Integrating crack geometry, specimen thickness and stress ratio effects in fatigue crack growth studies: closing the gap between laboratory and real c
Giovanna Calvín García
- DIRECTORS: Haritz Zabala Rodríguez and Miguel Muñiz Calvente
- UNIVERSITY: Universidad de Oviedo
Industrialization has transformed society with the introduction of mechanization and new technologies. However, such progress has also presented challenges to structural integrity due to increased production and the extensive use of machinery. The materials used in components often contain defects that evolve into larger cracks, resulting in human risks and causing economic losses or even catastrophic failures.
This issue persists, and it is estimated that a significant portion of industrial failures are attributed to material fatigue problems. Therefore, it is crucial to carefully assess the structural integrity of critical components or structures with defects subjected to cyclic loads over time. In this context, predicting fatigue crack growth (FCG) could be the solution to preventing catastrophic failures in order to make decisions about repairs and replacements or schedule reliable inspections. However, making a concise and reliable FCG prediction is not an easy task. To achieve this, it is essential to properly select the fracture parameter governing crack propagation and adequately characterize the crack growth law. The success of this prediction depends on the accurate consideration of factors such as geometry, loading conditions, or material behavior.
Metallic materials are often affected by plastic deformations that promote phenomena like plasticity-induced crack closure (PICC). PICC leads to a reduction in crack growth rate due to contact between the crack surfaces and must be included in both the crack growth law characterization and the component life prediction. Currently, the most common method to determine PICC is the one proposed by Newman, through the relationship between the crack opening load and maximum applied load (Pop/Pmax). This parameter is included in the most widely used fracture parameter for evaluating FCG in metallic components, known as the effective range of stress intensity factor, ΔKeff. Various analytical, numerical, and experimental approaches can be found in the literature to determine it.
This thesis aims to enhance the transferability of FCG curves determined in laboratory conditions to predictions of crack growth in real components. This is achieved by integrating the effect of crack geometry, specimen thickness, and stress ratio into the calculation of the ΔKeff fracture parameters. To accomplish this, available methods in the literature for calculating the components of ΔKeff (Kmax y Pop/Pmax) are critically evaluated, paying special attention to the assumed hypotheses. Having observed the limitations of the previous models, a procedure based on the use of three-dimensional advanced numerical models capable of more accurately capturing PICC is proposed.
The results of this research i) highlight the importance of considering the three-dimensional crack geometry for FCG curve characterization, ii) demonstrate that a characterization based on standardized procedures (ASTM) cannot be considered a material property as it is influenced by specimen thickness, iii) show that characterization independent of specimen thickness can be obtained through the developed methods, and iv) reveal inconsistencies resulting from the use of the nodal displacement method for PICC calculation under negative load ratios. This study then proposes a procedure to address this issue.