Search academic texts

Information × Registration Number 0221U106805, 0117U000523 , R & D reports Title Development of scientific bases for diagnosing of steel's technical condition by quantitative estimates of fracture damage features of structural elements operated under the influence of corrosive environments popup.stage_title Head Vorobel Roman A, д.т.н.Student Oleksandra Zynoviyivna, д.т.н. Registration Date 21-12-2021 Organization Physico-Mechanical Institute named after GV Karpenko of the National Academy of Sciences of Ukraine popup.description2  A set of experimental physical and mechanical studies to assess the high-temperature degradation of low-alloy Cr-Mo-V steel of TPP steam pipes has been carried out. The availability and non-destructive nature of the characteristics gives grounds to recommend the dependence of hardness on the grain size of the steel used to diagnose the current technical condition of steels of steam pipes during routine inspections of their external surfaces. It is established that the characteristics of brittle fracture resistance are the most sensitive to the assessment of degradation of steam material. Computerized methods for detection and quantitative analysis of ductile fracture dimpls, inclusions on their bottoms and traces of them on fractographic images of heat-resistant steels have been developed. It is shown that the assessment of the degree of steel degradation is possible by the ratio of the number of dimpls with inclusions on their bottom relative to the total number of available dimpls. It is established that the common signs of high-temperature degradation of low- and high-alloy heat-resistant steels at the structural level are the separation and coagulation of carbides at the grain boundaries. The dependences between mechanical and fractographic parameters for low- (12Х1МФ and 15Х1М1Ф) and high-alloy (15Х11МФ) heat-resistant steels are constructed. A method for automated estimation of fatigue striation spacing of heat-resistant steels based on the fractogram image has been developed. A method for classifying images of brittle and ductile fractures of heat-resistant steels using deep learning neural networks has been developed, which has made it possible to achieve classification accuracy of 89%. Product Description popup.authors Berehulyak Olena R. Vorobel Roman A. Zvirko Olʹha Ivanivna Krechkovsʹka Halyna V. Kurnat Ivan M. Mandziy Teodor S. Nykyforchyn Hryyorij M. Svirsʹka Lesya M. Solovej Petro R. Student Oleksandra Zynoviyivna Tsyrulnyk Oleksandr T popup.nrat_date 2021-12-21 Close