Localization of steel fractures based on the fractal model of their metallographic images

Abstract

There are a number of tasks that require assessment of the condition of the material and its mechanical characteristics. Such tasks may arise at the production stage, when there’s a need to control the content of various components of the material, strength, hardness, etc. Also similar tasks arise during exploitation of materials, which is especially relevant today, when most of the responsible products and structures in the field of nuclear energy, chemical industry, machine-building industry are on the verge of wearing down. Previously defectoscopy methods were mainly used to assess the reliability of such materials and products. These methods provided information on the presence or absence of a defect. But to prevent accidents, information about the pre-defective state of the material itself and the degree of its degradation is needed. Approaches involving methods and means of solid state physics, mechanics, chemistry, materials science and other scientific disciplines have become more informative for describing the state of degradation. However, these methods are quite laboursome and time consuming and cannot be applied to transient processes. Therefore, it is important to develop a method that would be based on the analysis of the microstructure of the material would allow to obtain its numerical mechanical characteristics. This approach would be used at the production stage of materials to determine their components and mechanical characteristics and at the stage of exploitation to determine the degree of degradation of the material. It is known that the fractal dimension of each microstructure of the material is an indicator of its qualitative characteristics. Thus, the numerical value of the fractal dimension establishes the relationship between the structure and the mechanical properties of the material. In this work the method of localization of fractures of heat-resistant steels on the basis of fractal models of metallographic images is developed and its advantages in comparison with other known approaches are analyzed.