Rotating Element Defect Detection by Order-Based Vibration Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting defects in rotating elements, such as roller bearings, struggle with low signal-to-noise ratios and non-linear distortion, making it difficult to accurately diagnose defects using solid-borne sound analysis, especially when rotational speeds are variable.
Innovation Solution
An apparatus and method that form multiple analysis functions based on vibration signals and estimated rotational speeds, using correlation functions to examine the state of rotating elements, allowing for defect detection even at low signal-to-noise ratios and variable rotational speeds by correlating vibration signals with harmonic analysis functions and varying time shifts and frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional amplitude demodulation is used for envelope detection, then the method is simple to implement, but the signal-to-noise ratio requirement is high (+10 dB minimum) and non-linear distortion occurs
Solution Approach 1:
The patent changes the fundamental parameter of the detection approach by replacing amplitude demodulation with correlation-based detection. Instead of detecting envelope amplitude variations, the system correlates the vibration signal with reference signals at different frequencies and orders, transforming the detection mechanism to operate effectively at much lower signal-to-noise ratios (down to -60 dB).
Solution Approach 2:
The patent substitutes the mechanical/envelope detection approach with a signal processing correlation approach. By using correlation functions instead of amplitude demodulation, the system replaces a method that requires high signal energy with one that can detect weak signals through statistical correlation, eliminating the non-linear distortion inherent in envelope detection.
2Measurement precision
If integral transforms (Hilbert, Fourier, Wavelet) are used to improve signal-to-noise ratio, then detection sensitivity increases, but the method requires constant rotational speed for up to half an hour
Solution Approach 1:
The patent introduces dynamic adaptation by making the correlation analysis frequency-dependent on the current rotational speed. Instead of requiring constant speed over long periods, the system continuously adjusts the reference frequency based on the measured rotational speed, allowing defect detection to adapt to speed variations in real-time while maintaining high signal-to-noise ratio performance.
Solution Approach 2:
The patent performs preliminary correlation analysis at multiple frequencies and orders before selecting the optimal detection parameters. By pre-calculating correlation functions across a range of frequencies and identifying the most promising defect indicators, the system prepares the detection framework in advance, enabling rapid adaptation when rotational speed changes without requiring long constant-speed periods.
3Device complexity
If conventional methods are used, then the system works with standard equipment, but defect detection is impossible at signal-to-noise ratios below +10 dB
Solution Approach 1:
The patent introduces correlation functions as an intermediary between the raw vibration signal and the defect detection process. By correlating the vibration signal with reference signals at specific frequencies and orders, the system acts as a mediator that amplifies weak defect-related components while suppressing noise, enabling detection at signal-to-noise ratios as low as -60 dB without requiring additional sensors or complex hardware.
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
This approach enhances diagnostic reliability by improving the signal-to-noise ratio and enabling early detection of defects in rotating elements, even under conditions of temporally variable rotational speeds, without requiring additional sensors for rotational speed measurement.
Implementation Method 1
correlation means for calculating a correlation function, the correlation means being configured to calculate a correlation result from the vibration signal and each analysis function
Data Source
AI summary
An apparatus for analyzing an element rotating at a rotational speed based on an vibration signal originating from the rotating element has: an analysis function processor for forming a plurality of analysis functions, the analysis function processor configured to form each analysis function based on an analysis frequency, on the vibration signal or on a measured or estimated rotational speed and on a predetermined order factor differing for each analysis function; a correlator configured to calculate a correlation result from the vibration signal and each analysis function, each correlation result being associated to the order factor which the analysis function, with which the correlation result was calculated, is based on, and the correlation results representing a correlation function; and an examiner configured for examining the correlation function or examining information derived from the correlation function so as to make a statement on a state of the rotating element.