On detection of crack-like welding defects by existing quality control methods

Introduction. The research materials devoted to the operability assurance of welded structures of hazardous technical devices — components of oil and gas production equipment are summarized. The factors determining the operational strength of welded joints and structures are systematized and diagrammed. The causes for the decrease in their service properties are described. We have defined the role of volumetric and crack-like welding defects, which were not detected during quality control at the manufacturing stage and in determining the residual life while in operation, in ensuring the life cycle of the structure. Measures to improve the practice of detecting crack-like defects under visual- and-dimensional and ultrasonic methods of quality control of welded joints are proposed.Materials and Methods. In accordance with SNiP 2.05.06–85, the structures of main oil and gas pipelines, vertical cylindrical tanks, and other oil and gas production equipment (OGPE) are made of dead-killed and semikilled low- carbon and low-alloy steels with tensile strength of up to 686.5 MPa. Regardless of the class and strength level of steel, it should be well welded by all methods prescribed by standard process documentation (SPD). At the same time, regardless of the state of supply, the carbon equivalent Сэкв should not exceed 0.46 %. The research methods are calculation-experimental ones. To calculate the stress-strain state of welded joints, the following methods were used:finite elements (FEM),fracture mechanics using the stress intensity coefficient Kt.Methods of mathematical statistics were used to estimate the geometric dimensions of crack-like welding defects. Welding defects were detected by standard quality control methods prescribed by the SPD and GOST standards.Results. Welded structures operating in the fatigue mode are considered. Some factors characteristic of welded joints are analyzed. It is shown how they affect the formation of strength performance properties. The role of dangerous crack-like defects, which with high probability can be formed in welded joints in the manufacture of structures and during operation, is established. Often, the reason is a decrease in mechanical properties due to aging and loss of plasticity caused by accidental mechanical actions. Note that it is impossible to identify these defects by existing quality control methods, both during the control process under production and during diagnostics while in operation. This reduces the accuracy of predicting the operational life of the welded structure.Discussion and Conclusion. It is proposed to include the following requirements in the SPD:to the quality of welded joints of hazardous technical devices of OGPE,to detection of sharp crack-like defects with a radius of curvature in the range of 1-0.25 mm.However, the reliability of detection of such defects by the control methods used remains low due to the human factor. To increase the reliability of detection of sharp crack-like defects, the directions of improvement of ultrasonic quality control of welded joints are determined.

Interaction of a screw dislocation with a parabolic cavity and a semi-infinite crack

We derive an analytic solution to the problem of a screw dislocation interacting with a parabolic cavity and a semi-infinite sharp crack using conformal mapping techniques and the method of images. Closed-form expressions for the image force acting on the screw dislocation, the mode III stress intensity factor at the crack tip and the generalized mode III stress intensity factor for the parabolic cavity are obtained. The correctness of the solution is validated by comparison with existing solutions in the literature.

Scattering on a square lattice from a crack with a damage zone

A semi-infinite crack in an infinite square lattice is subjected to a wave coming from infinity, thereby leading to its scattering by the crack surfaces. A partially damaged zone ahead of the crack tip is modelled by an arbitrarily distributed stiffness of the damaged links. While an open crack, with an atomically sharp crack tip, in the lattice has been solved in closed form with the help of the scalar Wiener–Hopf formulation (Sharma 2015 SIAM J. Appl. Math. , 75 , 1171–1192 ( doi:10.1137/140985093 ); Sharma 2015 SIAM J. Appl. Math. 75 , 1915–1940. ( doi:10.1137/15M1010646 )), the problem considered here becomes very intricate depending on the nature of the damaged links. For instance, in the case of a partially bridged finite zone it involves a 2 × 2 matrix kernel of formidable class. But using an original technique, the problem, including the general case of arbitrarily damaged links, is reduced to a scalar one with the exception that it involves solving an auxiliary linear system of N × N equations, where N defines the length of the damage zone. The proposed method does allow, effectively, the construction of an exact solution. Numerical examples and the asymptotic approximation of the scattered field far away from the crack tip are also presented.

