In offshore jacket design, it has long been recognized that an accurate global structural model requires implementation of the effects of local joint flexibility (LJF). However, there is still no general method for implementing these effects accurately and efficiently without complicating the application of loads. The literature describes several techniques for determining LJFs using parametric formulas and implementing these in global models of a jacket structure. These techniques are simple but associated with uncertainties and a risk of compromising the accuracy of the global model. Alternative methods, such as the use of superelements, provide very accurate results but complicate the consistent application of external loads as well as postprocessing. This paper introduces a new methodology which is called the Correction Matrix Methodology. This allows the effects of LJF from detailed three-dimensional (3D) finite-element (FE) shell or solid models to be incorporated in a global beam FE model via a simple correction matrix. The effectiveness of the methodology is improved by using interpolation between a limited number of correction matrices. The new methodology provides exact results when correction matrices associated with the actual geometry are applied. When using the interpolation procedure, the methodology provides accurate results and computational efficiency when the database has been established. The Correction Matrix Methodology is a significant improvement of the conventional methods for modelling LJF and is currently being implemented in a general form for arbitrary joints in Rambolls Offshore Structural Analysis Program (ROSAP).

Tubular joints are complex geometries present in many offshore structures. Fatigue testing of tubular joints requires either downsizing of the dimensions of the joints or the use of a simplified geometry, to avoid prohibitive costs. For the development of a welding process optimized for the fatigue performance of tubular joints, one needs to use a representative sample: same material, same thickness, use of similar welding positions, same level of restraint as in a real structure, combined with the supplementary requirement of ease of manufacturing and testing. A novel geometry was developed to fulfil all these requirements. Then, different welds were produced using robot and manual welding in different positions. The fatigue tests proved to be very reproducible, and indicate a strong influence of the welding position on the fatigue resistance. For that reason, an optimized welding procedure and sequence was developed, in which the welds in the most fatigue sensitive locations are produced first in optimum condition, while the less sensitive parts are produced by subsequent welding in position.

Fatigue occurs in structures due to the stresses from cyclic environmental loads. Offshore environmental loads being highly cyclic and recurring in nature, fatigue analysis with high degree of accuracy is required for reliable and optimized design of offshore structures. The main aim of this paper is to automate the process of identification of fatigue critical tubular joints of an offshore jacket structure using deterministic fatigue analysis with emphasis on the Hot Spot Stress Range (HSSR), an important measure in estimating fatigue damage, calculated using three different approaches for each tubular joint. The first approach determines HSSR at the time of maximum base shear of the jacket, the second, by calculating the difference between maximum and minimum Hot Spot Stress (HSS) and the third, at all time-instants of the wave cycle. Thus fatigue damage and fatigue life of the tubular joints are estimated using the highest HSSR value and the joints with lower fatigue life are identified as fatigue sensitive joints. This ensures effective identification of critical tubular joints of the offshore jacket structure which needs detailed investigation or redesign for fatigue. The deterministic approach discussed in this paper is applicable to large jackets which contains more number of tubular joints where sophisticated fatigue assessment at the preliminary stage is computationally intensive and manual identification of fatigue critical joints is laborious.

This paper represents the condensed version of an original Master Thesis written in 2016 [1]. The main aim lays in the structural analysis of Tension Leg Platforms (TLP) as support structures for Floating Wind Turbines (FWT). Special focus is given to a methodology study of different static analysis methods for components of a FWT support structure by Finite Element Method (FEM). The analysis considers Ultimate Limit States (ULS) only. Based on a given open source TLP design [2], the support structure is fragmented in its components. Each component is categorized, based on whether it can be analyzed by parametric equations or by computational methods. For the components, verifiable by parametric equations, the relevant standards are examined. The analysis of the decomposed support structure shows that existing standards can be applied for many of the components. One specific tubular joint of the support structure is selected and analyzed by computational methods. Considering the complexity of Finite Element Analysis (FEA) for FWT, the performed methodology study compares and evaluates various simulation methods on how loads are applied onto the structure. The methodology study results show that for ULS consideration, all load driven simulation methods are generally applicable. In the prospect of fatigue analyses further developments and improvements are considered crucial. This especially includes a better coherence of the stiffness representation between global load simulation model and local structural analysis model.

