Description:
TECHNOLOGY AREA(S): Materials
OBJECTIVE: Troubleshoot machine operation issues through ‘Monitoring and Diagnosis via Machinery Vibration Auditing
DESCRIPTION: Machine faults can be diagnosed by the changes of the system parameters or modal parameters, such as the natural frequency, damping, stiffness, etc. Since most manufacturing process generates vibrations, vibration analysis plays a major role in detecting machinery degradation before the equipment fails and potentially damages other related equipment for the ultimate purpose of avoiding unwanted breakdowns and downtime. Vibration analysis can help increase the lifetime of equipment when degradation is detected and then dealt with at an early stage. Vibration analysis of a rotating table top model has shown that some faults might exist even though they are not visible to the naked eye. The statistical features of the vibration signals in time, frequency and time–frequency domains have different representation capabilities for fault patterns. Singularity point detection, fault feature extraction, weak signal extraction, and system identification can be implemented based on vibration signals. A sophisticated vibration-fault relation model can be developed based on the vibration feature analysis. Industry has performed extensive work on smart seismic networks and data analytics through collaboration with geophysicists from NASA, USGS, energy exploration industry and academic. Characteristics of machinery vibration signals (including amplitude, frequency, phase) can be efficiently extracted using signal decomposition methods. Industry has also developed innovative intrinsic oscillation mode analysis or signal feature extraction methods, which can directly apply to machine health monitoring and diagnosis. Advanced signal processing and machine learning methods could be explored to enhance the sensitivity, robustness, reconstruction accuracy, classification specificity, and efficiency.
PHASE I: Develop statistical feature extraction methods from vibration sensor signals. Time domain measurement and the corresponding frequency domain spectrum are capable of separately describing machinery vibration in terms of time and frequency. For jointly representing vibration features, it would be required to extract time-frequency domain features for signal processing and analysis.
PHASE II: Develop monitoring and diagnosis software via vibration auditing. Based on the features extracted from Phase I, focus would be channeled toward optimizing statistical features in different domains from different types of faults in different diagnostic applications. Once a relationship between vibration features and faults is built, root cause diagnosis can be discerned based upon vibration signals. Then, fault diagnosis experiments on real devices could be conducted. Thereafter, based on the vibration-fault model, typical machinery systems could be constructed to validate the proposed approach.
PHASE III: Monitoring and diagnosis via vibration auditing would have many commercial applications. A successful system could be marketed to commercial manufacturing, aerospace industry, as well as other defense customers. Additional markets might include “oil and gas” and homeland security.
REFERENCES:
1. Peng, Z. K., and F. L. Chu. "Application of the wavelet transform in machine condition monitoring and fault diagnostics: a review with bibliography." Mechanical systems and signal processing 18, no. 2 (2004): 199-221.; 2. Yan, Ruqiang, Robert X. Gao, and Xuefeng Chen. "Wavelets for fault diagnosis of rotary machines: A review with applications." Signal processing 96 (2014): 1-15.; 3. Li, Chuan, René-Vinicio Sánchez, Grover Zurita, Mariela Cerrada, and Diego Cabrera. "Fault diagnosis for rotating machinery using vibration measurement deep statistical feature learning." Sensors 16, no. 6 (2016): 895.KEYWORDS: Manufacturing System, Monitoring And Diagnosis, Vibration
