The study started with the analysis of different types of mechanic-welded frames designed and manufactured by KIMUA, for sea transport of heavy cargo. From these case studies, we conducted a deeper analysis of the fatigue life in welded joints, considering the Residual Stresses (RS). The results of this study were presented in the 34th Encounter of the Spanish Fracture Group (GEF), which has been celebrated recently in Santander, Spain.
The main objective of the work presented in Santander was to compare the fatigue life prediction in High Cycle Fatigue (HCF) regime in three different cases. Applying the DNV GL IV-7-2 standard, which does not take into account the RS; considering yield stress as RS value, and determining the RS value by a numerical model developed by the Structural Mechanics and Design team of MGEP. This model has been experimentally validated and presented in previous works.
Due to the difficulty in the theoretical or experimental determination of the RS, they are usually not considered in the fatigue life prediction. This results in oversized designs, with the corresponding increase in material, time and transport costs.
Moreover, the difference in life prediction can be observed in a recent work (Figure 1), where we published the comparison of the fatigue curves for welded joints given by several standards with experimental results.
Figure 1. Comparison of fatigue life prediction by several standards and experimental points. Published in A. Lopez, et al., Comparación de diferentes métodos de predicción de vida a fatiga de un componente soldado a tope y su validación experimental, GEF31, El Escorial, 2014.
In the work developed, first, we determined the RS pattern in function of the plate thickness. For the analysed case, the RS value is around the 87% of the yield strength of the material. Subsequently, we estimated the fatigue life by the three proposed methods.
The results obtained show the tight restrictions imposed by the standards because of not considering the effect of the RS, in HCF regime (figure 2). In fact, according to the DNV GL IV-7-2 standard, the endurance limit of these joints is 25 MPa; while considering the RS the endurance limit increases to 77 MPa for the yield strength approach and to 93 MPa for the RS prediction model. In addition, it was observed that for plates with thickness less than 50 mm, the yield strength approach can be considered valid. For those thicknesses, the real RS value is close to the yield strength, so the work needed to carry out the numerical prediction can be avoided.
Figure 2. Comparison of fatigue life prediction obtained in this study.
The use of these methods for fatigue life prediction considering RS could lead to a reduction in the quantity of material needed in the welded joints, an important factor in the case of big mechanic-welded structures as the ones developed in KIMUA. Thus, KIMUA and the Structural Mechanics and Design team of MGEP will continue working together in order to generate new technical knowledge, with the aim of developing more efficient processes that enable a reduction of costs of material, manufacturing process and transport, for the benefit of our clients.
- M. Larrañaga, J.A. Esnaola, I. Llavori, U. Etxebarria, I. Oreja, I. Alzate, L. Martin, Predicción de vida a fatiga de una estructura mecanosoldada para sector off-shore teniendo en cuenta las tensiones residuales, 34º Encuentro del Grupo Español de Fractura (GEF34), Santander, Marzo 2017.
- Lopez-Jauregi, J.A. Esnaola, I. Ulacia , I. Urrutibeaskoa , A. Madariaga, Fatigue analysis of multipass welded joints considering residual stresses, Int. J. of Fat., Vol. 79, pp. 75-85, 2015.
- J. A. Esnaola, I. Ulacia, D. Ugarte, A. Lopez-Jauregi, I. Torca, J. Larrañaga, Analysis of the fatigue performance of thick T-joint samples considering residual stresses, ASME IMECE, Phoenix (USA), 2016.