Methods for Coupled Analysis of TLP's

Abstract

ABSTRACT This paper investigates means by which the customary methods used to model TLP's in the frequency domain may be improved through more detailed consideration of tendon dynamics. The objective is to improve upon the method of modeling tendons as simple equivalent springs, and thus to improve the accuracy of deepwater TLP analysis. Two approaches are addressed:a technique is derived whereby a non-linear, finite-element program with Morison-type hydrodynamic loading can be extended via impulse-response considerations so as to include forcing derived from a radiation-diffraction model.Frequency-dependent equivalent spring coefficients are derived that can be used with a frequency- domain motion program. Derivation of the methods, application to a particular test case, and a comparison of results obtained by several methods are presented. Significant differences between the analysis results were found. The differences can have important implications for design. INTRODUCTION During the TLP design process, global performance analysis must be conducted to provide fundamental force and motion data for sizing or selecting:Main particulars of the hull (including deck height)Tendon system particularsRiser system particulars. Overall TLP design, often referred to as lLP Global Design, is generally governed by the following responses:minimum tendon tensionmaximum tendon tension (and stress)maximum motions (offsets)tendon fatigue lifeminimum deck clearance. Each of these quantities is the sum of static and dynamic components. Dynamic components are in turn the sum of high-frequency, wave-frequency, and low-frequency (slowly-varying components). Our main concern in this study is the computation of the wave-frequency dynamic component. The proportion of the total response due to the wave-frequency component is different in each response. According to our experience, the proportions are:Table (Available in full paper) Accurate computation of wave-frequency dynamic components is therefore seen to be important, but is clearly not the whole story. Generally speaking, tension dynamic amplitude is the most important of the above quantities, since minimum tension generally governs a design. Wave-frequency surge, which is often the first thing compared when considering analysis methods, is of decidedly less importance to design. Further, the above implies that comparison of analytical methods should be made on the basis of their impact on total response, rather than just on comparison of response amplitude operators (RAO's). The current practice in TLP global analysis relies on frequency-domain dynamic solutions. It may be stated that programs suitably model hull first-order hydrodynamic effects, but, beyond some limits, cannot fully model the tendon's restoration effects. These limiting conditions are directly related to the flexibility and mass of the tendons and are a result of non-linear effects (Fylling and Larsen1). Such effects become more pronounced with increasing water depth and result from:tendon curvature due to distributed loads: gravitational, buoyancy, and current/wave action all tend to give the tendon a catenary shape which effect vertical and horizontal stiffnesstendon distributed mass and damping effectstendon rotation effects at the upper flex jointplatform setdown as influenced by the above and by tendon stretching.