Quantitative Analysis Of Fracture Surfaces Research Articles

Deep learning based initial crack size measurements utilizing macroscale fracture surface segmentation

To this date the safety assessment of materials, used for example in the nuclear power sector, commonly relies on a fracture mechanical analysis utilizing macroscopic concepts, where a global load quantity K or J is compared to the materials fracture toughness curve. Part of the experimental effort involved in these concepts is dedicated to the quantitative analysis of fracture surfaces, which is a cumbersome and time-consuming process. In addition, however, it is affected by subjective bias. Within the scope of our study, we have introduced a methodology that employs deep learning models for macroscopic fracture surface segmentation. To fight subjective bias and enable improved reproducibility, we set up three distinct and unique datasets labeled by different experts that correspond to typical isolated laboratory conditions and complex real-world (multi-laboratory) scenarios. To demonstrate the benefit of our approach for the fracture mechanics assessment, we employed the models for initial crack size measurements utilizing the area average method. Furthermore, the influence of structural similarity on the segmentation capability, differing due to the miscellaneous materials, specimen types, as well as imaging-induced variance has been analyzed. For semi-supervised learning a weak-to-strong consistency regularization was implemented. We were able to train robust and well-generalizing models that learned feature representations from images across different domains without observing a significant drop in prediction quality. Our approach reduced the number of labeled images required for training by a factor of 6. The deep learning assisted measurements proved to be as precise as manual measurements. For the multi-laboratory data, mean measurement deviations smaller 1 % could be achieved, showcasing the enormous potential of our approach.

Toward quantitative fractography using convolutional neural networks

The science of fractography revolves around the correlation between topographic characteristics of the fracture surface and the mechanisms and external conditions leading to their creation. While being a topic of investigation for centuries, it has remained mostly qualitative to date. A quantitative analysis of fracture surfaces is of prime interest for both the scientific community and the industrial sector, bearing the potential for improved understanding on the mechanisms controlling the fracture process and at the same time assessing the reliability of computational models currently being used for material design. With new advances in the field of image analysis, and specifically with machine learning tools becoming more accessible and reliable, it is now feasible to automate the process of extracting meaningful information from fracture surface images. Here, we propose a method of identifying and quantifying the relative appearance of intergranular and transgranular fracture events from scanning electron microscope images. The newly proposed method is based on a convolutional neural network algorithm for semantic segmentation. The proposed method is extensively tested and evaluated against two ceramic material systems (Al2O3,MgAl2O4) and shows high prediction accuracy, despite being trained on only one material system (MgAl2O4). While here attention is focused on brittle fracture characteristics, the method can be easily extended to account for other fracture morphologies, such as dimples, fatigue striations, etc.

The mechanical properties of heat-treated rocks: a comparison between chert and silcrete

In archaeology, heat treatment of stone is the process of “making” a new material for tool production. Its invention in the African Middle Stone Age was an important step in the evolution of transformative technologies and the cultural evolution of early humans in general. Although the chemical and crystallographic transformations in silica rocks, the only material class heat-treated in the Stone Age, begin to be well known, many of the mechanical transformations and their chemical origins remain a subject of controversy. The difference between different silica rock categories is also only poorly understood. In this paper, we investigate the thermally induced changes of three mechanical properties in the two silica rock types chert and silcrete: fracture strength, indentation fracture resistance (approximating fracture toughness) and elastic modulus. These tests are complemented by statistical analyses (Weibull modulus) and a quantitative fracture surface analysis. The results show that heat treatment transforms these silica rocks in terms of their fracture toughness and the uniformity of fracture. A comparison with published data on the structural transformations in the same samples identified the loss of chemically bound water and subsequent defect healing to be the chemical mechanism behind these mechanical transformations. These findings have important implications for the study of the interactions between chemical and structural processes and the mechanics of natural rocks or ceramics.

MICROMECHANISMS OF FERRITE-PEARLITE STEELS FRACTURE UNDER CYCLIC AND IMPACT LOADING

During operation, the railway axle is exposed to static, cyclic and dynamic loading that is accompanied by influence of corrosive environments. This causes the accumulation of structural and mechanical damages in the material, formation of corrosion and mechanical micro-defects that are potential sites of fatigue crack nucleation and growth. This requires the development of failure analysis methods that allow determining main mechanisms of ferrite-pearlite steels fracture on the micro scale level. The railway axle’s material – the OSL steel – main regularities of temperature influence on fracture mechanisms in ferrite-pearlite steel type are shown. Qualitative and quantitative analysis of fracture surface of Charpy specimens tested at 20°C and -40°C was performed on micro scale level using SEM investigation.

