Cristobal Ruiz-Carcel

Use of Spectral Kurtosis for Improving Signal to Noise Ratio of Acoustic Emission Signal from Defective Bearings

The use of acoustic emission (AE) to monitor the condition of roller bearings in rotating machinery is growing in popularity. This investigation is centered on the application of spectral kurtosis (SK) as a denoising tool able to enhance the bearing fault features from an AE signal. This methodology was applied to AE signals acquired from an experimental investigation where different size defects were seeded on a roller bearing. The results suggest that the signal to noise ratio can be significantly improved using SK.

A Comparative Study of the Effectiveness of Adaptive Filter Algorithms, Spectral Kurtosis and Linear Prediction in Detection of a Naturally Degraded Bearing in a Gearbox

AbstractDiagnosing bearing faults at the earliest stages is critical in avoiding future catastrophic failures. Many techniques have been developed and applied in diagnosing bearings faults; however, these traditional diagnostic techniques are not always successful when the bearing fault occurs in gearboxes where the vibration response is complex; under such circumstances, it may be necessary to separate the bearing signal from the complex signal. In this paper, an adaptive filter has been applied for the purpose of bearing signal separation. Four algorithms were compared to assess their effectiveness in diagnosing a bearing defect in a gearbox, least mean square (LMS), linear prediction, spectral kurtosis and fast block LMS. These algorithms were applied to decompose the measured vibration signal into deterministic and random parts with the latter containing the bearing signal. These techniques were applied to identify a bearing fault in a gearbox employed for an aircraft control system for which endurance tests were performed. The results show that the LMS algorithm is capable of detecting the bearing fault earlier in comparison with the other algorithms.

Application of Linear Prediction, Self-Adaptive Noise Cancellation, and Spectral Kurtosis in Identifying Natural Damage of Rolling Element Bearing in a Gearbox

The ability to detect and diagnose faults in rolling element bearings is crucial for modern maintenance schemes. Several techniques have been developed to improve the ability of fault detection in bearings using vibration monitoring, especially in those cases where the vibration signal is contaminated by background noise. Linear prediction (LP) and self-adaptive noise cancellation (SANC) are techniques which can substantially improve the signal to noise ratio of a signal, improving the visibility of the important signal components in the frequency spectrum. Spectral kurtosis (SK) has been shown to improve bearing defect identification by focusing on the frequency band with higher level of impulsiveness. In this paper, the ability of these three methods to detect a bearing fault is compared, using vibrational data from a specially designed test rig that allowed fast natural degradation of the bearing. The results obtained show that the SK was able to detect an incipient fault in the outer race of the bearing much earlier than any other technique.

Size differentiation of a continuous stream of particles using acoustic emissions

Procter and Gamble (P&G) requires an online system that can monitor the particle size distribution of their washing powder mixing process. This would enable the process to take a closed loop form which would enable process optimization to take place in real time. Acoustic emission (AE) was selected as the sensing method due to its non-invasive nature and primary sensitivity to frequencies which particle events emanate. This work details the results of the first experiment carried out in this research project. The first experiment involved the use of AE to distinguish sieved particle which ranged from 53 to 250 microns and were dispensed on a target plate using a funnel. By conducting a threshold analysis of the peaks in the signal, the sizes of the particles could be distinguished and a signal feature was found which could be directly linked to the sizes of the particles.

Effectiveness of adaptive filter algorithms and spectral kurtosis in bearing faults detection in a gearbox

Bearing faults detection at the earliest stages is vital in avoiding future catastrophic failures. Many traditional techniques have been established and utilized in detecting bearing faults, though, these diagnostic techniques are not always successful when the bearing faults take place in gearboxes where the vibration signal is complex; under such circumstances it may be necessary to separate the bearing signal from the complex signal. The objective of this paper is to assess the effectiveness of an adaptive filter algorithms compared to a Spectral Kurtosis (SK) algorithm in diagnosing a bearing defects in a gearbox. Two adaptive filters have been used for the purpose of bearing signal separation, these algorithms were Least Mean Square (LMS) and Fast Block LMS (FBLMS) algorithms. These algorithms were applied to identify a bearing defects in a gearbox employed for an aircraft control system for which endurance tests were performed. The results show that the LMS algorithm is capable of detecting the bearing fault earlier in comparison to the other algorithms.

