Material Degradation Prediction Using Microstructure and LMP Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for evaluating metal material degradation in power plants and mechanical facilities require destructive testing, which is inefficient and disruptive.
Innovation Solution
A degradation prediction device and method that uses a non-destructive approach by analyzing microstructure images and environmental data to predict a degradation index and Larson-Miller Parameter (LMP) value through a regression model, employing a bag of images, non-parametric clustering, and machine learning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If destructive testing is used to evaluate metal material degradation, then measurement precision of degradation level is improved, but the material is destroyed and cannot be used again
Solution Approach 1:
The patent creates a digital replica (virtual model) of the metal material's microstructure through image processing and machine learning. This digital copy allows for repeated degradation assessment without physically damaging the actual material. The system captures microstructure images, processes them through clustering algorithms, and generates a virtual representation that can be analyzed multiple times while the physical material remains intact and usable.
2Measurement precision
If expert visual observation with electron microscopes is used for degradation assessment, then measurement precision is improved, but the process is time-consuming and requires specialized expertise
Solution Approach 1:
The patent replaces the mechanical/optical system of electron microscope observation with an automated image processing and machine learning system. Instead of relying on expert visual inspection through complex microscope equipment, the system uses digital image processing, clustering algorithms, and regression models to automatically assess degradation. This substitution dramatically reduces assessment time while maintaining or improving measurement precision through consistent algorithmic analysis.
Solution Approach 2:
The system enables self-service degradation assessment by automating the entire evaluation process. The machine learning models automatically process microstructure images, perform clustering analysis, extract degradation indices, and generate assessments without requiring expert intervention. This automation eliminates the time-consuming nature of manual expert observation while providing consistent, repeatable results.
3Reliability
If conventional degradation evaluation methods are used, then reliability of equipment operation is maintained through regular assessment, but the destructive nature disrupts normal operation and incurs high costs
Solution Approach 1:
By creating and analyzing a digital replica of the material's microstructure, the system enables degradation assessment without removing or damaging the actual component. The virtual model allows for continuous monitoring and evaluation while the equipment remains in service, maintaining both reliability through regular assessment and productivity through uninterrupted operation.
Data Source
AI summary
Disclosed is a degradation prediction device predicting a lifetime of a target material, including: at least one processor, and the at least one processor is configured to train a degradation index prediction model based on the representative value of a degradation index of a material by environmental conditions and the environmental information of a material subjected to a destructive testing, train a LMP (Larson-Miller Parameter) value prediction model based on the representative value of a degradation index of a material by environmental conditions and a theoretical value of LMP at a destructive testing, predict a degradation index for a target material using the degradation index prediction model for which training is completed based on environmental information of the target material and predict a LMP value of the target material using the LMP value prediction model for which training is completed based on the predicted degradation index.