This study evaluated, by time-domain simulations, the fatigue life of the jacket support structure of a 3.6 MW wind turbine operating in Fuhai Offshore Wind Farm. The long-term statistical environment was based on a preliminary site survey that served as the basis for a convergence study for an accurate fatigue life evaluation. The wave loads were determined by the Morison equation, executed via the in-house HydroCRest code, and the wind loads on the wind turbine rotor were calculated by an unsteady BEM method. A Finite Element model of the wind turbine was built using Beam elements. However, to reduce the time of computation, the hot spot stress evaluation combined FE-derived Closed-Form expressions of the nominal stresses at the tubular joints and stress concentration factors. Finally, the fatigue damage was assessed using the Rainflow Counting scheme and appropriate SN curves. Based on a preliminary sensitivity study of the fatigue damage prediction, an optimal load setting of 60-min short-term environmental conditions with one-second time steps was selected. After analysis, a sufficient fatigue strength was identified, but further calculations involving more extensive long-term data measurements are required in order to confirm these results. Finally, this study highlighted the sensitivity of the fatigue life to the degree of fluctuation (standard deviation) of the wind loads, as opposed to the mean wind loads, as well as the importance of appropriately orienting the jacket foundations according to prevailing wind and wave conditions.

The recommended practice DNVGL-RP-C203 on fatigue design of offshore steel structures has been revised a number of times since it was first issued in 2001. The 2016 revision includes additional information on a number of items asked for by the industry such as: Residual stresses in tubular sections made from cold forming, S-N data for subsea application, change in validity of S-N curve in seawater with cathodic protection, relation between surface roughness and requirements for roughness for coating, S-N curves and stress concentration factors for fatigue assessment of pipelines, relation between fabrication tolerances and S-N curves in different types of structural details and hot spot stress analysis for rainflow counting. The background for this is explained in the paper. In addition the paper includes some information on recommendations for fatigue design of single-sided tubular joints that was included in the 2014 revision of the Recommended Practice.

The use of jacket structures to support offshore installations has for a long time been a popular choice in places with appropriate water depths. In recent years the use of jacket structures as offshore wind turbine foundations has also attracted increasing attention and is becoming a feasible alternative to traditional monopile foundations. One of the key challenges in jacket design is optimizing tubular joints in terms of fatigue resistance. As it is not practically possible to include detailed FEM joint models in global jacket models designers are forced to look for alternative methods to obtain realistic joint representations. This is done by calculating influence factors (INF) and stress concentration factors (SCF) to be applied to simplified models of relevant tubular joints in global models in order to achieve a realistic force flow in the structure. One simple and widely used method is to apply parametric formulas, e.g. those suggested by Efthymiou. However, these approximating formulas have a fairly limited validity range. Therefore, on complex joint the most reliable way to determine INF’s is by setting up refined FEM models of relevant joints and then subsequently using the calculated factors in the global model. This strategy is computationally demanding and hence, very time consuming, as a new detailed FEM analysis of the tubular joint must be conducted for each step in the optimization process. The present paper demonstrates how this time consuming procedure can be avoided by use of artificial neural networks (ANN) trained to estimate INF’s on tubular joints. The neural network is trained on a pre-generated library of detailed FEM joint models and is then able to predict INF’s on joints that are not part of the library — and thereby providing a significant reduction in calculation time during the jacket/joint optimization process. The analysis is conducted on a typical joint on a three legged jacket structure. The joint is located on a jacket leg and has two incoming braces. Such a joint has a finite number of free design variables, e.g. chord diameter/thickness, brace diameter/thickness, brace angle, gap etc. Each of these free variables can be considered as a dimension in the joint design space. Having a sufficient number of FEM joint models in the library the neural network can be trained to recognize and predict underlying patterns in this design space. The method is demonstrated on a limited number of design variables but should easily be extended to cover all variables as the joint library is expanded to include all dimensions.

