Project

Experimental-Numerical Study on Offshore Fatigue Analysis of Steel Structures Under Variable Load Conditions Influenced by Environmental Parameters

Code
178TW1514
Duration
01 January 2014 → 31 December 2017
Funding
Regional and community funding: IWT/VLAIO
Research disciplines
  • Engineering and technology
    • Structural engineering
    • Other civil and building engineering
Keywords
Offshore Steel Structures renewable energysources offshore wind
 
Project description

Due to actual international concern about climate change, renewable energies found a privileged position as they neither pollute the air nor consume water and they are considered to be inexhaustible. The recent confirmation of the first-ever universal, legally binding global climate deal achieved at the Paris climate conference (COP21) in December 2015 strengthened that position. Offshore wind prevailed amongst these alternative energy sources because of its maturity and previously gained onshore experience. In fact Europe’ cumulative installed capacity grew exponentially from the early 2000 reaching by the end of 2016 a total capacity of 12,631 MW, across a total of 3,589 wind turbines. The rated capacity of offshore wind turbines has grown 62% over the past decade. Of course the higher the expected capacity, the larger and more robust the turbine and its support structure must be. Present concern is related to the inclusion of new generation steels for the support structures in the relevant design codes which presumably can reduce the total weight and therefore the manufacturing and installation costs of these structures. However a deeper understanding of these materials’properties subjected to close to real offshore conditions (in terms of dynamic loads and harsh environment) is needed. As a matter of fact, fatigue plays a key role during all project phases (design, transport, installation) and furthermore during operation. Recent state of the art reviews have highlighted the need of expanding the material database included in the design codes and moreover to enhance the standardized design methodologies. Major enhancement areas are related (but no limited) to: development of testing methodologies aiming for the faster inclusion of new generation materials in the design codes, admittance of nonlinear damage estimation approaches that could account for the effect of variable amplitude loading on the fatigue life, quantification of scale influence on the fatigue crack growth rate and a better understanding of the interaction of corrosion and fatigue damage processes. In this work first attention is directed towards a critical interpretation and comparison of fatigue design codes in order to detect possible enhancement items. Based on the obtained conclusions it was decided to focus on the experimental development of fatigue testing methodologies and setups for a better understanding of the following topics: - Determination of the fatigue limit and the entire S-N curve using multiple instrumentation techniques and a reduced number of samples. - Interaction effects occurring during variable amplitude and ΔK block controlled tests. - Scale effects not considered in fatigue design codes. - Accelerated estimation of the corrosion-fatigue damage process. Hereunder a brief summary of each of these topics is given. An accurate estimation of the entire S-N curve plays an important role in fatigue structural design. Traditional experimental methods used to this end are highly time consuming. In recent years the application of infrared (IR) technology for accelerated fatigue damage evaluation has increased. However, this technology is not always deployable in aqueous environments, e.g. due to lack of space or accessibility. Under such circumstances, the electrical potential drop (PD) technique has shown to be applicable for crack initiation and crack growth assessment. Attention was paid to the description and comparison of the established IR and a novel PD method to determine the fatigue limit of two HSLA steels in an effective and efficient way. The results of both methods showed very good correlation. Conventionally a load spectrum of varying stresses is reduced into a series of blocks with constant amplitude stress reversals. The block load variations generate non-linear transients in the crack growth rate, i.e. crack growth acceleration or retardation effects. However these are generally ignored and a linear damage accumulation method is used to estimate fatigue lifetime. This can have large cost implications for example in the planning of maintenance intervals based on a fracture mechanics based assessment. To investigate the effect of block loading transients, different ΔK block loading schemes with changes in magnitude similar to those expected for a structure operating in the North Sea environment have been elaborated. Two offshore steel grades have been subjected to variable amplitude fatigue crack growth tests according to these testing programs. Retardation of crack growth has been evidenced in socalled high-low sequences. This effect was most pronounced for larger jumps in stress amplitude and lower ΔK values which eventually can cause crack arrest. Fatigue design curves have typically been derived based on testing small scale standardized specimens. A correction factor is applied to account for differences in thickness and weld geometry. However crack growth behaviour in tubular elements is complex and it is not totally clear if this can accurately be represented by testing small-scale coupons. Large scale fatigue tests aim to characterize the fatigue behaviour in a more realistic manner. However conducting such tests is very expensive. Hence, a medium scale strip type specimen with dimensions between small scale standardized specimens and a full pipe was designed. It has the same curvature and thickness as the pipe from which it is extracted and a semi-circular notch is introduced. In this way it can account for most of the scale effects. Furthermore it can be tested in conventional test rigs at rather high frequencies. By comparing the corresponding da/dn-ΔK curves of both small and medium scale specimens, it was possible to conclude that the standardized specimens overestimate the crack growth rate. Additionally methodologies are proposed for a more accurate use of small scale specimen properties to assess notched components. The corrosive nature of the marine environment is an important factor to be considered during the fatigue design of offshore structures. Thereto, S-N curves must be determined in close to real conditions which is highly time consuming. It is hypothesized that if the corrosion process is accelerated at approximately the same rate as the fatigue frequency, testing time could be highly reduced. Corrosion acceleration is possible by modifying physical and/or electrochemical properties involved in the redox reactions. In this research the first option was chosen. Based on a literature review temperature and dissolved oxygen level were concluded to be the most influencing parameters. Several tests scenarios with different combinations of sea water temperature and dissolved oxygen level have been defined. Corresponding SN curves have been constructed for two high strength low alloyed specimens immersed in natural seawater. The direct current potential drop technique was used to quantify the damage evolution for all tested scenarios. Additionally, a reference S-N curve for immersed behaviour was determined at a temperature and frequency close to North Sea conditions. Comparison of the experimental results indicated that an average acceleration of the corrosion assisted fatigue damage process of around 80% could be obtained. Future research opportunities are identified as follows. The experimental investigation of a broader range of parameters for a confirmation of the obtained results; such as R-ratio effects, randomization effect and quantification of the corrosion acceleration. Additional attention should be directed to the analysis of welding details since they represent a weak point in real structures. A final point of attention is the effect of corrosion and hydrogen embrittlement produced by micro-organisms and the investigation of hybrid material joints for marine applications