Reliability analysis of Tension-Leg Platform Tendon with Respect to Fatigue Failure under Environmental Condition of Caspian Sea

The primary objective of this paper is probabilistic quantification of the fatigue life of tension-leg platforms (TLP) using reliability methods. The need for such methods stems from the significant uncertainty in the loads exerted on offshore structures. The scope of this paper is limited to the study of fatigue in TLP tendons. For this purpose, nonlinear time-history of force response of the TLP tendon under random-wave load is computed via MOSES software and the damage due to fatigue is estimated in accordance with the Palmgren-Miners rule. Assuming a Rayleigh distribution for stress variation and eight different sea states, the ultimate fatigue damage is computed by accumulating the damage over all individual sea states. This cumulative damage enters the limit-state function that is based on the Palmgren-Miners rule. Prevailing sources of uncertainty in this problem are those in the estimation of fatigue stresses, fatigue strength, and the Palmgren-Miners rule. Finally, reliability analysis is carried out for four different service lives using the first- and second-order reliability methods (FORM and SORM) and Monte Carlo sampling. The results indicate that FORM computes the probability of failure sufficiently accurate. It is concluded that the probability of failure increases drastically with the service life. The importance vector from the sensitivity analysis in FORM reveals that the model error is the most influential source of uncertainty on the probability of failure.