Papers by Keyword: Steel Weldments

Abstract: Scanning electron microscopy (SEM) observation of the fracture surface of CT specimens was carried out at CEMUP in the context of a FEUP R&D programme for the characterization of fatigue crack growth (FCG) of base material and welded steel CT specimens, carried out under constant load range, reported in a companion paper. The inclusion of specimens of welded material in the FCG characterization programme justified the need for relatively large size specimens. The thickness of the specimens (B=32mm) was approximately identical to the thickness of the structural application of interest, since with thinner specimens the possible effects of residual stress fields would be at least partially lost. The CT specimens were designed according to ASTM E647 standard. The FEI QUANTA 400 FEG / EDAX Pegasus X4M available at CEMUP was used for the SEM observations, carried out in the fatigue fracture surfaces of base and welded (HAZ) steel. Both in base material as well as in welded specimens, it was found that there are regions of the fracture surfaces where no striations are observed, mixed with other regions (patches) with striations. Some regions are featureless (ropey appearance). It was not possible to find a preferential orientation of the striation spacing, that might be directly related with the macroscopic direction of crack propagation; on the contrary, the regions presenting striations seem to have a mainly random orientation. This fact contributes to the absence of correlation between the distance between striations (s) measured using SEM, and the macroscopic fatigue crack growth rate da/dN. Only for substantial values of the crack length/specimen width (a/W) ratio a reasonable agreement between s and da/dN is found. The relationship s/(da/dN) presents values that decrease up to approximately one, as a/W increases. For low values of a/W, s is one or two orders of magnitude greater than da/dN. In the specimens analyzed, fatigue crack growth rates below approximately 2 to 3 x10^‑7 m/cycle are associated with approximately constant s values. This border is over the transition between regions I and II of the typical sigmoidal log-log plot of the da/dN vs. DK (I corresponding to the near threshold regime and II to the Paris law regime), and this fact should be accounted for in failure analysis involving SEM of fracture surfaces. Finally, the distance between striations s proved to be insensitive to the presence of important residual stresses in the case of the welded specimens.

Abstract: This paper reports on the use of the Modified Wöhler Curve Method (MWCM) applied along with the Theory of Critical Distances (TCD) to estimate fatigue lifetime of steel welded joints subjected to both uniaxial and multiaxial cyclic loading. In a recent work [1] we have proved that the above engineering method is highly accurate when calibrated by using standard fatigue curves characterised by a probability of survival equal to 50%. In order to better check its accuracy and reliability, in the present study our approach is systematically applied to a large amount of experimental data by calibrating it using standard fatigue curves having a probability of survival equal to 97.7%. This exercise allowed us to prove that the in-field application of such an engineering procedure results in estimates which fully comply, from a statistical point of view, with Eurocode 3’s recommendations. This result strongly supports the idea that our approach can safely be employed to perform the fatigue assessment of real mechanical assemblies, with the advantage over other existing methods that fatigue lifetime under any kind of fatigue loading can be estimated by simply post-processing linear-elastic Finite Element Models.