Prediction Methods For The VIV-induced Fatigue Damage In Deep Sea Risers

With the increased interest in the deep sea waters in the field of petroleum and natural gas production, VIV-induced fatigue damage in deep sea risers gradually becomes the common key point both in the industry field and in the academic field. Furthermore, it is necessary for riser design to carefully consider the VIV-induced fatigue damage.The present study is a part of the project"Theoretical and Experimental Study on the Key Dynamic Characteristics of Deep Sea Spar"supported by the Science and Technology Commission of Shanghai Municipality under Grant No.05DJ14001, and mainly consists of the following aspects:(1) A literature review of the related research works, including the prediction methods of the VIV response, cumulative fatigue damage theory, fatigue crack propagation theory, and the prediction methods for the VIV-induced fatigue damage, is presented, in which a description of the existing approaches, development procedure and applied conditions is introduced.(2) Prediction methods for the VIV in deep sea risers. The basic concept and parameters of VIV are introduced and the empirical models and CFD models for VIV prediction in deep sea risers are also provided. Furthermore, the prediction method for VIV by our group is highlighted and the VIV-indued fatigue damage prediction in the next step is based on this model.(3) Study on the prediction methods for the VIV-induced fatigue damage in deep sea risers based on the S-N curves. The VIV response is obtained through the VIV prediction program developed by our group, and the long term distribution of the alternating stress induced by VIV is assumed to be Rayleigh distribution. Combined the suitable S-N curves with the Palmgren-Miner linear cumulative fatigue damage rule, the prediction method for the VIV-induced fatigue damage in deep sea risers is proposed. Compared with the experimental data, a reasonable prediction method for the multiple mode VIV-induced fatigue damage is recommended for the engineering application.(4) Effect of various factors on the VIV-induced fatigue damage in deep sea risers. With the above proposed prediction method for the VIV-induced fatigue damage, effects of top pre-tension, inner fluid density, current velocity profile, outer diameter, riser wall thickness, and elastic modulus on the VIV-induced fatigue damage are studied in detail. Some valuable conclusions are drawn and can be referred in riser design and engineering application.(5) McEvily model based on the fatigue crack propagation theory. McEvily model which shows promising capability is provided in detail. Based on three group of fatigue experimental data, determination of parameter reflecting crack closure development in the McEvily model is studied through the nonlinear least squares curve fitting method. This is the basis of the modified McEvily model which can account for the load ratio effect.(6) Study on the relation between threshold effective stress intensity factor range,ΔK eff,th and load ratio, R .ΔK eff,th under different load ratios is commonly viewed as constant. However,ΔK eff,th is assumed to be variable for different load ratios in this paper, and the relation betweenΔK eff,th and load ratio is further studied based on the following three aspects: the simple model ofΔK eff,th proposed by Schmidt and Paris (1973), experimental results ofΔK eff,th based on the full crack closure model, and experimental data ofΔK eff,th based on the partial crack closure model.(7) Modified McEvily model. Though McEvily model can explain most of the fatigue phenomena observed in tests, some other problems can not be solved well. In the paper, modifications on the McEvily model are mainly the following two aspects: (a) the slope of the fatigue crack growth rate curve for different materials is viewed as variable, rather than constant; (b) bothΔK eff,th and the maximum stress intensity factor at the crack opening level, K op,max are regarded as variables for different load ratios. The experimental results are adopted to validate the modified McEvily model. This is the foundation of the prediction method for the time domain VIV-induced fatigue damage in deep sea risers.(8) Prediction method for the time domain VIV-induced fatigue damage in deep sea risers. Since the prediction method for the time domain VIV response is not available, the random data are employed to simulate the time domain response procedure. The modified McEvily model is adopted to predict the time domain VIV-induced fatigue damage. The predicted results not accounting for the load ratio effect are also provided. The corresponding comparisons highlight the requirement for considering the load ratio effect when the time domain fatigue life prediction is performed.(9) Summary and forecast. Summarizes all the work mentioned in the dissertation and proposes the future development trends of the related fields.The innovative research works in the dissertation are primarily represented as prediction method for the multiple mode VIV-induced fatigue damage, study on the relation betweenΔK eff,th and load ratio, modified McEvily model accounting for the load ratio effect, and application of fatigue crack propagation theory on the time domain VIV-induced fatigue damage prediction in deep sea risers, with the associated conclusions being drawn as follows:(1) For multiple mode VIV response, the maximum fatigue damage of all excited modes agrees well with the corresponding experimental results, and this method can be adopted for engineering application.(2) Top pre-tension, outer diameter, elastic modulus, and current velocity profile all have great effect on the VIV-induced fatigue damage. The larger the top pre-tension, the lower the VIV-induced fatigue damage. The VIV-induced fatigue damage increases along with the riser outer diameter. In order to reinforce the fatigue capability of the riser, the material with lower elastic modulus should be selected if possible. As the variety ranges of inner fluid density and riser wall thickness are very limited, effects of these two parameters on the VIV-induced fatigue damage are not significant.(3)ΔK eff,th should be viewed as variable under different load ratios.ΔK eff,th tends to increase as load ratio increases when the load ratio is lower than the critical load ratio; However,ΔK eff,th will decrease as load ratio increases when the load ratio is larger than the critical load ratio. Furthermore, the relation betweenΔK eff,th and load ratio is in accordance with Lorentz distribution in general.(4) The modified McEvily model proposed in the dissertation can account for the load ratio effect well, and can correctly predict the load sequence effect under the classical two step fatigue loading. The extended McEvily model shows the potential value for engineering application.(5) Comparisons indicate that prediction results not accounting for the load ratio effect are greater than the corresponding results considering the load ratio effect, and it is dangerous for engineering application to predict fatigue damage without load ratio effect. Therefore, it is necessary for the time domain VIV-induced fatigue damage prediction to reasonably account for the load ratio effect.