Gas Turbine Core Rub Diagnostics Using Multi-Signal Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current engine diagnostics systems for gas turbine engines are inadequate in detecting and categorizing various types and severities of engine rubs, as they often rely on single-pronged analyses that can only detect a subset of rub types and fail to predict rubs before they occur, leading to reactive maintenance and inadequate categorization.
Innovation Solution
A system employing a three-pronged vibration analysis using machine learning pattern recognition to compare fundamental mode placements, vibration spectra, and operation parameters with baseline data, generating a rub indicator to detect, predict, and categorize engine rubs, and an on-ground digital twin for proactive maintenance recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single-pronged vibration analysis is used, then the system complexity is reduced, but the rub detection accuracy and categorization capability deteriorate
Solution Approach 1:
The patent divides the vibration analysis into three distinct analytical prongs: (1) vibration spectrum analysis examining frequency content and peaks, (2) time-domain waveform analysis studying amplitude patterns and temporal characteristics, and (3) operational parameter correlation analysis matching vibration data with engine operating conditions. Each prong independently evaluates different aspects of rub detection, and their results are integrated to achieve comprehensive and accurate rub categorization while maintaining manageable system complexity through modular architecture.
2Device complexity
If a single-pronged vibration analysis is used, then the device complexity is reduced, but the ability to predict and categorize various types of engine rubs deteriorates
Solution Approach 1:
The diagnostic system is segmented into three independent analytical prongs that each evaluate different characteristics of engine rubs: vibration spectrum analysis for frequency-based patterns, time-domain waveform analysis for amplitude and temporal patterns, and operational parameter correlation for context-aware interpretation. This segmentation enables the system to detect and categorize multiple rub types (light rubbing, heavy rubbing, blade tip rubbing, rotor-stator rubbing) with high reliability while keeping each analytical module relatively simple and manageable.
Solution Approach 2:
The patent merges the results from three separate analytical prongs into a unified diagnostic conclusion. By combining vibration spectrum data, time-domain waveform characteristics, and operational parameter correlations, the system achieves comprehensive rub detection and prediction capability that exceeds what any single analysis method could provide alone, while maintaining device complexity through efficient integration architecture.
3Ease of operation
If traditional vibration analysis methods are used, then the ease of operation is maintained, but the productivity of maintenance operations deteriorates due to reactive maintenance
Solution Approach 1:
The system performs preliminary diagnostic actions by continuously analyzing vibration data and operational parameters to detect early signs of rub conditions before they escalate into serious failures. The three-pronged analysis methodology identifies patterns indicative of light rubbing, heavy rubbing, blade tip rubbing, or rotor-stator rubbing in advance, enabling maintenance teams to schedule interventions proactively rather than reactively, thereby improving maintenance productivity while keeping the system easy to operate through automated analysis.
Data Source
AI summary
Systems and techniques that facilitate predictive core rub diagnostics are provided. A sensor component can collect real-time operation parameters of a gas turbine engine. An analysis component can statistically combine first values, second values, and third values to yield a rub indicator for the engine. The first values can be based on a first comparison of fundamental mode placements of the engine, derived from the real-time operation parameters, and baseline fundamental mode placements. The second values can be based on a second comparison of a vibration spectrum of the engine, derived from the real-time operation parameters, and a baseline vibration spectrum. The third values can be based on a third comparison of the real-time operation parameters of the engine and baseline operation parameters. A classification component can generate a rub classification report indicating presence of rubbing in the engine, based on the rub indicator.