To guarantee the structural integrity of oil and gas transporting pipelines, a detailed analysis of the pipe’s structural response has to be performed. This is of particular importance for offshore applications. As large scale testing is costly and time consuming, one often relies on FE (Finite Element) modelling to accomplish, at least, part of this task. Properties that typically need to be evaluated are compressive strain capacity, collapse resistance and ovalization during reel-lay installation. Furthermore, it can be assumed that those properties are influenced by the pipe forming process, as pipe forming will change the mechanical properties and the level of anisotropy and will modify/introduce residual stresses. Therefore, a first logical step is to simulate pipe forming before evaluating the pipe’s structural performance, to account for these effects.
The reliability of FE simulations largely depends on the capability of the constitutive model to accurately describe the mechanical behaviour of the material being studied. Most commercial FE codes only offer combined kinematic-isotropic hardening models. Those models cannot capture the so-called cross-hardening effect and can therefore not predict the evolution of anisotropy during pipe forming. The present paper discusses the implementation and calibration of a more advanced constitutive model, more specifically the Levkovitch-Svendsen model, which accounts for isotropic, kinematic and distortional hardening. The model was implemented in Abaqus/Implicit through a UMAT user subroutine. An inverse modelling approach was applied to calibrate the constitutive model, whereby an extensive set of mechanical tests, involving cyclic tension-compression tests and tests with changing strain paths, was conducted.
To assess the model’s performance, it was used in two case studies. The first study focused on the evolution of mechanical properties during spiral pipe forming. The results show that the Levkovitch-Svendsen model allows prediction of the properties in both the transverse and longitudinal direction on pipe. When applying a kinematic-isotropic hardening law only, the properties in the longitudinal direction are significantly underestimated. In the second study, different hardening models were used to predict the evolution of ovality during reel-lay installation. It was observed that the predictions made with the Levkovitch-Svendsen model were much closer to the experimental values than the results obtained by means of a kinematic-isotropic hardening model.
