“…In the last specimens of the S5 series shown in Figure 9k-l, it is relevant that once the imperfections on both sides of the weld beads were removed, the fatigue failure corresponded to pores; surface pores located on the lateral side for the specimen with shorter In the last specimens of the S5 series shown in Figure 9k,l, it is relevant that once the imperfections on both sides of the weld beads were removed, the fatigue failure corresponded to pores; surface pores located on the lateral side for the specimen with shorter life (56,343 cycles) and internal pores for the specimen with longer life (207,400 cycles). The apparent scatter previously mentioned for these specimens is actually the effect of the location of the pores in the specimens, as reported in the literature for pores near the surface and for internal pores [37,38]. From the previous review on fractured surfaces, for the stress level examined, it can be inferred that in the fatigue life the type and shape of the imperfections are less important than the size, quantity and location, and there was a certain order in the fatigue lives of the specimens, from the lowest to the highest: a continuous very small surface imperfection along of the width of specimen; various surface imperfections covering the width of the specimens or imperfections near to the border that produce fatigue in the lateral side, but growing semi-elliptically; surface pores or pores near to the surface in the lateral side; one big or a few small surface imperfections that do not cover the width of the specimen, that growth as one semi-elliptical crack; a condition similar to the previous one but with very small imperfections and one or various internal imperfections (pores).…”