Oil and gas pipelines are designed using recognised pipeline design codes, which typically limit the applied stresses and strains to below yield. Pipelines may be subjected to higher plastic strains before they enter service and during operation, either intentionally, as a result of the method of installation, or a requirement of the design and operation, or accidentally; examples include reeling, denting, ground movement, frost heave and earthquake loading. The material subjected to high plastic strain, or pre-strain, will have different material properties to those of the ‘virgin’ material. Plastic deformation causes work hardening; the yield strength is increased, but the strain hardening capacity and the ductility of the material are reduced. The published literature on the effects of pre-strain indicates that it has a detrimental effect on the fracture toughness of steel. Pre-strain reduces the resistance to crack initiation and crack growth, and increases the transition temperature. Therefore, pre-strain will affect the response of the material to operational loads and its resistance to defects.

Conventional design practice does not explicitly consider the effect of pre-strain on the material properties. The trends towards using high strength line pipe steels, and to designing to operate at higher stresses and strains, mean that historical experience and empiricism may become inappropriate or non-conservative. The development of more accurate methods for assessing mechanical damage, to replace the existing semi-empirical methods, will require that the effects of pre-strain be explicitly considered.

OUTLINE OF THE STUDY This thesis describes an investigation of the effect of static, tensile pre-strain on the fracture toughness of line pipe steels, with the objective of understanding and quantifying the effect of pre-strain on toughness. The study comprises experimental, numerical and theoretical analyses. The study has identified the reasons for the effect of pre-strain on toughness, and the properties of the virgin material that determine the severity of the effect. A model for predicting the effect of pre-strain on toughness has been developed.

The material tested in the experimental study was an API 5L X80 line pipe steel. Samples of the virgin material were subject to pre-strains of approximately 2.7 percent and 6.5 percent engineering strain. The lower level of pre-strain is similar to that in a pipeline subject to frost heave, whilst the higher level of pre-strain is similar to that caused by denting. For comparative purposes, samples of the virgin material were also artificially strain aged. Tensile, notched tensile, fracture toughness (J-integral and crack tip opening displacement), and Charpy V-notch impact tests of virgin, pre-strained and strain aged material were conducted.

The numerical study was based on the material and test specimen geometries considered in the experimental study. The finite element (FE) analyses were conducted using ABAQUS/Standard v6.2. Large scale, non-linear geometry effects were considered. The constitutive model assumed isotropic hardening. The effect of softening due to void growth was included using the modified Gurson-Tvergaard model of porous metal plasticity. Axi-symmetric FE analyses of the notched tensile geometries and the experimental results were used to calibrate the plane strain FE model of a compact tension test specimen. The effect of the properties of the virgin material on the reduction in toughness caused by pre-strain was investigated.

A theoretical model of the effect of static, tensile pre-strain on fracture toughness was derived using the local approach to fracture. The effect of pre-strain is expressed in terms of the ratio of the fracture toughness of the pre-strained material to that of the virgin material. The HRR singularity was used to describe the stress and strain field around the crack tip. A stress-modified, critical strain-controlled model was used for ductile fracture. A critical stress-controlled model was used for cleavage (brittle) fracture.

RESULTS OF THE STUDY The trends of the published data, the experimental data, the numerical analyses and theoretical model are consistent. Tensile pre-strain increases the yield and tensile strength (the yield to tensile ratio tends to unity), and reduces the strain at the tensile strength, the percentage elongation at fracture and the true fracture strain. It reduces the critical fracture toughness and the fracture initiation toughness, and reduces the tearing resistance at higher levels of pre-strain. In the experimental study, the 2.7 percent pre-strain reduced the toughness of the virgin material (expressed in terms of δm) by approximately 14 percent (average of six tests), and the 6.5 percent pre-strain reduced the toughness by approximately 30 percent (average of four tests). Reasonable agreement was obtained between the predictions of the theoretical model for ductile fracture and the test data reported here and in the published literature.

The effect of tensile pre-strain on toughness can be attributed to: • the increase in the yield strength, • the decrease in the strain hardening capacity, and • the decrease in the fracture strain. This conclusion is supported by the observation that the trends in the tensile properties and toughness of strain aged material, or material subject to high strain rates, are similar to those seen in pre-strained material. Material damage, in the form of the nucleation and growth of voids during the introduction of pre-strain, is not a significant factor in the reduction in toughness caused by pre-strain.

The stresses and strains around the crack tip are higher in the pre-strained material compared to the virgin material. If the fracture mechanism is ductile, the void growth rate is higher, so crack initiation and stable crack growth occur at lower applied values of J and δ. If the fracture mechanism is cleavage, the Weibull stress is higher, with similar implications for crack initiation and unstable crack growth. Consequently, the fracture toughness is reduced by pre-strain. The increase in the transition temperature caused by pre-strain can also be explained in terms of the effect of pre-strain on the stress and strain field around the crack tip.

The properties of the virgin material that influence the effect of pre-strain on toughness are: • the yield strength, • the ductility, • the strain hardening behaviour, • the volume fraction of void nucleating particles, and • the transition temperature. The effect of pre-strain will be greatest when the fracture mechanism changes from ductile to cleavage, i.e. when the fracture mechanism of the virgin material is ductile, and then the pre-strained material is within the transitional region or on the lower shelf.

Fracture toughness tests of different steels subject to tensile pre-strain can be compared if the test data is expressed in terms of two ratios: • the ratio of the toughness of the pre-strained material to that of the virgin material, and • the ratio of the true pre-strain to the true fracture strain (measured in a tensile test) of the virgin material. The true fracture strain can be replaced by the true strain at the tensile strength of the virgin material, albeit with an increase in the scatter of the test data.

A semi-empirical relationship, based on the theoretical model, is proposed for predicting the effect of pre-strain on toughness. The relationship is expressed in terms of the true strain at the tensile strength of the virgin material. The relationship is conservative with respect to the test data reported here and in the published literature.

Tensile pre-strain reduces toughness. The study has shown that the effect of pre-strain on toughness is greater for virgin materials with a low ductility (defined by the strain at the tensile strength or the fracture strain), or a low strain hardening capacity. Consequently, the effect of pre-strain on toughness will be more severe in older line pipe steels, compared to modern steels, and more severe in modern high grade line pipe steels, compared to modern low grade steels.