PSI - Issue 57

Yevgen Gorash et al. / Procedia Structural Integrity 57 (2024) 611–617

615

Y. Gorash et al. / Structural Integrity Procedia 00 (2023) 000–000

5

Fig. 5. Typical small welding pore with Ø = 75 µ m found on fracture surfaces of USF testing specimens cut from weldments.

Fig. 6. Typical big welding pore with Ø = 470 µ m which is 640 µ m away frow surface of USF testing specimens cut from weldments.

4. Analysis of results

Based on the summary of USF testing results for S275JR + AR welds illustrated in Fig. 4, it is obvious that the welding porosity has the defining detrimental e ff ect on the fatigue life of weldments. An absolute majority of the weld samples have a fatigue life significantly lower than the parent material. Just a few weld samples tested in the stress range of 340-360 MPa surpass the parent SN curve with slightly longer fatigue life, which can be explained by the absence of “big & medium” pores and the location of the crack origin on the surface. Another important observation is that detrimental e ff ect of the corrosion on the fatigue life is stronger than the e ff ect of welding porosity on the fatigue life, with a majority of the pre-corroded samples having a crack origin from a corrosion pit on the surface. The obtained fatigue data points for both polished and pre-corroded surface condition are scatted quite uniformly around the semi-logarithmic diagram in Fig. 4 proving a random nature of the damaging e ff ects caused by both welding pores and corrosion pits. The damaging e ff ect of corrosion pits over the time was measured using the surface roughness machine Mitutoyo SV 600 as reported by Gorash et al. (2023). The detrimental e ff ect of welding porosity can be studied by measuring a detailed 3D topography using high-resolution X-ray nano-CT (computed tomography) system – SkyScan 2211 from Bruker. X-ray nano-CT scanning technology was used to create a map of pores in each sample including their relative location and size. The diagram in Fig. 4 summarises the USF testing of welded samples for S275JR + AR grade reporting the details on the samples with the following induced environmental e ff ect: 1) 40 samples with polished surface; 2) 6 samples pre-corroded for 2 weeks; 3) 12 samples pre-corroded for 1 month. There is a significant variation in both the size and location of pores which has a massive e ff ect on the fatigue life of each welded sample. The size of pores that initiated the crack growth and resulted in failure varies from 50 µ m to almost 1 mm, with a distance to the surface varying from the centre of the sample cross-section to a few micrometres or partly open pore. At the same nominal stress amplitude level, the fatigue life can vary from 1 million cycles for a