Fatigue crack growth at welded joints often propagates from as many as tens to hundreds of small weld toe cracks along the weld toe line in offshore welded structures. This paper will present a fatigue algorithm for modeling many small weld toe cracks propagating from a welded joint. Cracks usually initiate at the weld toe region of the structures and propagate as surface cracks at the stress concentration regions of the weld-toe line. The presence of such weld defects or crack-like flaws can have a severe detrimental effect on their fatigue life and fracture resistance. Currently, there is a lack of studies that considers the effects of multiple cracks and their distribution density in welded joints. This work focuses on the fatigue analysis and modeling of multiple weld toe cracks, specifically in T-butt joints. Fatigue crack growth prediction is usually determined by the stress intensity factor range and crack propagation rate through Paris law. To predict the Stress Intensity Factor (SIF) of a weld toe crack, the magnification ( M k ) factor was used. The M k factor is influenced by the size of the welded attachment, as well as the size and depth of the weld toe crack. Simplified solutions for practical prediction of M k factors were determined from 3D extended finite element method (XFEM) by modelling a semi-elliptical weld toe crack in a T-butt weld for cracks of different dimensions. The accuracy of the M k factor solutions was verified by comparison to HSE fatigue data on 16 mm thick tubular joints. The M k factor solutions were used to predict the growth of fatigue cracks using a model based on Paris Law and SIF solutions by Newman and Raju with plastic zone size corrections. Fatigue life was predicted for plates with and without attachments. It could be seen that the predicted life of a weld toe crack was severely reduced with the addition of a welded attachment. The model was extended to the multiple surface cracks commonly observed at the weld toe, where each crack is treated as independent, following established code procedures. The multiple cracks will coalesce as they propagate, until a single dominant crack emerges and fracture occurs. In this paper, the relationship between the fatigue life and the number and density distribution of the initial cracks was investigated. Fatigue life was predicted for plates with attachments with 1, 2, 10 and 100 cracks initially. The results show that as the number of cracks increases, the predicted fatigue life decreases.

The operation and maintenance cost of offshore wind turbine substructures contributes significantly in the cost of a kWh. That cost may be lowered by application of reliability- and risk-based maintenance strategies and reliability updating based on inspections performed during the design lifetime. Updating the reliability of a welded joint can theoretically be done using Bayesian updating. However, for tubular joints in offshore wind turbine substructures when considering a two dimensional crack growth and a failure criterion combining brittle fracture and material strength, the updating is quite complex due to the wind turbine loading obtained during operation. This paper solves that updating problem by using the Failure Assessment Diagram as a limit state function. It is discussed how application of the updating procedure can be used for inspection planning for offshore wind turbine substructures, and thus also for reducing the required safety factors at the design stage.

Acquiring the hot spot stress (HSS) in dangerous points of the structure (normally the welding seams of the structure) is the foundation of the offshore platform fatigue strength design. Because of the fact that it is very difficult and unpractical to install sensors in the welding seams, the most widely used method of acquiring the HSS currently is extrapolation, a method that obtains the HSS by extrapolating from stress of distant points. In order to figure out the feasibility of the method of extrapolation as well as the best ways to implement extrapolation in real practice, it is important to carry out numerical calculation of welded tubular joints in advance. To study the tubular joints, the finite element method (FEM) is the most widely adopted approach. In previous studies, shell elements are utilized to simulate tubular joints instead of solid elements concerning the difficulty. However, shell element simulation merely simulates the mid-surfaces of the tube walls so that shear stress in the direction of tube thicknesses are ignored, making the calculation of the stress concentration factor (SCF) inaccurate. In this study, the recommended ways of simulating the tubular Y- and T-joints are displayed and then the comparison of calculation results between different weld type, different mesh situations as well as different elements are given in order to bring up advice on numerical calculation for the study of extrapolation. The results of numerical calculation show that the extrapolation results are more accurate when we only consider the change of parameter α (chord length-to-chord radius ratio) and also, it is suggested to do the fitting in the area of 2t-4t away from the welding seam. These suggestions are verified by applying the method of extrapolation to a real jacket platform, which turns out that the fitting degree of extrapolation is very pleasant.