Estimation of Fine and Oversize Particle Ratio in a Heterogeneous Compound with Acoustic Emissions

The final phase of powder production typically involves a mixing process where all of the particles are combined and agglomerated with a binder to form a single compound. The traditional means of inspecting the physical properties of the final product involves an inspection of the particle sizes using an offline sieving and weighing process. The main downside of this technique, in addition to being an offline-only measurement procedure, is its inability to characterise large agglomerates of powders due to sieve blockage. This work assesses the feasibility of a real-time monitoring approach using a benchtop test rig and a prototype acoustic-based measurement approach to provide information that can be correlated to product quality and provide the opportunity for future process optimisation. Acoustic emission (AE) was chosen as the sensing method due to its low cost, simple setup process, and ease of implementation. The performance of the proposed method was assessed in a series of experiments where the offline quality check results were compared to the AE-based real-time estimations using data acquired from a benchtop powder free flow rig. A designed time domain based signal processing method was used to extract particle size information from the acquired AE signal and the results show that this technique is capable of estimating the required ratio in the washing powder compound with an average absolute error of 6%.

Bearing Natural Degradation Detection in a Gearbox: A Comparative Study of the Effectiveness of Adaptive Filter Algorithms and Spectral Kurtosis

Bearing faults detection at the earliest stages is vital in avoiding future catastrophic failures. Many traditional techniques have been established and utilized in detecting bearing faults, though, these diagnostic techniques are not always successful when the bearing faults take place in gearboxes where the vibration signal is complex; under such circumstances it may be necessary to separate the bearing signal from the complex signal.The objective of this paper is to assess the effectiveness of an adaptive filter algorithms compared to a Spectral Kurtosis (SK) algorithm in diagnosing a bearing defects in a gearbox. Two adaptive filters have been used for the purpose of bearing signal separation, these algorithms were Least Mean Square (LMS) and Fast Block LMS (FBLMS) algorithms.These algorithms were applied to identify a bearing defects in a gearbox employed for an aircraft control system for which endurance tests were performed. The results show that the LMS algorithm is capable of detecting the bearing fault earlier in comparison to the other algorithms.Copyright

WITHDRAWN: Estimation of powder mass flow rate in a screw feeder using acoustic emissions

Screw feeders are widely used in powder processes to provide an accurate and consistent flow rate of particles. However this flow rate is rarely measured or controlled. This investigation explores the use of generalised norms and moments from structural-borne acoustic emission (AE) measurements as key statistics indicators for the estimation of powder mass flow rate in a screw feeder. Experimental work was carried out acquiring AE measurements from an industrial screw feeder working with four different types of material at different dispensation rates. Signal enveloping was used in first place to eliminate high frequency components while retaining essential information such as peaks or bursts caused by particle impacts. Secondly a set of generalised norms and moments is extracted from the signal, and their correlation with mass flow rate was studied and assessed. Finally a general model able to estimate mass flow rate for the four different types of powders tested was developed.

Size Differentiation Of A Continuous Stream Of Particles Using Acoustic Emissions

Procter and Gamble (P&G) require an online system that can monitor the particle size distribution of their washing powder mixing process. This would enable the process to take a closed loop form which would enable process optimisation to take place in real time. Acoustic Emission (AE) was selected as the sensing method due to its non-invasive nature and primary sensitivity to frequencies which particle events emanate. This work details the results of the first experiment carried out in this research project. This experiment involved the use of AE to distinguish between the sizes of sieved polyethylene particle (53-250microns) and glass beads (150-600microns) which were dispensed on a target plate using a funnel. By conducting a threshold analysis of the impact peaks in the signal, the sizes of the particles could be distinguished and a signal feature was found which could be directly linked to the sizes of the particles.

Improved condition monitoring using fast-oscillating measurements

In this paper, a technique of merging typical process data with variables containing fast periodic oscillations is proposed for the purpose of detecting faults in industrial systems working under variable operating conditions. Analysing windows of the fast-oscillating signals allowed key features to be extracted from the data at the same rate at which the process variables are sampled. This allows the fusion of both types of data acquired at different sampling rates in a single data matrix. The data is then analysed using canonical variate analysis (CVA) looking for deviations in any parameter that can point at a fault in the system. The dynamic characteristics of CVA allow the detection and diagnosis of faults in systems working under variable operating conditions. This approach was tested using experimental data acquired from a compressor test rig where the compressor surge process fault. Results suggest that the combination of both types of data can effectively improve the detectability of faults in systems working under variable operating conditions.