Fred Nilsson

FEM-remeshing technique applied to crack growth problems

The remeshing option of the commercial FE-code ABAQUS is used to study crack propagation in elastic-plastic materials. Propagation is accomplished with a combination of remeshing and nodal relaxation. The results are compared with analyses using nodal release only. Substantial crack growth is achieved with less computer time and higher accuracy than in conventional FE analysis.

A path-independent integral for transient crack problems

Abstract The problem of a stationary crack in a visco-elastic body under plane strain, subjected to dynamic loading, is formulated within the realms of the classical theory. By introducing the Laplace-transform with respect to time the problem is reformulated in the transformed space. A line-integral is defined and it is shown that it is path-independent. The relation between this integral and the transformed stress-intensity factor is derived. It is indicated that the integral may be valuable for calculations of dynamic stress-intensity factors. In order to illustrate this, a simple example is solved and some numerical results are presented.

Generalizing the safety factor approach

Safety factors (uncertainty factors) are used to avoid failure in a wide variety of practices and disciplines, in particular engineering design and toxicology. Although these two areas have similar ...

Risk-based approach to plant life management

The growing interest in risk-based management is driven by the need of developing strategies that lead to an optimal safety versus cost balance. The general philosophy behind risk-based management is briefly described here and discussed. It is formally shown that the core damage frequency can be factored into a system (PSA) part and component failure frequency part. Some of the procedures, currently applied in risk-based inspection, are discussed. The basic elements for failure frequency calculations are presented and discussed. A quantitative risk-based inspection study performed for the Oskarshamn 1 unit is briefly presented as an example of how risk-based procedures can be applied.

A perturbation method used for static contact and low velocity impact

Abstract An approximate method for the solution of static and dynamic contact problems between bodies with non-linear material behaviour is described. The method is a perturbation technique based on the linear elastic quasi-static solution. Here the method is applied to the problem of a sphere in contact with a half-space which means that the Hertz solution is used. The governing equations are rewritten so that the problem for the perturbed variables is one with surface forces in the contact region and volume forces inside the bodies. The latter are due to accelerations and strain gradients calculated from the quasi-static solution and the equation of motion. The contact condition results in an integral equation for the surface forces. Results are compared with FEM calculations, which show very good agreement for the dynamic case, both with linear elastic and non-linear (plastic) material behaviour. For the static case with non-linear material behaviour the results are good approximately up to the point where the inelastic zone reaches the surface of the bodies.

A procedure for estimation of pipe break probabilities due to IGSCC

Abstract A procedure has been developed for estimation of the failure probability of weld joints in nuclear piping susceptible to intergranular stress corrosion cracking. The procedure aims at a robust and rapid estimate of the failure probability for a specific weld with known stress state. Random properties of the crack initiation rate, the initial crack length, the in-service inspection efficiency, and the leak rate, are taken into account. A computer realization of the procedure has been developed for user-friendly applications by design engineers. Some examples are considered to investigate the sensitivity of the failure probability to different input quantities.

JIC-testing of A-533 B— statistical evaluation of some different testing techniques

The purpose of the present study was to compare statistically some different methods for the evaluation of fracture toughness of the nuclear reactor material A-533 B. Since linear elastic fracture mechanics is not applicable to this material at the interesting temperatures (275°C), the so-calledJIc testing method was employed. Two main difficulties are inherent in this type of testing. The first one is to determine the quantity J as a function of the deflection of the three-point bend specimens used. Three different techniques were used, the first two based on the experimentally observed input of energy to the specimen and the third employing finite element calculations. The second main problem is to determine the point when crack growth begins. For this, two methods were used, a direct electrical method and the indirect R-curve method. A total of forty specimens were tested at two laboratories. No statistically significant different results were obtained from the respective laboratories. The three methods of calculating J yielded somewhat different results, although the discrepancy was small. Also the two methods of determination of the growth initiation point yielded consistent results. The R-curve method, however, exhibited a larger uncertainty as measured by the standard deviation. The resulting JIc value also agreed well with earlier presented results. The relative standard deviation was of the order of 25%, which is quite small for this type of experiment.

Crack propagation experiments on strip specimens

Abstract A method for measuring dynamic fracture properties is briefly described. The method employs the strip specimen and this configuration is discussed from a practical point of view. Results from experiments performed to determine the velocity variation of the fracture energy are presented. Some observations are made regarding the physical nature of the fracture process.

Closure effects on fatigue crack growth rates at constant and variable amplitude loading

Abstract Accurate measurements of fatigue propagation rates after single and during repetitive overloads were performed. It was shown that plastic deformations in the crack tip region and in the crack wake cause crack closure that is of major importance for crack growth rates. The consequences of this for methods to determine certain material properties are discussed. An improved method to measure closure free propagation rate data was utilised. The possibilities and shortcomings of compliance methods to determine crack closure levels are pointed out. The relevance of using ΔKI,eff=KI,max−KI,lim, where KI,lim possibly could be interpreted as the crack closure level was investigated and is discussed. A method to predict average crack propagation rates at block loadings consisting of well-defined load cycles of different amplitudes and R-factors was suggested. Three-dimensional effects after overloads were observed. These observations confirmed that the crack closure behaviour also depends on geometry of the specimen.

Influence of specimen geometry on crack propagation and arrest toughness

A number of crack propagation experiments on four different specimen geometries were performed in order to investigate the existence of a unique relationship between crack propagation toughness and crack-tip velocity. Forone geometry, dynamic finite-element method (FEM) calculations were found to be necessary in order to obtain proper stress-intensity factors, whereas quasi-static FEM-calculations were found to be sufficient for the other geometries. Crack-tip velocities were determined from continuous recordings of crack length versus time, which was measured by an impedance method. The stress-intensity factor and the instantaneous crack-tip velocity were obtained for a number of crack lengths for each experiment. The experimental results do not contradict the hypothesis of a unique relationship between these two quantities at low load levels; neither do they contradict the hypothesis of a geometry-independent crack arrest toughness. At high load levels a deviation from the uniqueness was observed. The height of the specimens was found to influence the crack propagation toughness level at the beginning of deviation from the unique relationship.