Journal of Achievements of Materials and Manufacturing Engineering · 2007 · 16 citations · 7 references
EngineeringMaterials EngineeringMaterials ScienceMachine LearningMaterials CharacterizationMetallurgyDamage MechanicsFailure AnalysisImage AnalysisCreep ResistanceDamage Evolution VisualizationImage-based Condition AssessmentImage-based Damage DetectionImage-based Damage RecognitionLow Alloy SteelsMaterial Failure Analysis
Purpose: The aim of the paper is to present the computer assisted method for analysis of scanning electron microscope metallographic images of elements after long time operating in creep conditions. Design/methodology/approach: The development of internal damages in low alloyed chromium-molybdenum steels were discussed and illustrated. As a tools of evaluation of the internal damages’ classes of power industry installation elements from 16Mo3 steel operating in creep conditions a method based on image analysis, geometrical coefficient and artificial neural networks was used. Findings: Combining of several methods making use of the image analysis, geometrical coefficients and neural networks will make it possible to achieve the better efficiency of class recognition of damages developed in the material. Practical implications: This method can find a practical application for evaluation and classification of creepdamages of elements power industry installations. Originality/value: The use of images analysis and neural networks to identification and classification of internal damages of low alloyed chromium-molybdenum steels working in creep conditions.
7
L. A. Dobrzański, Michał Kowalski, J. Madejski · Journal of Materials Processing Technology · 2005 · 61 citations
Natural Sciences, Materials Engineering, Materials Science +16