The Battelle structural stress method (BSSM) for fatigue life evaluation is examined for multi-axial stress states that develop due to uniaxial loading in welded structures. The resultant multi-axial stress state due to simple uniaxial loading is easily observed in common joint types such as a plate with a welded tube or a plate with an angled attachment. In these joint types, under simple loading, the stress distribution at the location of failure along the weld line shows significant in-plane shear stress (parallel to the weld line) as well as normal stress (normal to weld line). Although the fatigue data, as exemplified by the inverse slope of the S-N curve for the subject joints under uniaxial loading, are observed to be similar to that for normal-loading-mode dominant (Mode I) failures in welded joints, when only the normal structural stress is considered for these joints the predictions of both the fatigue failure location and the fatigue life using the master S-N curve approach are inaccurate because the in-plane shear stress plays a significant role in the development of the crack. The slope of fatigue data exhibited in S-N curves taken from weld fatigue data for resultant multi-axial stress state generated by uniaxial loading is different from multi-axial fatigue loading conditions for tubular joints as discussed in the recent work [OMAE2014-23459]. In this article, the fatigue behavior of welded joints with multi-axial stress states is evaluated using an effective equivalent structural stress range parameter that is formulated as a von Mises form of the combined normal and in-plane shear equivalent structural stress ranges. When the effective equivalent structural stress range parameter is employed, the fatigue failure location can be predicted correctly. It is also found that the cycles-to-failure data from the subject joint types are comparable with the master S-N curve for Mode I loading dominant behavior (inverse slope of 3.125). Therefore, the master S-N curve that was developed for Mode I failures can be equally applicable for fatigue life prediction for these joints by replacing the equivalent structural stress range with effective equivalent structural stress range on the ordinate axis.

Tubular joints are intensively used in off-shore structures for shallow waters. Depending on the sea conditions and the type of structure, the design can be fatigue driven. This is particularly the case for off-shore wind turbines, where turbulences are generating a fatigue loading. Any improvement of the fatigue performance of the tubular joint would be beneficial to reduce the weight and the cost of the structure. To assess efficiently the fatigue resistance of the tubular joint, a testing method has been developed based on the resonance principle. The complete circumference of the welded joint can be loaded, successively in the in-plane and out-of-plane modes at a frequency close to 20Hz. Finite element computations were used to investigate the feasibility of the concept. Then, an X-node was made and successfully tested to investigate the stress distribution along the weld. The experimental results were compared with finite element computations, giving a good agreement.

Fracture toughness of tubular welded joints is one of the critical factors affecting the structural integrity and reliability of offshore structures, such as platforms and subsea pipelines. The factors affecting the design fracture toughness of these structures are related to, both, the welding process as well as the chemical composition of the weld metal. The welding process in this application typically comprises of depositing weld metal in the tubular joints of varying thicknesses through series of weld passes. The number of weld passes required for welding these joints subjects the weld metal to repetitive cycles of heating and cooling. The effect of the thermal cycling introduces significant heterogeneity in the microstructure. This is further exacerbated by the presence of micro-alloying elements such as Niobium (Nb) and Vanadium (V) that form complex carbides, nitrides and carbo-nitrides during post weld heat treatment (PWHT). The focus of this work is to evaluate the effect of micro-alloying elements on the ductile to brittle transition temperature and the mode of fracture at temperatures relevant to offshore applications. A threshold Nb and V level has been determined for achieving acceptable weld metal toughness. The improvement in the fracture toughness using this approach has been quantified by Charpy V-Notch (CVN) and Crack Tip Opening Displacement (CTOD) measurements. The Ductile to Brittle Transition Temperature (DBTT) has been shown to be shifted to lower temperatures by 25 °C after post weld heat treatment in the welds where the total amount of Nb and V are controlled to less than 40 ppm. A wet precipitate extraction technique was used to extract precipitates from the welds to establish the presence of fine Nb rich precipitates in the welds with the higher DBTT. The weld deposited with controlled levels of Nb and V was further tested in different joint configurations and base plate thickness. The fracture toughness was evaluated by CTOD testing of the weld in two different thicknesses (50 mm and 70 mm). Increased specimen thickness resulted in lower CTOD values.