Correlative Fractography: Combining Scanning Electron Microscopy and Light Microscopes for Qualitative and Quantitative Analysis of Fracture Surfaces

Correlative fractography is a new expression proposed here to describe a new method for the association between scanning electron microscopy (SEM) and light microscopy (LM) for the qualitative and quantitative analysis of fracture surfaces. This article presents a new method involving the fusion of one elevation map obtained by extended depth from focus reconstruction from LM with exactly the same area by SEM and associated techniques, as X-ray mapping. The true topographic information is perfectly associated to local fracture mechanisms with this new technique, presented here as an alternative to stereo-pair reconstruction for the investigation of fractured components. The great advantage of this technique resides in the possibility of combining any imaging methods associated with LM and SEM for the same observed field from fracture surface.

파면거칠기 유도 균열닫힘에 의한 혼합모드 피로균열의 전파거동 및 파면에 대한 평가

In this study, we have investigated the closure behavior of fatigue cracks in SAPH440 steel plates under a mixed-mode I+II loading. A crack image capturing system as a direct measuring method was used to measure the closure levels at a crack tip. The crack closure levels in the fluctuation and stable sections were increased with the increase of the mode mixture ratio. The mode mixture ratio independent fatigue crack propagation rates equation was calculated by considering mixed-mode crack closure levels. The equation was examined according to the application method of crack opening ratio. The fracture surface analysis by C-scan method was also performed in order to investigate the closure mechanism and propagation mode of crack under the mixed-mode I+II loading. The crack closure under the mixed mode I+II is confirmed as a surface roughness closure by the quantitative analysis of fracture surface using the proposed surface roughness parameter.

The Characterization of Local Deformation and Fracture Properties – a Tool for Advanced Materials Design

Novel experimental techniques have been recently developed to measure local deformation and fracture properties. The techniques comprise quantitative fracture surface analysis and in‐situ loading experiments in combination with digital image analysis. Examples are presented, demonstrating how these methods can be used to investigate the damage evolution in materials, to improve tools for the numerical simulation, and for the design of more fracture resistant materials and components.

The effect of inclusion size on the local conditions for void nucleation near a crack tip in a mild steel

The local conditions for void initiation are determined for individual MnS-inclusions near the crack tip in a mild steel St37. First, the crack tip opening displacement at the moment of void initiation, CODvi, is measured by quantitative fracture surface analysis; subsequently, the maximum interfacial stress, σmaxinterf, is calculated from models by Argon et al. [Metall Trans 1975;6:825] and Beremin [Metall Trans 1981;12:723]. A strong effect of the size of the MnS-inclusions on the CODvi- and the σmaxinterf-values is revealed.

ランレングス統計量による球状黒鉛鋳鉄の定量的破面解析

Fractography is the most popular method for failure analysis. This method, however, has been mainly applied to structural steels, and has not been always effective to cast irons. Since many factors like graphite, matrix, cast defects, and so on affect fracture of cast iron, they made it difficult to estimate failure causes from the fracture surfaces. In this study, therefore, quantitative fracture surface analysis was carried out on spheloidal graphite cast iron to improve accuracy of the estimation.When fracture surface analysis is conducted on images obtained with scanning electron microscopes (SEM) that are commonly used in the failure analysis fields, the characteristic values extracted from the images can possibly depend upon the contrast of the images. To minimize the contrast influence, we proposed the fracture surface analysis method that edges are extracted from the SEM images by differentiating them and fracture surface analysis is conducted on the texture images. The proposed method was applied to the fracture surfaces formed by a fracture toughness test. By computing run length statistics which are parameters widely used in texture analysis, we could divide the fatigue fracture surface produced in introducing a pre-fatigue crack from the rapid fracture surface produced during the fracture toughness test.

Characterization of self-affine surfaces from 3D digital reconstruction

A new experimental method aiming to the characterization of self-affine rough surfaces is presented. Based on a three dimensional reconstruction of the relief from stereo pairs obtained with a scanning electron microscope, it allows a non-destructive quantitative analysis of fracture surfaces, with a xy and z resolution only slightly lower than the original images. The roughness index of the fracture surface of an aluminum alloy is recovered, as well as the Hurst exponents of reconstructed computer-generated fractional Brownian reliefs.

