Fatigue Behavior of Automatically Welded Tubular Joints for Offshore Wind Energy Substructures
| Autor: | Schürmann, Karsten Wolfgang |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
fatigue strength
bionic optimization Schweißnahtgeometrie Dewey Decimal Classification::600 | Technik::620 | Ingenieurwissenschaften und Maschinenbau technology industry and agriculture Automatically welded tubular joints Automatisiert geschweißte Rohrknoten Bildkorrelationsverfahren bionische Optimierung respiratory system Ermüdungsfestigkeit reale Kerbspannungen digital image correlation real notch stresses ddc:620 weld geometry |
| Popis: | To successfully fight climate change, it is essential to replace fossil fuels with renewable energies. Offshore wind energy will significantly contribute to this transition. Since the required future bottom-fixed offshore wind farms will be located further away from shore and in greater water depth, in addition to (XL-) monopiles, jacket foundations are a reasonable alternative. To increase the competitiveness of jacket substructures, a modular jacket concept was developed within the offshore industry, combining prefabricated robot-welded tubular joints with standardized pipes. With regard to fatigue strength in particular, the automatically welded tubular joints have a large potential due to the reproducible fabrication process resulting in highly uniform welds. However, this potential cannot be adequately considered for the fatigue assessment due to the lack of suitable S-N curves. Up to now, serial fatigue tests to determine statistically validated S-N curves have only been conducted on manually welded tubular joints. Nevertheless, some robot-welded tubular joints were tested regarding fatigue strength, but neither the influence of an inner root welding nor the existing weld geometry was systematically evaluated. Against this background, the fatigue resistance of automatically manufactured tubular joints is determined within this thesis. Furthermore, this work focuses on the characterization of the uniform weld seam geometry as a prerequisite for an additionally proposed weld geometry optimization using bionic approaches. To determine a statistically validated S-N curve, serial fatigue tests were performed on 16 single- and 16 double-sided automatically welded tubular X-joints, whereby these joints were medium-scaled compared to real jacket dimensions. During these tests, the fatigue damage evolution was digitized using the digital image correlation method, enabling a detailed analysis of the tubular joints' fatigue behavior including crack initiation. The obtained fatigue resistance of the robot-welded tubular X-joints was moderately improved compared to the currently valid design S-N curve. The corresponding scatter was significantly reduced in comparison to the experience of manually welded tubular joints. In contrast, no significant impact of the inner root welding on the fatigue strength could be observed. In addition to the fatigue tests, the geometry of the robot-fabricated welds was systematically evaluated with regard to its reproducibility. The outcomes were then compared to reference values of manually welded tubular joints. For this purpose, analytical investigations were performed to determine the notch radius and flank angle distributions. Additionally, a reverse engineering application was developed to enable a real notch stress analyses of the actual weld geometry. The obtained statistics representing the distributions of the flank angles and real notch stresses confirmed the optical impression of a uniform and highly reproducible weld geometry when compared to the manually welded tubular joints. However, with respect to the minimum size of the achieved notch radii, no significant advantage of the robot-based welding could be determined. Finally, considering the statistically confirmed reproducibility of the weld geometry, a bionic optimization of the weld geometry profile was proposed, which resulted in a significant reduction of the decisive fatigue loads. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|