High-Pressure Tank AE Waveform Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional pressure testing methods for high-pressure tanks do not effectively detect signs of destruction during pressurization, risking tank breakage and damage to testers and peripherals.
Innovation Solution
A pressure testing method that extracts and classifies acoustic emission waveforms using a machine-learned classifier to differentiate between macrocracks and microcracks, stopping pressurization when a predetermined threshold of macrocrack activity is reached to prevent tank destruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional AE-based pressure testing is used, then the testing method is simple, but it cannot detect destruction signs and may cause tank breakage
Solution Approach 1:
The patent replaces conventional mechanical/electronic AE analysis methods with a machine learning-based waveform classification system. The classifier automatically distinguishes between safe and dangerous waveforms, enabling reliable detection of destruction signs while maintaining operational simplicity through automated decision-making algorithms.
Solution Approach 2:
The patent introduces a machine learning classifier as an intermediary between AE sensor detection and pressure control decisions. This intermediary processes raw AE waveforms, identifies dangerous patterns, and triggers appropriate responses, bridging the gap between simple detection and complex safety decision-making.
2Manufacturing precision
If pressurization continues to predetermined pressure, then testing completeness is improved, but tank destruction risk increases
Solution Approach 1:
The patent implements real-time feedback control by continuously monitoring AE waveforms during pressurization and dynamically adjusting the testing process. When dangerous waveforms are detected, the system provides feedback to stop or reduce pressurization, ensuring testing completeness while preventing tank destruction through adaptive control.
Solution Approach 2:
The patent performs preliminary classification of AE waveforms to identify dangerous patterns before they lead to catastrophic failure. By detecting and responding to early signs of macrocrack development, the system takes preliminary protective action to prevent complete tank destruction while maintaining testing integrity.
3Measurement precision
If AE waveforms are classified in real-time, then destruction detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies partial classification by focusing computational resources on identifying the most critical dangerous waveform patterns rather than performing exhaustive analysis of all waveform characteristics. This selective approach maintains high detection accuracy for destruction signs while reducing overall processing time through targeted analysis.
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
Accurately determines the state of a high-pressure tank before potential destruction, preventing tank breakage and ensuring safety during testing by stopping pressurization based on the classification of AE waveforms.
Implementation Method 1
an AE sensor configured to detect AE waves generated in the high-pressure tank while increasing a pressure inside the high-pressure tank
Data Source
AI summary
Provided is a pressure testing method for a high-pressure tank capable of avoiding a destruction of the high-pressure tank during a pressure test. A pressure testing method includes: extracting a plurality of AE waveforms from output waveforms of an AE sensor while increasing a pressure inside the high-pressure tank; and testing the high-pressure tank based on the extracted plurality of AE waveforms. The method includes: while increasing the pressure inside the high-pressure tank, classifying the extracted AE waveforms into first waveforms and second waveforms with a classifier that is machine-learned so as to classify the plurality of AE waveforms into the first waveforms derived from a macrocrack that increases immediately before destruction of the high-pressure tank, and the second waveforms derived from a microcrack smaller than the macrocrack; and stopping pressurization of the high-pressure tank based on the number of the first waveforms.