Analysis of Low-Frequency Damping in Mooring Lines for Floating Production Units
In this thesis, the relationship between mooring line damping and system responses, due to varying environmental conditions, is investigated. The responses focus mainly on the low-frequency (LF) components and on the modelling capabilities of coupled and uncoupled analysis.
Moored vessels are exposed to time varying environmental loads, such as wind, waves and current, which give rise to large dynamic motions. The presence of mooring introduces a low stiffness and a high natural period in the horizontal plane. Due to second order difference frequency effects, one will get a low frequency resonant excitation from the waves. This effect is enhanced by the slowly varying wind force. The total dynamic response will therefore be due to a combination of the first order wave frequency (WF) motion and of the resonant low frequency (LF) motion. The LF motion requires an accurate assessment of the damping in the system, and mooring lines will in many cases constitute a large portion of the damping.
The dynamics of mooring lines can be analysed by different methods and one usually distinguish between frequency domain (FD) and time domain (TD) calculations, and between coupled and uncoupled analysis. The choice of method is usually a compromise between accuracy and computational effort, where a non-linear coupled simulation in TD will yield the most accurate results. The goodness of the results depends largely on the system, and approximate methods such as FD and uncoupled analysis must be handled with great care, as the accuracy of the results may be dissatisfactory.
Coupled and uncoupled analysis models of a turret-moored FPSO at 320 meters water depth are established, and calculations are performed in TD. Damping from mooring lines are estimated from coupled analyses and applied to the corresponding uncoupled model as linear vessel damping coefficients. Mean current forces are also estimated from coupled analyses, and applied as constant forces acting on the vessel. Model applicability and limitations are discussed through a comparison of coupled and uncoupled results.
The main case investigated is an extreme 100-year storm with head sea conditions. Parameter studies are also performed, in order to study the effect on the damping estimate and on the responses. Among these are a seed variation of the main case, a less severe condition with reduced peak period and significant wave height, an alternative current profile, an alternative heading, and a damping sensitivity study.
The uncoupled model show a good agreement with the coupled model for the less severe condition. Discrepancies become more apparent as the current velocity is increased, or as WF motions becomes more prominent. This is because the uncoupled model uses a linear coefficient to represent the damping in the mooring lines, while in the coupled model the correct non-linear damping contributions from the lines are automatically included. The discrepancies caused by introducing such a linear coefficient becomes larger as the damping level increases. Both current velocity and WF excitation will increase the damping contribution from the mooring lines.
Reducing the peak period and wave height results in a decrease in WF response and damping estimate. This confirms that WF motion is an important contributor to the damping in mooring lines. WF motion decreases both because of lower values in the first order transfer functions, and because the energy in the wave spectrum is proportional to the wave height and to the square of the peak period. Contrary to the WF motion, LF motion increases when the peak period is reduced, because of higher values in the drift coefficients. The increase in drift coefficient values outweighs the effect from a reduction in spectral energy.
Interestingly, a reduction in peak period increases both the resonant LF motion and the accuracy in the uncoupled results. The accuracy in LF motion is greater despite an increase in motion magnitude because the overall damping is reduced. Therefore, an increase in LF motion does not itself necessarily lead to an increased sensitivity, or a poorer representation of this motion in the uncoupled model.
Results from the sensitivity study show that a low uncertainty in the mooring line damping estimates may lead to large or moderate uncertainties in the LF response, and by consequence in the total response. Extreme response is affected through the uncertainties of the LF response, as LF constitutes a large portion of the total response. This confirms that the LF response is highly sensitive to mooring line damping. Accurate estimates are need for both LF and extreme response.
Place, publisher, year, edition, pages
Institutt for marin teknikk , 2014. , 115 p.
IdentifiersURN: urn:nbn:no:ntnu:diva-26498Local ID: ntnudaim:10768OAI: oai:DiVA.org:ntnu-26498DiVA: diva2:748307
Larsen, Carl Martin, ProfessorLone, Erling Neerland