Stress concentration factors for tubular joints were computed using solid quadratic elements. The results of the computations are compared with experiments reported in the literature and with expressions reported in the literature and in design codes. An influence of element size and element type was observed, which leads to recommendations regarding element size of four quadratic elements in thickness, which is finer than in different published recommendations. A parametric study was performed, showing that stress concentration factors from the literature are not always conservative, particularly at the crown toe of the chord, while they tend to be overconservative at the chord saddle. The stress concentration factor for the inside of the member was also computed; it is found that it can be close to the stress concentration factor at the weld toe for both the in plane or out of plane bending modes.

The Hot Spot Stress (HSS) distribution around welded tubular joints in truss legs is essential information for the fatigue assessment of jack-up platforms. Both Finite Element (FE) analysis and model test were carried out on a multi-planar tubular KK joints under asymmetric axial loads in our research. The study shows that the HSS distribution around welded joints is different from common K or T joints, and the maximum HSS appears between crown and saddle on the chord surface at the chord-brace intersection. Moreover, some sensitive factors which influence the locations of the maximum HSS were investigated by FEA in this paper. The analysis results are useful for the optimal structure designing and crack initial prediction of the offshore jack-up platforms legs.

The application of a fracture mechanics based methodology for integrity assessment of complex structural components containing flaws often requires computational simulation where the flaws have to be modeled explicitly due to the absence of the appropriate analytical solutions. Moreover, for large flaws, re-distribution of loads may occur and therefore for an accurate assessment the flaw needs to be embedded within the structural model. A critical issue that must be addressed in a three-dimensional finite element simulation of a structural component containing a crack-like defect is that of mesh generation. In this paper, an efficient methodology has been developed based on the use of general purpose finite element packages and other commercial software to automatically generate the meshes. The model reflects the full complexity of a 3-D structure ensuring at the same time the required mesh resolution around the crack front for the determination of parameters used in fracture mechanics assessment. The approach is demonstrated through computational simulations of a tubular K-joint containing a crack-like defect with elastic and elastic-plastic material properties. From these analyses, the parameters required for construction of a geometry-dependent and material-specific failure assessment diagram (FAD) are derived. Results of the fully computational fracture mechanics assessment are compared with those generated in accordance with Annexes B and P of BS 7910:2005. It is demonstrated that the level of conservatism in fracture integrity assessment of tubular joints can be reduced if the crack driving force is determined directly from the computational analysis of the cracked geometry. A discussion on the application of different limit load and reference stress solutions is presented and the importance of the appropriate selection of the limit load solution is highlighted.

When load causes stress above the yielding stress, stress concentration factor from linear conditions can no longer be applied. Neuber’s rule has been traditionally used to estimate the non-linear stress concentration factors to avoid elaborate numerical analysis. However, the multiaxial stress state around the tubular intersection has been found to be too complex for Neuber’s rule to be applied directly. To facilitate calculation of non-linear stress concentration for tubular intersection, this paper proposed an effective relative notch factor for tubular connection stress concentration based on available linear stress concentration factors and fatigue experimental studies. This approach is validated by detailed FE analysis for three representative tubular joint configurations, two T joints and one Y joint. It has been found that although the stress states are different from typical notch specimen, the modified Neuber’s method using notch stress concentration factors Kt = Ktn as a product of tubular hot spot SCF and relative notch factor of 1.53 matches true stress closely initially but progressively become less and less accurate. When the applied load is about 70% of the plastic collapse load, the method give 30% under prediction at the worst case. This may be due to unaccounted non-linear effect from local buckling and excessive plasticity.

