Fakultät Maschinenbau
Refine
Document Type
Language
- English (5)
Has Fulltext
- no (5)
Is part of the Bibliography
- no (5)
Keywords
- Pipeline (5) (remove)
Institute
Begutachtungsstatus
- peer-reviewed (2)
A comprehensive experimental program is carried out to determine material parameters for fracture initiation and propagation in X100 pipeline steels. The quadratic Hill’48 yield function along with an isotropic hardening are used to describe plastic flow at large deformation and a phenomenological fracture criterion to predict fracture initiation. Fracture mechanics SENT specimens are used to calibrate post-initiation softening parameters necessary for ductile crack propagation in thick components. Once the material model parameters set is complete a final comparison is conducted with ring expansion test on same material.
Investigation of running ductile fracture in gas transmission pipelines and the derivation of reliable crack arrest prediction methods belong to major topics in pipeline research. The yet available crack arrest criterion, known as the Battelle Two-Curve Method (BTCM), leads to reliable predictions up to grade X70 line pipe steels for which it has been validated. This includes specific limits in terms of mechanical properties, pressure and geometry. The application of this criterion to modern pipeline steels, i.e. especially grades X80 and beyond in combination with larger diameters and high pressure, has led to mispredictions of the BTCM. Hence, in order to ensure safe design of pipelines, new methods are required based on in depth knowledge and appropriate characterization of material resistance. This paper presents a procedure for the assessment of dynamic ductile fracture resistance based on combined experimental and numerical investigations. The procedure involves quasi-static and dynamic drop- weight tear testing (DWTT) on modified specimens with pre-fatigued crack for grades X65, X80 and X100 materials, and the application of cohesive zone (CZ) and Gurson-Tveergard-Needleman (GTN) models to describe ductile material damage. The damage model parameters are calibrated on basis of DWTT results and subsequently used to simulate dynamic crack propagation in a pipeline. The influence of material properties (strain hardening, toughness), pipe geometry, usage factor and decompression behaviour on ductile fracture propagation behaviour is studied and evaluated. The results will contribute to an enhanced understanding of major parameters controlling ductile fracture propagation and will help to establish a reliable procedure for safe design of new high-capacity pipelines with regard to crack arrest.
Fracture behavior of seamless pipeline material X65Q acc. to API 5L has been studied both experimentally and numerically at different loading conditions (quasi-static vs. dynamic) and temperatures. The recent findings have shown difficulties in applying well established methods for determination of transition behavior or prediction of ductile crack arrest for the new generation of high-toughness steels. The irregular fracture performance (e.g. so-called "abnormal inverse fracture" appearance, significant scattering in ductile-to-brittle-transition-temperature region, etc.) suggests that the influence of pipe dimensions, loading parameters, crack initiation resistance as well as testing procedure on the fracture behavior has been neither understood nor properly described. This work aims to shed light on these questions regarding the applicability of conventional methods and to better illuminate most relevant parameters affecting fracture behavior of high toughness steels. To achieve this goal, experimental data basis for analysis of fracture behavior in transition and upper shelf regime has been established by conducting quasi-static fracture mechanics tests and dynamic tests on Battelle Drop Weight Tear (BDWT or DWT) specimens at different temperatures. The evaluation of obtained test results in upper shelf has been additionally complemented by numerical simulation of damage behavior. The results highlight the influence of stress conditions on fracture behavior with reference to pipe dimensions and loading conditions and, subsequently, may be used as a basis for revision of existing design methods.
The fracture propagation and arrest control for pipelines transporting rich natural gases and high vapor pressure liquids is based on the Battelle Two-Curve Model (BTCM). Distinct limitations of this model were demonstrated for past and modern steels and gas mixtures. These can be related to the insufficient description of individual physical processes and interactions between the pipe material and transported mixture during the running ductile fracture. In the past, fluid-structure interaction (FSI) models enabled a more sophisticated, coupled analysis of the failure scenario. To quantify their capability of describing the multi-physical processes, the FSI models need to be verified by experimental data from full-scale burst tests (FSBT). Therefore, this paper deals with the simulation of five FSBTs from the literature on API grade X65 pipes with different pipe geometries, mixtures and initial conditions. The FSI is modeled by the coupled Euler-Lagrange (CEL) method. The modified Mohr-Coulomb (MMC) model is implemented in the CEL framework to describe the deformation and ductile fracture in the X65/L450 pipes. 3D Euler equations are used to calculate the mixture decompression with the GERG-2008 equation of state defining the volumetric behavior of a CO2-rich mixture, CH4 and H2. The extended model considers the effect of soil backfill on the pipe deformation and inertia. The numerical predictions agree well with the experimental findings in terms of the crack propagation speed and arrest length underlining the capability of the developed numerical tool.
As part of current design standards, the Battelle Two-Curve Model (BTCM) is still widely used to predict and secure ductile crack arrest in gas transmission pipelines. For modern linepipe steels and rich natural gases or CO2 mixtures, the BTCM might lead to incorrect predictions. On the one hand, it suffers from the insufficient description of the individual physical processes in the pipe material and fluid itself. Furthermore, the model does not account for fluid-structure-interaction (FSI) effects during simultaneous running-ductile fracture (RDF) and mixture decompression. Numerical FSI models allow for a more sophisticated, coupled analysis of the driving forces for the failure of pipelines. This paper deals with the development of an FSI model for the coupled prediction of 3D pressure profiles acting on the inner pipe wall during crack propagation. The coupled Euler-Lagrange (CEL) method is used to link the fluid and structure models. In a Lagrange formulation, the modified Bai-Wierzbicki (MBW) model describes the plastic deformation and ductile fracture as a function of the underlying stress/strain conditions. The fluid behavior is calculated in a 3D model space by Euler equations and the GERG-2008 reference equation of state (EOS). The coupled CEL model is used to predict the RDF in small-diameter pipe sections for different fluid mixtures. The calculated 3D pressure distributions ahead and behind the running crack tip (CT) significantly differ in axial and circumferential directions depending on the mixture composition. The predicted FSI between the pipe wall and fluid decompression in 3D CEL/FSI model provides reliable knowledge about the pressure loading of the pipeline during RDF.