Unbounded Flexible pipe used for offshore fields development are usually composed of different layers of polymer and steel, each layer having a specific function during the product service life. This multi-layer characteristic enables to tailor the cross-section of the pipe to meet project-specific requirement, and optimize the cost of the product for each application. In particular, the main function of the thermoplastic pressure sheath is to guaranty the sealing of the product. The material and the thickness of this pressure sheath mainly depend on the pressure and temperature of the bore, and the design choice is driven by the creeping of the sheath in the interstices of the pressure vault: it must be limited with regard to sheath thickness reduction, as per API17J design requirement.

Consequently, when developing new material for pressure sheath application, the early prediction of the creep performance over the full range of the targeted application is crucial. For this reason, before any full-scale test, a test campaign is required to evaluate the creeping of the material on small-scale material sample. In this development context, the use of advanced finite-element simulation for predicting the creeping behavior is quite useful to amplify the benefit of tests campaign results, and to give additional information on material performances. As far as the modelling is validated by correlation with small-scale tests, the numerical tool is used to multiply virtual creep tests configurations.

This paper will focus on the numerical challenges for developing such creeping simulation, based on ABAQUS commercial software. Firstly, the identification of the viscoelastoplastic parameters for polymer material law will be presented. This material law is a nonlinear viscoelastoplastic model consisting of multiple networks connected in parallel. The number of parameters of such law is not limited, but a compromise between law precision and identification robustness must be found. Then, the correlation process between small-scale test and finite-element results will be detailed. In particular, the influence of the experimental protocol has to be determined. Finally, a sensitivity study of the most influent parameters, based on parametric FEA model, will be presented to highlight the benefice of such model. The benefice of such model does not only consist on correlation with small-scale test. As the material modeling is intrinsic, it is also possible to use the same law for studying the creep behavior on very different geometrical configurations.

This content is only available via PDF.
You do not currently have access to this content.