The statistical uncertainty of the long-term distribution of wind- and wave-induced hot-spot stress ranges in multi-planar tubular joints of a fixed jacket offshore wind turbine designed for a North Sea site in a water depth of 70m has been assessed in this paper. The dynamic response of the jacket support structure due to wind and wave loads is calculated using a decoupled procedure. Hot-spot stresses at failure-critical locations of each reference brace for 4 different tubular joints (DK, DKT, X-type) are derived by summation of the single stress components from axial, in-plane and out-plane action. The effects of planar and non-planar braces are also considered. A two-parameter Weibull function is used to fit the long-term statistical distribution of hot-spot stress ranges by combination of time domain simulation for representative environmental conditions (wind / sea states) in operational condition of the wind turbine. The statistical uncertainty of the Weibull distribution of hot-spot stress ranges and the two parameters defining the Weibull distribution is assessed, based on 20 simulations for each representative environmental condition. The contributions to the uncertainty from wind loads and wave loads are analyzed by considering 3 different load cases: wind loads only, wave loads only and combination of wind and wave loads. The sensitivity of the long-term distribution of hot-spot stress ranges due to their stress components is also assessed.

Spectral fatigue analysis approach is highly recommended for fixed offshore platform design and reassessment by API. This method is a computationally efficient method, being able to handle the random nature of environmental ocean wave conditions during calculating wave loads on the offshore platforms and subsequent structural responses. However, its fundamental theory is based on the assumption of linearity of both structural system and wave loading mechanism. Although this method is critically appropriate to be applied in offshore platform design and fatigue assessment for deep water scenarios where wave and force nonlinearities are not very severe, it has still been widely utilized for the design and assessment of shallow water platforms in offshore industry without carefully considering possible errors caused by strong nonlinear factors between ocean waves and forces. The source giving rise to the errors is because of the difficulties in choosing suitably correct wave heights for a series of wave periods required for producing transfer functions between sea state spectra and structural response spectra. Therefore, the studies to justify the possible errors of the spectral fatigue analysis method for shallow water platforms have been provoked. This paper presents the results of the studies of investigating the errors from currently existing spectral fatigue analysis method. A new technical approach that can reduce the errors in the spectral fatigue analysis of shallow water platforms is introduced. The proposed technical approach is mainly focused on producing realistic transfer functions between sea state spectra and structural response spectra, which can reasonably reflect the individually local sea state data by using wave height-period joint probability density function. Hence the fatigue damage and life at the tubular joints of offshore platforms can be more precisely predicted. The spectral fatigue analysis of a practical shallow water jacket platform in the recent platform design project has been performed using the proposed approach and the results are discussed.

In a conventional method of structural analysis, for modeling and analysis of jacket type offshore platforms, member connections are assumed to be rigid. In this method, members are rigidly connected which means there is no axial or rotational deformation at the end of brace member relative to chord axis. However in reality local deformations occur at chord surface due to applied loads from braces, which mean tubular joints are considerably flexible especially in non linear range of deformations. Therefore results of analysis based on rigid connections assumption differ from real behavior of the structure. Various research works have been carried out in the past on tubular joints and different methods have been presented in order to include the effect of joint flexibility in structural analysis. Most of these methods are just valid in elastic range but some non-linear methods have also been developed for simple tubular joints. In order to carry out a nonlinear analysis on a 3-D model of an offshore platform with multi-brace / multi-planar tubular joints, none of these simplified methods is applicable. In this case a complete model of tubular joints by non-linear shell elements is the most accurate one which is not only valid for non-linear analyses but also covers all type of tubular joints. In this paper two samples of offshore platforms are studied. These platforms are modeled using the following approaches: 1. No modeling of joints as structural elements (rigid connections). 2. Modeling of joint can with nonlinear shell elements (flexible connection). Different types of static non-linear analysis (Push over) are carried out and results are compared. In order to evaluate the results and compare this type of modeling with simplified methods included in professional software for the analysis of offshore structures, aforementioned platforms are also analyzed using the Fessler and MSL models to include effects of joint flexibility. The results of these types of modeling are also compared with the previous ones.