Session: 02-06-02 Reliability of Mooring and Riser Systems 2

Paper Number: 81465

81465 - Analysis of Chain Fatigue Test Data for New and Used Mooring Chains 

DNV is currently running a Joint Industry Project, “NorMoor JIP”, which is now approaching the end of phase 3. In this phase the fatigue limit state of mooring lines with studless chains for permanent units has got the main attention. Phases 1 and 2 was covering the ultimate and accidental limit states respectively. The JIP has participants from energy companies, engineering companies, rig-owners, manufacturers and marine authorities. It is a global study covering Gulf of Mexico, Northern Europe and Brazil waters.

Our motivation for initiating a study on mooring line reliability was that all the global standards (API, ISO, DNV GL, others) are mostly based on work from late 1990s, when frequency domain analysis was prevalent. The reliability level implied by these regulations is not known, and we also see that the mooring standards are interpreted and applied differently.

Thus, there is a need for a mooring design code with a consistent analysis methodology and with safety factors that are in line with current more advanced analysis methodology and calibrated at an appropriate target reliability level. This is achieved through reliability-based calibration for a range of different units, mooring systems, water depths and geographical locations.

The focus in the present paper is on the chain fatigue strength.  NorMoor JIP has obtained access to a large number of tension-tension fatigue test data for studless chain of various sizes, both new and used. The tests are carried out at different mean load levels and load ranges, and provide a unique opportunity to study how the fatigue strength is influenced by mean load and degree of corrosion. Considerations have also been made to obtain consistency in the test results for tests with different number of links in the test sample.

The main result from analyzing the test data is an S-N based fatigue strength formulation, where the nominal stress range versus number of cycles to failure is given as a function of nominal mean stress and corrosion category. The data support keeping the slope of the S-N curve at m=3, hence, logA is fitted to the data as a function of mean stress and corrosion category. The overall fit to all data has been checked for sub-groups of data such as new chain, used chain of different corrosion category, different chain diameter and different mean load. Prior to performing the fit, the test data were adjusted to represent number of cycles to failure for the weakest out of 6 links. This adjustment was implemented for tests with number of links different from 6 and is based on extreme value statistics. For corrosion, seven corrosion categories are used, based on guidance provided by the LifeMoor project at Sintef/NTNU /??/. More work is needed to more precisely specify the categories. The formulation proposed uses nominal stress based on the reference diameter for new chain. The strength formulation from the present paper is used in the structural reliability analysis and code calibration presented in an accompanying paper at this conference, OMAE2022-81463.

Presenting Author: Limin Yang DNV

Authors:

Torfinn Hørte DNV
Limin Yang DNV
Siril Okkenhaug DNV