Infrared Sensor Thermal Data Machine Learning Crack Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting defects, such as cracks, in aircraft structures under repetitive stress are time-consuming and expensive, often requiring manual inspection and back-calculations to estimate defect onset, which are not efficient or timely.
Innovation Solution
A system utilizing an infrared sensor to monitor structural components and process thermal data with machine learning models, specifically unsupervised and supervised techniques, to detect the onset and propagation of cracks, providing automated and timely defect identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual inspection and back-calculations are used to detect defects, then defect detection can be performed, but the process is time-consuming and expensive
Solution Approach 1:
The patent replaces manual mechanical inspection methods with an automated optical sensing system. An infrared sensor captures thermal data from the structure during fatigue cycling, and a processor automatically analyzes the thermal data to detect defects. This substitution eliminates the need for manual inspection and back-calculations, significantly reducing inspection time while maintaining defect detection capability.
2Measurement precision
If manual inspection methods are used, then defect detection is possible, but highly skilled engineers are required making the process expensive
Solution Approach 1:
The system enables self-service defect detection by automatically capturing thermal data during fatigue cycling and processing it through machine learning models. The processor independently identifies defects without requiring highly skilled engineers to perform manual inspection and back-calculations. This automation reduces dependency on expensive expert personnel while maintaining accurate defect detection.
3Measurement precision
If traditional inspection methods are used, then defects can be detected, but it is rare to discover a defect at the moment it occurred
Solution Approach 1:
The system performs preliminary action by continuously monitoring the structure with infrared sensing during fatigue cycling before defects propagate to detectable sizes. The thermal data is captured in real-time during the fatigue process, allowing defects to be detected at or near their onset moment rather than at later stages during periodic inspections. This preserves the timing information of when defects occurred.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system enables automated and efficient detection of crack onset and propagation, reducing the need for manual inspection and back-calculations, thereby improving the speed and cost-effectiveness of defect identification in aircraft structures.
Implementation Method 1
monitoring the structure with an infrared sensor to provide thermal data of the structure within a field of view of the infrared sensor
Data Source
AI summary
Exemplary embodiments of a system and method are provided for detecting cracks and crack propagation in aircraft structures subject to repetitive stress. A method for detecting onset or propagation of defects in a structure includes monitoring the structure with an infrared sensor to provide thermal data of the structure within a field of view of the infrared sensor. A processor is used to process the thermal data memory to extract features from the thermal data and utilize at least one machine learning model to detect onset or propagation of defects in the structure. A system includes an infrared sensor having a field of view of the structure and a processor coupled to the infrared sensor and a memory, which contains instructions that cause the processor to process thermal data from the infrared sensor to extract features from the thermal data and utilize at least one machine learning model to detect onset or propagation of the defects in the structure.