AN INVERSE PROBLEM FOR PARAMETER ESTIMATION IN A BEND STIFFENER SYSTEM

摘要

The top connection of the flexible pipe attached to the platform supporting structure is considered to be a critical area as it sustains the highest forces and often the maximum curvature in the riser system. Bend stiffener, a polymeric structure with conical shape, is employed to limit the maximum curvature of the riser at the uppermost connection, and protect it against excessive bending and accumulative fatigue damage. In this work, an inverse problem methodology is proposed for estimating unknown parameters in the bend stiffener system, based on a large displacement beam theoretical model combined with the Levenberg-Marquardt Method. The global mathematical formulation is used for nonlinear analysis of the riser/bend stiffener system considering linear elastic symmetric material. A case study is given considering simulated angle measurements in five monitoring positions to estimate two unknown parameters in the system, top tension and polyurethane Young's modulus. Monte Carlo method is employed to analyze the statistic properties of the estimated parameters with measurement errors. The effects of sensor locations and measurement error ranges on the accuracy of parameter estimation are investigated. It is shown that the proposed procedure can estimate efficiently and accurately unknown parameters in a bend stiffener system. The parameter estimation procedure can also be used to assess other mechanical parameters of the bend stiffener system by angle measure- ments in certain monitoring positions in realistic production systems.