This paper waspresented at the 18th.Annual OTC inJ-loiJston, Texas, May 5-8,1986. The_material is supLeCtIlf2Prrection by the author. Permission to copy IS restricted to an abstract of not more than 300 wQ[Qs,A model is described for the calculation of wave loads on pile template offshore platforms in random directional waves. The model incorporates drag and inertia coefficients measured in laboratory and field. The force coefficients vary with amplitude to diameter ratio and surface condition of each member. A method is described for estimating the appropriate amplitude for a member in random waves and current, so that time varying coefficients are not required. Random directional waves are simulated using a fast Fourier transform simulator and linear wave theory. The linear wave theory kinematics are modified in order to produce reasonable velocities near the free surface. Conditioned simulation of forces on the Exxon Ocean Test Structure and the Shell COGNAC platform are compared to measurements on these platforms in severe storms. The conditioning helps eliminate much of the random variability and permits the comparison of wave force time series rather than probability distributions._

An experiment is described which was designed to investigate the forces normal to an isolated cylindrical member of an offshore structure. The presence of random directionally-spread waves, current, marine growth and structural motions were all modeled. This complex environment was simulated in the Maritime Dynamics Laboratory of SSPA Maritime Consulting AB (Swedish Maritime Research Centre at the time) by moving a 1 meter diameter instrumented cylinder along prescribed paths. Reynolds numbers over 2 × 106 and amplitudes of oscillation up to 15 meters were studied. It is observed that interaction of the cylinder with its own wake is extremely important. Strong vortex shedding effects throughout all of the experiments were evident. Large transverse forces were observed at Reynolds numbers over 106 " The Morison equation proves to be a good model of the forces. Comparisons are made with data from the Exxon Ocean Test Structure. INTRODUCTION Prediction of hydrodynamic forces on welded tubular structures in the ocean continues to be an area of active research. The force coefficients must be determined empirically due to the intricacy of the flow around a bluff body. Analytical models (l) of such flows are intractable and as yet unproven. The force coefficients for a fixed structure in storm-driven seas are not fully understood; this can be attributed to the difference between the regular harmonic fluid motions used in the laboratory and the random multidirectional motions found in the offshore environment. Reproducing the complex environment of oscillating flows in three dimensions, with mean flows due to currents and simultaneous structural motions at scales appropriate for design is extremely expensive. Developing theoretical tools for such problems has also proven elusive. Researchers have therefore concentrated on understanding simpler flows. It has been necessary to make in situ measurements and extrapolate these and laboratory results to the design environment. Hogben et al. (2) cite some 17 experiments relevant to the study of forces on cylinders in time varying flows. A recent review of the field is provided by Sarpkaya (3). Most laboratory experiments have been restricted to oscillations in one direction, and most of these experiments do not achieve scaled design conditions. Older field experiments, such as Wave Force Project II (4), did not have the benefit of measured fluid motions so that wave theories were necessary to deduce fluid motions and then force coefficients. Analysis of Ocean Test Structure (5, 6) data has determined force coefficients which are significantly larger than those previously observed in the field, especially for barnacle roughened cylinders. The Reynolds numbers and Keulegan-Carpenter numbers observed during OTS were somewhat less than those used in extreme wave design. And some differences in the force coefficients may exist at prototype conditions. As with all field experiments, the environment could not be specified for the OTS so that some conditions of interest were not observed. For instance, Reynolds numbers over 106 were not reached. And no measurements in simultaneous large current and waves were made.

The extreme responses of a turret moored tanker are sensitive to non-aligned wind, wave and current conditions. Such conditions commonly occur in the Gulf of Mexico during the passage of the eye of a hurricane. Conventional design practice often relies on a collinear or, at best, a "guessed" non-collinear combination of 100-year environmental return period wind, wave and current conditions. Hence there is a need to derive response-based design criteria, i.e. that particular combination of wind, waves and current which most likely yields the 100-year return period response. The long term response characteristics of a turret moored tanker in deep water Gulf of Mexico conditions are investigated through the use of a comprehensive hurricane hindcast database. The effects of turret location and wave spreading are considered. The 100-year long term responses are compared against the short-term 100-year design responses derived from a 100-year hurricane design analysis. Response-based design criteria are then derived. Introduction Turret moored tanker based FPSO systems are widely used in many deepwater areas. Conventional design of such systems often relies on the assumption of a design storm event comprised of a collinear (or at best a guessed non-collinear) combination of 100-year environmental return period wind, waves and current. However, it is well known that the extreme responses of a turret moored tanker are sensitive to non-collinear wind, waves and current (Refs. 6-8). With few exceptions, the effect of non-collinear wind, waves and current has received little attention. Such events commonly occur in the deepwater Gulf of Mexico during the passage of the eye of a hurricane. The resulting effects on the motions and mooring line tensions may be significant as such systems have a natural tendency to weathervane, i.e. align themselves against the prevailing direction of wind, waves and current. This then poses the question of how well the "conventional" design recipe really works for these systems in significantly non-collinear environments. In order to address this problem the actual long-term response characteristics need to be investigated and the 100-year return period responses need to be derived. Response-based design criteria may then be stipulated to capture specific response characteristics, e.g. the 100-year maximum offset storm is that particular combination of wind, waves and current that most likely yield the 100-year return period offset. Notice that the associated wave height, wind speed and current speed for such a non-collinear design event may well be lower than those normally referred to as 100-year return period environmental criteria. Also, the 100-year return period offset storm, say, is merely intended to estimate the 100-year offset while other responses (e.g. roll or mooring line tension) should be ignored. Theory The long term Gulf of Mexico environment is described by means of a hurricane hindcast database of 35 storms over an 85 year period since 1900 (Ref. 3). The original database contains some 240,000 records with the hourly values of wind, waves and current parameters.

No abstract

No abstract