Abstract
In the oil and gas industry, fixed platforms are commonly applied in shallow water production. In-place environmental conditions generates cyclic loads on the structure that might lead to structural degradation due to fatigue damage. Fatigue is one of the most common failure modes of offshore structures and is typically estimated when dimensioning of the structure during design phase. However, in times when life extension of existing offshore structures is being a topic in high demand by industry, mature fields may represent an interesting investment, especially for small companies.
Concerning fixed platforms, composed mainly by welded tubular joints, the assessment of hot spot stresses is considered to predict structure fatigue. The estimation of welded joint hot spot stresses is based on the stress concentration factors (SCFs), which are given by parametric formulae, finite element analysis (FEA) or experimental tests. Parametric formulae may be defined as a fast and low-cost method, meanwhile finite elements analysis may be time consuming and experimental tests associated with higher costs. Given these different characteristics, each method is applied according to the study case, which will rely on the joint geometry and associated loads.
Considering simple joint geometries several sets of parametric equations found in the literature may be applied. On the other hand, the SCFs calculation of non-studied yet complex joints consider known formulae adapted according to the under load joint behavior and geometry. Previous analysis shows that this adaptation may furnish different results compared to those obtained by FEA. Furthermore, it is observed that even for simple joints the results derived from the different methods may differ.
Given their importance for the oil and gas industry, since they are the basis for the assessment of the fatigue life of welded tubular joints which may impact on additional costs related to maintenance and inspection campaigns, the estimation of SCFs must be the most accurate as possible. Therefore, this paper intends to investigate the differences between results derived from parametric formulae and different FEA studies.
