Payne, Jessica (2023). Qualitative condition monitoring of rail steel. University of Birmingham. M.Sc.
|
Payne2023MScByRes.pdf
Text - Accepted Version Available under License All rights reserved. Download (15MB) | Preview |
Abstract
To meet the growing demand for rail travel and keep customers satisfied, disruption to services must be minimised. Currently used rail inspection techniques require rail access which prevents 24h operation of rolling stock. The replacement of these disruptive techniques with passive remote monitoring methods can provide economic benefits to those who use and supply rail travel whilst improving customer satisfaction.
Acoustic Emission (AE) is an inspection technique that can passively and continuously collect health degradation data while attached to the surface of a material. It has previously been identified as capable of detecting and locating growing defects within rails and crossings and recent work has focused on its use to quantify fatigue cracks as they propagate. Active inspection can be used to quantify defects that have previously propagated whilst AE testing can provide information on damage as it occurs. To predict future health degradation, it is necessary to use modelling. Recent work regarding the modelling of rails have included improvements to the accuracy of modelling the material behaviour whilst keeping the processing effort reasonably low (due to the large number of cycles that rails operate over). Links have also started to be made between Finite Element Analysis (FEA) and AE, with a goal of aiding AE data interpretation.
A three-point bending fatigue test was performed on ten notched and pre-cracked samples of cast manganese steel (otherwise known as Hadfield steel). AE was measured by two sensors, the outputs of which were parametric and waveform analysis respectively. The crack length was measured throughout the test using Direct Current Potential Drop (DCPD). A similar experiment was performed in recent work on R260steel samples; data from the experiment was further analysed in this work. The waveform AE data and recorded force for each event was used to calibrate the parametric AE data to the true frequency so that a more accurate cycle number for each event was used and the cycle stage at which hits were measured could be analysed. For the Hadfield steel, the parametric AE data was compared to the crack growth rate to attempt quantification using the AE count rate. Finite Element Analysis was also performed to simulate a crack growing through an inclusion in R260 steel and investigate the resultant energy release for comparison with previous work in the literature that related the location of an inclusion along the surface of a fatigue crack to the time at which a significant step in cumulative AE energy was measured.
In both Hadfield steel and R260 steel, AE events that remained after filtering were determined to primarily be generated by crack propagation and friction as the crack opened and closed. Events that were measured during loading (due to crack propagation and friction of surfaces during crack opening) occurred earlier in the cycle as the crack grew, whilst events that were measured during unloading (due to friction during crack closure) appeared to occur at a consistent cycle stage.
Contrary to what was previously observed for R260 steel samples, the Hadfield steel samples showed no relationship between the count rates or energy rates and the stress intensity factor range (ΔK). As a result, no link could be made between the measured AE parameters and the crack growth. Therefore, crack length quantification using AE could not be achieved for the Hadfield steel samples. The FEA inclusion model showed that soft MnS inclusions can cause fast crack propagation and a related step in energy released. This supported the suggestion in previous work that an observed step in cumulative AE energy was a result of fast crack propagation caused by an inclusion.
Finally, a digital twin for rail is conceptualised which uses inputs of measured mechanical loads and AE for a section of track to reproduce defect growth and material behaviour. Implementation of an inspection technology such as this, if accurate, would vastly reduce inspection and maintenance costs whilst raising confidence in rail health.
Type of Work: | Thesis (Masters by Research > M.Sc.) | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Award Type: | Masters by Research > M.Sc. | ||||||||||||
Supervisor(s): |
|
||||||||||||
Licence: | All rights reserved | ||||||||||||
College/Faculty: | Colleges (2008 onwards) > College of Engineering & Physical Sciences | ||||||||||||
School or Department: | School of Metallurgy and Materials | ||||||||||||
Funders: | Other | ||||||||||||
Other Funders: | University of Birmingham, Engineering and Physical Sciences Research Council, European Commission | ||||||||||||
Subjects: | T Technology > TF Railroad engineering and operation T Technology > TN Mining engineering. Metallurgy |
||||||||||||
URI: | http://etheses.bham.ac.uk/id/eprint/13418 |
Actions
Request a Correction | |
View Item |
Downloads
Downloads per month over past year