Matrix replica method and artificial neural networks as a component of condition assessment of materials for the power industry

Purpose: use of matrix replica method and artificial neural networks in condition assessment of materials working under creep conditions. Demonstration of scanning electron microscopy usefulness for correct assessment of the occurrence of internal damages. Design/methodology/approach: For material after long-term service the microstructural investigations were carried out on metallographic specimens and matrix replicas using light and scanning electron microscope. A computer program was used for material condition assessment. Findings: The way of material condition assessment by matrix replica method using a computer program as an example of the application of computer materials science methods was presented. The correct development of methodology for assessment of the degree of internal damages in steel working under creep conditions was found a result of verification. Critical comments on condition assessment of structure observed using light microscopy were presented. Practical implications: The presented method can be used for evaluation and qualification of structural changes in power station boiler components operating in creep conditions. Originality/value: The presented results of changes in the mechanical properties, structure and in the precipitation processes are applied to evaluation the condition of the elements in further industrial service.