In situ blocked force measurement in gearboxes with potential application for condition monitoring

Abohnik, AA 2018, In situ blocked force measurement in gearboxes with potential application for condition monitoring , PhD thesis, Salford University.

[img] PDF - Submitted Version
Restricted to Repository staff only until 5 May 2018.

Download (6MB) | Request a copy

Abstract

The use of gearboxes for power transfer is widespread throughout industry. However, machines today are operating at higher speeds than ever before and gear failures such as wear or tooth breakage is serious and legitimate concerns. Incipient fault detection in gears has thus become the subject of intensive investigation and at this stage of development, there are many competing condition monitoring methods based on vibration signal analysis.

This thesis summarizes the research steps taken after a review of (i) current maintenance strategies, (ii) gearbox condition monitoring techniques and gear vibration fundamental and common gearbox failure modes, (iii) new approach called in-situ blocked force.

A test rig was built, designed and fabricated for experimental data collection. The experimental work was carried out using a healthy spur gear and one suffering from tooth breakage with two levels of faults; 25% and 85%.

This study reports the use of the blocked force to characterize the gear mesh interactions; the advantage is to remove the effect of the housing, and to get a signal, which is more representative of the source mechanisms from which it is generated. Under certain assumptions the blocked forces are an intrinsic property of the vibration source. For example, a given vibration source, such as gear, hypothetically operating in the same conditions could produce different vibration signals when installed in different housings, however, the blocked forces theoretically are the same in both cases.

The blocked force represents a property independent of the noise generating mechanisms and is therefore invariant to the gearbox housing. It is proposed that this invariance yields a signal more amenable to fault detection.

In this thesis, assessment of the condition of a gearbox in a test rig is based on vibration analysis but, contrary to standard condition monitoring techniques, this research uses the blocked force signal instead of acceleration signals.

FFT has applied to transform the time domain signal to frequency domain and identify the spectrum of the shaft and mesh frequency. However, the low pass filter has been applied to filter the signal above the 1000 Hz and subjected to statistical parameters.

Conventional parameters using the time domain of the vibration signal (kurtosis, RMS, crest factor and skewness) were used for detecting and diagnosing the faults by applying them to filtered signals. As a result, the noise might be removed or reduced but the effect of the housing remains. Then, the total energy was also applied to detect the presence of the faults by combination with EMD, and the results compared with those obtained by the conventional parameters.

However, the blocked force signal obtained through the inverse procedure was filtered using the same filter which was used for fault detection alternatively to conventional signal. Moreover, the aim was to use BF as a signal for condition monitoring purposes.

Parameters (namely Kurtosis, crest factor and total energy) were then applied to filtered BF signals to identify the condition of the machine. The results obtained based on filtered BF signals were compared to ones obtained based on conventional signals.

The comparison between the results obtained from the acceleration signals and BF signals shows that (a) the blocked force can be used to remove or eliminate the effect of the housing; (b) the trends of kurtosis and crest factor and total energy are more consistent with the severity of the fault. Additionally, the parameters applied to the blocked force signals can offer more effective way of all those tested to detect faults than conventional acceleration signal at least for this case study.

Item Type: Thesis (PhD)
Schools: Schools > School of Computing, Science and Engineering > Salford Innovation Research Centre (SIRC)
Depositing User: Alsdeg A Abohnik
Date Deposited: 05 Apr 2018 08:15
Last Modified: 05 Apr 2018 08:15
URI: http://usir.salford.ac.uk/id/eprint/46144

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)

Downloads

Downloads per month over past year