Security Hostile Attacks

Attacks on people at events and crowds in general are found around the world. It has completely changed the security at events, and event planning as a whole. From cement bollards to bag checks, it is an irritant to every event attendee. The extra security cost of events has risen so high that many events have been cancelled. The celebratory element of the event has been diminished. However it is not the new phenomena that the media seems to assume. Many countries have lived with political/social inspired attacks for years. If a country has hostile neighbours, terror attacks will occur. Regardless of the statistical evidence and the probability of an attack, people are so worried that the sound of a sharp crack in a crowded place can cause panic and a stampede. This chapter describes some of the attacks on crowds at events, and what the security agencies and governments require from those who manage the crowded spaces. It is well to remember that no matter what is written here or in government recommendations, the attackers are ‘free agents’ and can adapt their actions to the conditions trying to prevent them.

Comparison of Fracture Assessments of Corrosion Pits Using Sharp and Blunt Notched Crack Procedures

Historically, when determining failure criteria for a cracked structure, the structure has been assessed with the crack tip assumed to be infinitely sharp. For scenarios where it can be shown that the crack tip is blunt, such as for corrosion pits, this may lead to overly pessimistic assessments which can have implications for remnant life assessments of structures. Recent research has been carried out to evaluate structures with the crack tip as a blunt notch with a radius, and to develop a notch acuity parameter. That research led to formulae for adapting either the failure assessment curve or the failure assessment points utilising the notch acuity parameter for ferritic materials. This paper analyses various configurations of crack and specimen geometry, crack tip notch acuity parameter and crack spacing. It takes a finite element model of a cracked structure with an infinitely sharp crack and compares the failure parameters to those of equivalent notched cracks. The comparison is completed using the formulae derived in recent literature and those determined using finite element models with the blunt notched radii at the crack tip incorporated into the model.

Limits of Applicability of the Notch Failure Assessment Diagram

Structural integrity assessment codes such as R6 [1] and BS7910 [2] provide guidance on the assessment of flaws that are assumed to be infinitely sharp using the Failure Assessment Diagram (FAD). In many cases, such as fatigue cracks, this assumption is appropriate, however it can be pessimistic for flaws that do not have sharp tips such as those associated with lack of fusion, porosity or mechanical damage. Several Notch Failure Assessment Diagram (NFAD) methods have been proposed in the literature to quantify the additional margins that may be present for non-sharp defects compared to the margins that would be calculated if the defect were assumed to be a sharp crack. This paper uses mechanistic modelling to define the limits of applicability of the NFAD approach in terms of ρ/a, where ρ is the notch root radius and a is the notch depth. The work concludes that the NFAD can be used to assess notches with ρ/a values of up to unity.

Using Virtual Testing to Study the Influence of Constraint on Fracture Properties

This paper presents a study in which the influence of constraint is investigated using a ductile damage model based on a stress modified failure strain. This model, proposed by Oh et al, has been used to successfully predict the J-R curve of standard compact tension (CT) test pieces for a range of different materials. Further work has shown how the model can also predict the fracture behaviour of non-standard CT test specimens, in which a blunt notch is used in place of a sharp crack. Continuing on from this work, the paper predicts the J-R curves of single edge notch tension (SET) test pieces with both a sharp crack and blunt notches.

Validation of the Proposed R6 Method for Assessing Non-Sharp Defects

Structural integrity assessment codes such as R6 [1] and BS7910 [2] provide guidance on the assessment of flaws that are assumed to be infinitely sharp using the Failure Assessment Diagram (FAD). In many cases, such as fatigue cracks, this assumption is appropriate, however it can be pessimistic for flaws that do not have sharp tips such as lack of fusion, porosity or mechanical damage. Several Notch Failure Assessment Diagram (NFAD) methods have been proposed in the literature to quantify the additional margins that may be present for non-sharp defects compared to the margins that would be calculated if the defect were assumed to be a sharp crack. This paper presents the second stage of validation work, using 3D Finite Element (FE) Analyses and a wide range of test data on non-sharp defects, to validate an NFAD method proposed for inclusion in R6 and to quantify the errors caused by various approximations in the method.