Analysis of cyclic fracture of the threaded joints

Abstract

The initial production defects having been transferred into a structure and developed during its operation, it is crucial for labour safety to prevent such defects from reaching critical limits. First, the defects should be measured and then calculations of the parameters affecting the extent of their propagation are to be made. Provisions should be made to control damage during production and maintenance process. Evaluation of the structural element strength, once the defect or crack is detected, is rather a complicated procedure requiring sometimes additional theoretical solution or even an experimental research. The limit state conditions favouring stopping the crack find their way to practical application. If such conditions are maintained, it is estimated that the structural element containing a crack in a state of stoppage is able to perform fully its functions. In the paper, the resistance of threaded joints to cyclic loading is defined and based on the criteria of fracture mechanics describing non-permissible operational limit states: formation of macrocrack; transition from stable crack propagation to a dynamic one, ie fracture of the structural element; stable crack propagation exceeding the permissible crack size; transition from the defect (stable crack) to steadily propagating crack, ie continuous accumulation of damage. The regularities of the fracturing process, taking into account constructive peculiarities, production technology and loading conditions were studied in threaded joints within the range of M8 to M48. Threaded joints were tested according to the system ‘nut-stud-nut’ and ‘nut-stud-body’ made of steel 25XIMΦ and 20XIMIΦITP which are widely used to manufacture elements for fixing power equipment. The indicators of mechanical properties are as follows: σ0.2=780÷1020 MPa; σu=830÷1120 MPa; Ψ=58÷62%. Nuts were manufactured from the same kind of steel as studs, or from another kind of steel with different mechanical properties. The characteristic stages of crack propagation and its rate along the cavity and deeper into the cross-section are presented in the works of the authors. In the experiments, the criterion of fracture mechanics K1 was applied to describe the crack propagation rate and to investigate the conditions of brittle fracture. A special investigation was carried out to substantiate such a decision. On the basis of the refraction obtained, the S. Jarioma equation was applied to calculate the stress intensity ratio. The parameters of kinetic the fatigue diagrams (a threaded joint and cylindrical specimen 20 with a ring notch imitating the cavity of thread M20) described by means of the P. Paris equation are presented in the works of the authors. The presented experimental data of the cyclic fracture strength of models M20 are characteristic of threaded joints of other size too, not only of those with the size ranging from M8 to M48. Similar regularities were observed in natural specimens of threaded joints M 140x6. The low-cycle fatigue strength of threaded joints manufactured from steel with different mechanical properties (studs made of steel 25XIMΦ, nuts of steel 20.40X and TC) change insignificantly. A more uniform distribution of notches were observed, however, it has a minimal effect on threshold and crack propagation. To determine specific peculiarities which are characteristic of threaded joints, additional experiments were carried out. It is known that a properly constructed threaded joint ‘nut-stud- nut’ undergoes fracture in the stud bar during the construction loading (as in case of tension). Provided there is a big crack in the stud, the fracture can occur on the plane of the cavity during the construction loading. Specimens of three types were studied as follows: cylindrical specimens with a crack; studs with a crack and threaded joints ‘stud-nut’ with a crack in their studs. The results of the experiment depending on the depth of the crack are presented. Different regularities of the threshold, crack propagation and its brittle fracture can be explained as follows: a different stressed state within the cavity of the stud thread (the total result of the stud tension and turn bending) and in the ring notch of the cylindrical specimen; a different front of crack propagation; a different angle of crack propagation (perpendicular to loading in the cylindrical specimen and along the angle of the rise of a turn in the stud); the effect of the first joint turns of the stud and the nut which limits crack opening and influences the processes taking place on the top of a crack. All these factors determine a certain difference between the intensity ratios of critical stresses: materials Kc; studs Kcb and studs joint to the nut Kct. Making use of earlier and additional experimental investigation of threaded joints made of steel 25XIMΦ (normalising), the part of the diagram of kinetic fatigue was drawn in which the number of loading cycles exceeds 107. This part characterises the propagation of short cracks. The marginal value ΔKth of the intensity of stresses is a threshold below which cracks do not propagate. One more peculiarity of the experiment results should be indicated—at least two specimens (stud-nut) were tested under equal conditions, but the cracks propagated differently. In the opinion of the authors, cracks propagating in the stud change the flows of the internal resultant in the system ‘stud-nut’ as a consequence of which the crack propagation rate either decreases or terminates. Making use of the experimental data and analysis of the stressed state, the study of the region close to the threshold (ΔKth) was conducted by means of finite element method. The amplitude of the endurance limit of an even specimen is regarded as the limit state of the threshold. A satisfactory correlation of the results was observed with the stress ratio r = 0; 0.3; 0.6. The difference in the results with the stress ratio r = 0 shows the complexity of the fatigue process as well as the peculiarities of the effect of the nut upon the bolt. According to the experimental data, the length of a non-propagating crack ≈ 5 mm, and the depth exceeds 0.2 mm. Experiments carried out on studs with such cracks showed that non-propagating cracks have no effect on the static strength and fracture takes place in an even part. The determined size of non-propagating cracks is easily measured by means of non-distrusting control methods both under laboratory and operational conditions. © 2000 Taylor & Francis Group, LLC.