Component Distress Prediction Using Iterative Damage Modeling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for predicting distress on physical components, particularly in high-stress environments like aircraft engines, lack accuracy and reliability, leading to unplanned maintenance and operational disruptions.

Innovation Solution

A method and system for predicting distress on physical components by obtaining distress data, determining a distress rank, incorporating environmental and operational data, formulating a kernel using a cumulative damage model, and iteratively tuning a predictive model until the difference between the distress output and rank is within a threshold.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If existing prediction methods are used, then the process is simple, but the prediction accuracy and reliability are insufficient

Engineering Contradiction:
Improveprediction accuracyVSAvoidmodel complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The prediction system is segmented into distinct functional modules: distress data acquisition module, environmental data acquisition module, operational data acquisition module, kernel formulation module using cumulative damage model, and iterative tuning module. This segmentation allows complex prediction tasks to be divided into manageable components, improving accuracy while maintaining systematic organization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The predictive model employs dynamic parameter adjustment through iterative tuning, where kernel parameters are continuously optimized based on the difference between predicted distress output and actual distress rank. This dynamic adaptation enables the system to improve prediction accuracy by adjusting parameters in response to performance feedback.

Inventive Principle:
Principle #15Dynamics

2Reliability

If a simple model is used, then the model is easy to implement, but it cannot provide accurate distress predictions

Engineering Contradiction:
Improveprediction reliabilityVSAvoidtuning process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system implements a feedback mechanism where the predicted distress output is compared against actual distress rank, and the difference (error) is used to adjust kernel parameters. This closed-loop feedback process continues iteratively until the prediction error falls within an acceptable threshold, thereby improving prediction reliability through continuous optimization.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The model performs preliminary actions by formulating the kernel using cumulative damage models before iterative tuning begins. This preliminary structuring based on established damage theories provides a solid foundation that guides subsequent parameter optimization, making the complex tuning process more systematic and reliable.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If multiple data types are incorporated, then the prediction comprehensiveness improves, but the data processing complexity increases

Engineering Contradiction:
Improvedistress measurement accuracyVSAvoiddata integration complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system merges multiple data sources including distress data, environmental data, and operational data into a unified predictive framework. By combining these diverse data types through the kernel formulation, the model achieves comprehensive distress assessment that leverages information from all sources simultaneously, improving measurement accuracy through data integration.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4707963A1Model for predicting distress on a component
Publication Date: 2026.03.11 GENERAL ELECTRIC CO
  • EP4707963A1 patent drawingFigure 1
  • EP4707963A1 patent drawingFigure 2
  • EP4707963A1 patent drawingFigure 3

AI summary

An apparatus (10) and method (100) for predicting distress on a physical component (30). The method (100) can include obtaining (102) distress data. The distress data can be used to determine (104) a distress rank. The distress rank can be compared (112) to a distress output provided by a kernel (108) that use parameters related to the physical component (30). The comparison (112) can result in a predictive model (120) for the physical component (30).