Quantitative analysis of fracture surface by roughness and fractal method

In recent years there has been extensive research and great development in Quantitative Fractography, which acts as an integral part of fractographic analysis. A prominent technique for studying the fracture surface is based on fracture profile generation and the major means for characterizing the profile quantitatively are roughness and fractal methods. By this way, some quantitative indexes such as the roughness parameters R{sub L} for profile and R{sub S} for surface, fractal dimensions D{sub L} for profile and D{sub S} for surface can be measured. Given the relationships between the indexes and the mechanical properties of materials, it is possible to achieve the goal of protecting materials from fracture. But, as the case stands, the theory and experimental technology of quantitative fractography are still imperfect and remain to be studied further. Recently, Gokhale and Underwood et al have proposed an assumption-free method for estimating the surface roughness by vertically sectioning the fracture surface with sections at an angle of 120 deg with each other, which could be expressed as follows: R{sub S} = {ovr R{sub L}{center_dot}{Psi}} where {Psi} is the profile structure factor. This method is based on the classical sterological principles and verified with the aid of computer simulationsmore » for some ruled surfaces. The results are considered to be applicable to fracture surfaces with any arbitrary complexity and anisotropy. In order to extend the detail applications to this method in quantitative fractography, the authors made a study on roughness and fractal methods dependent on this method by performing quantitative measurements on some typical low-temperature impact fractures.« less

Fractography: Determining the sites of fracture initiation

Fractography is the analysis of fracture surfaces. Here, it refers to quantitative fracture surface analysis (FSA) in the context of applying the principles of fracture mechanics to the topography observed on the fracture surface of brittle materials. The application of FSA is based on the principle that encoded on the fracture surface of brittle materials is the entire history of the fracture process. It is our task to develop the skills and knowledge to decode this information. There are several motivating factors for applying our knowledge of FSA. The first and foremost is that there is specific, quantitative information to be obtained from the fracture surface. This information includes the identification of the size and location of the fracture initiating crack or defect, the stress state at failure, the existence, or not, of local or global residual stress, the existence, or not, of stress corrosion and a knowledge of local processing anomalies which affect the fracture process. The second motivating factor is that the information is free. Once a material is tested to failure, the encoded information becomes available. If we decide to observe the features produced during fracture then we are rewarded with much information. If we decide to ignore the fracture surface, then we are left to guess and/or reason as to the cause of the failure without the benefit of all of the possible information available. This paper addresses the application of quantitative fracture surface analysis to basic research, material and product development, and “trouble-shooting” of in-service failures. First, the basic principles involved will be presented. Next, the methodology necessary to apply the principles will be presented. Finally, a summary of the presentation will be made showing the applicability to design and reliability.

Advanced quantitative analysis of fracture surfaces

A new classification of fracture surfaces based on the modified degree of orientation is presented. The proposed classification system is more accurate and allows a better choice of parameters for further analysis. In particular it covers fracture surfaces with overlaps and secondary cracks. As secondary cracks have hardly been considered in the existing studies, the main features such as main and secondary cracks and overlaps are precisely defined. To quantity them a method of dividing the fracture profile into the main profile and secondary cracks is developed and described. New parameters such as the mean secondary crack length and density of secondary cracks are defined and their estimators are given. The problem of real fracture surface area is discussed with respect to the classification introduced. Corrected boundary values for the relations between linear and surface roughnesses are derived and various relations for evaluation of the real fracture surface area are discussed. Two simple parameters for quantifying particles observed in fracture surfaces are defined and their possible application is briefly discussed.

New Developments in Fracture Surface Analysis Using X-Ray and Laser

Recent developments in fracture surface examination are surveyed. Emphasis is placed on quantitative analysis of fracture surfaces. An X-ray analysis of a fracture surface is first described. The successful use of the residual stress distribution under the fracture surface is demonstrated for the estimation of the applied stress intensity factor under which the fracture occurred. A new technique using a laser profile meter is described for the measurement of fracture surface topographical information. A bird's-eye view of fracture surfaces is presented. The successful use of topographical information is demonstrated for matching of the conjugate fracture surfaces.

Theoretical basis for the quantitative analysis of fracture surfaces

A theory is proposed for the determination of several basic quantities in fractographic analysis. These are the true area of a fracture surface, the surface fraction, and the volume fraction of some identifiable features intersected by a random-curvature fracture surface. The objective of this analysis is to make it possible for fractographers to take all necessary measurements on the micrographs of the analyzed fracture surfaces. It is shown that simple and exact relations can be derived when such micrographs are taken at a zero-tilt angle in the scanning electron microscope or prepared by surface replication and transmission electron microscopy. There is excellent agreement between conclusions based on the proposed theory and experimental observations.