for the existing structure was conducted. However, any cracks weren’t initiated from the connection between a trough rib, a cross rib and a deck plate.

*DXJHV

&UDFNV

$ UXEEHU SODWH

Fig. 7. Fatigue cracks on the surface of a deck plate

Fig. 8. relationship between loading cycles and length of fatigue cracks which were initiated from welded joints between a trough rib and a cross rib