Anomalies concerned with interpreting fatigue data from two-parameter crack growth rate relation in fracture mechanics

作者：Sih G C^*

来源：Theoretical and Applied Fracture Mechanics, 2008, 50(2): 142-156.

DOI：10.1016/j.tafmec.2008.07.014

• Macromechanical
• Microstructural
• Nanochemical
• Through crack
• Surface crack
• Embedded crack
• Transitional behavior
• Threshold
• Sigmoidal
• Stable
• Unstable
• Slow and fast crack speed
• Micro/macro interaction
• Dual scale model
• Form invariant
• Linear interpolation
• Data interpretation
• Size and time scale

Early studies on metal fatigue have concerned primarily with testing the bulk properties of smooth or notch specimens. The fine line crack configuration became a standard ASTM specimen when local effects were recognized to play a decisive role in failure initiation. The size of the failure initiation crack can range from a few cm to a few mm or smaller depending on the application that can involve through cracks in metal sheets, surface cracks extruding grain boundaries and embedded cracks interior to the material microstructure. The state-of-the-art in fracture control for the most part is to monitor the length of an equivalent crack, say "a", such that the corresponding rate da/dN can be related to the integrity of the system depending on the geometry, load and material.
Consistent data interpretation has been a concern because of the enormous range of physical defect size that the design must cover to assure the effective application of the failure criterion from nano to structural scale size. The design may have to demonstrate that nanocracks are controlled in miniature devices as macrocracks are kept in the sub-critical state for larger structures, As more nanostructures are being used, consideration of small cracks in the damage evolution process is becoming the rule rather than the exception. Whether the ordinary two-parameter da/dN versus Delta K representation referred to Regions I-III would still accommodate microcrack and/or nanocrack behavior is an issue that deserves attention. Because of the empirical nature of the current da/dN approach, different physical meanings can be attached to the same crack growth rate data. The final decision will invariably favor the best fit of the data to certain pre-conceived condition. Non-uniqueness of the physical interpretation must therefore be kept in mind. Often, it is the a priori assumption based on physics that determines the outcome. Some of the possible differences are discussed with emphases placed on using an equivalent crack length and a corresponding energy density state such that the segmented state of affairs at the different scales can be connected.
The time history of material damage including the degradation of the microstructure should be the primary concern in determining the system integrity. Scale segmentation avoids over stretching the ability of theories, most of which are short ranged. The dual scale approach will be offered as the basis for developing multiscale models. Advent of the modern electronic computer offers the opportunity for optimizing several unknown variables at the same time to arrive at the desired solution that can correspond to the design specification.

• 出版日期2008-10
• 单位东华大学