Goertzel Order Analysis for Continuous Machine Fault Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current order analysis methods for machine condition monitoring are limited in providing comprehensive information about machine health, particularly in detecting issues like resonance and misalignment, especially during periodic motion changes.
Innovation Solution
An order analysis system that includes an input module for receiving periodic motion and machine characteristic data, a processor module applying a Goertzel algorithm to create an order data set showing magnitude and phase of machine characteristics versus normalized periodic motion, and a display for presenting this data graphically.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional order analysis methods are used, then basic frequency analysis is provided, but comprehensive machine health information including resonance and misalignment detection is limited
Solution Approach 1:
The patent segments the machine characteristic data into different order components (first order, second order, third order, etc.) corresponding to different harmonics of the periodic motion. This segmentation allows detailed analysis of each harmonic component to detect specific machine conditions such as unbalance (first order), misalignment (second order), and resonance (higher orders), thereby providing comprehensive machine health information without losing critical diagnostic details.
Solution Approach 2:
The patent transforms the analysis from traditional frequency domain to order domain by normalizing frequencies relative to the periodic motion frequency. This dimensional transformation enables the system to track and analyze machine characteristics across varying operating speeds, providing a new perspective that reveals machine health issues that are not apparent in conventional frequency analysis.
2Reliability
If run-up or run-down tests are performed to detect resonance, then resonance conditions can be observed, but continuous monitoring during steady operation is not achieved
Solution Approach 1:
The patent implements a dynamic order analysis system that continuously adapts to changing operating conditions by normalizing frequency data relative to the instantaneous periodic motion frequency. This dynamic approach allows the system to detect resonance conditions at any operating speed, not just during run-up or run-down tests, enabling continuous monitoring during steady-state operation while maintaining resonance detection accuracy.
Solution Approach 2:
The system performs continuous order analysis throughout the machine's operation, including during steady-state operation, by continually normalizing the periodic motion frequency and analyzing machine characteristic data across all orders. This continuous analysis provides uninterrupted monitoring capability, allowing resonance and other machine health issues to be detected at any point in the operating cycle, thereby eliminating the need for separate run-up or run-down test procedures.
3Measurement precision
If detailed analysis of all harmonic components is performed, then comprehensive machine state information is obtained, but system complexity increases
Solution Approach 1:
The patent extracts and analyzes only the relevant order components from the machine characteristic data, focusing on specific harmonics that correspond to known machine faults (e.g., first order for unbalance, second order for misalignment). This selective extraction approach provides comprehensive machine state information by concentrating on diagnostically significant frequencies while avoiding the complexity of analyzing all frequency components in detail.
Solution Approach 2:
The order analysis system serves multiple diagnostic functions simultaneously by analyzing machine characteristic data across all orders. A single analysis pass provides information about unbalance (first order), misalignment (second order), resonance (higher orders), and other machine conditions, eliminating the need for separate specialized tests and reducing overall system complexity while maintaining comprehensive diagnostic capability.
Data Source
AI summary
An instrument for performing order analysis on a rotational machine. An input module receives periodic motion data, and machine characteristic data that is associated by time with the periodic motion data. A processor module receives the periodic motion data and the machine characteristic data, applies a Goertzel module to the periodic motion data and the machine characteristic data, and thereby creates an order data set comprising magnitude of machine characteristic data versus order of normalized periodic motion data.