Creep behavior of in-service flexible flowline polyamide 11

Abstract

The high cost (material, service, and production loss) involved to substitute a condemned flexible pipe whose pressure sheath has reached its theoretical preconized service life has motivated this study. Therefore, the main objective is to propose a constitutive equation for in-service aged polyamide 11 (PA11) describing the creep behavior as a function of temperature, stress level, and Corrected Inherent Viscosity (CIV), this latter parameter representing the level of material degradation due to hydrolysis. The constitutive equation may be employed for gap spanning analysis and also to subsidize the decision to extend the operational life of flexible pipes that have experienced more severe conditions or have been used for a longer time than designed. The current models to assess the remaining life of the sheath are based only on a single property decay based on corrected intrinsic viscosity (CIV) curves obtained from laboratory tests. To compare the result from the life-prediction model in use and the material mechanical behavior, an experimental campaign was performed using polyamide 11 (PA 11) samples retrieved from a 6″ gas production flexible flowline, which theoretically had reached a full-damaged condition after nearly 3 years operating at higher than specified temperature (80 °C). Dog-bone geometry specimens were machined from the internal, intermediate, and external layers of the flexible flowline pressure sheath. Once polymers are excellent thermal insulators, it was assumed that the material operated under different temperatures within the thickness and, therefore, presents different degradation degrees. CIV, tensile, and creep analyses were performed, confirming that the behavior is different for each region within the thickness of the pressure sheath. Differential scanning calorimetry (DSC), thermogravimetry analyses (TGA), and dynamic thermomechanical analysis (DMA) were performed to comparatively characterize the degree of crystallinity, amount of extractables and morphology of each section. A creep behavior model considering the gradient difference in the material is proposed. It is concluded that aging is different across the liner thickness, and the PA11 creep behavior may be expressed as a function of the CIV, temperature, and stress.