Kirk, Adam
(2021)
The role of debris in fretting: investigation into the interrelated effects of displacement amplitude and frequency in fretting wear.
PhD thesis, University of Nottingham.
Abstract
This thesis concerns an investigation of the role of debris in fretting contacts, which has been highlighted in fretting literature over several decades to be critical in affecting the development of wear in fretting contacts; retained debris plays a significant role in fretting contacts due to a significant proportion of the wearing interface remains covered throughout the displacement cycle, thus restricting the transport into the contact of active species involved in the formation of wear debris (e.g. oxygen) and the transport out of the contact of debris once it is formed. The processes involved in the formation, flow and expulsion of oxide debris in fretting contacts are examined via the effects of frequency and displacement amplitude on fretting wear of a high strength steel, two parameters which are widely reported to influence critical processes of debris formation and ejection, but to have significantly different impacts on these processes. Examining these two parameters together therefore enables insight to be gained into processes of debris formation, retention in the form of compacted beds and the ejection of material from the contact (i.e. wear).
The fretting behaviour of a cylinder-on-flat configuration was examined over a range of fretting frequencies and displacement amplitudes, and the nature of wear debris generated in fretting tests investigated, both in the form of compacted beds retained in the contact and particles ejected during the wear process, as well as the conditions affecting the formation of debris beds.
Extensive characterisation of the surface and subsurface regions of wear scars was conducted to gain insight into the mechanisms by which oxide debris beds form and the impact of their formation (or otherwise) on the development of the tribologically transformed structure.
At all conditions examined, debris ejected from the contact was found to consist predominantly of haematite, with a small (< 6%) fraction of metallic iron that exhibits a modest dependence on both frequency and displacement amplitude. Likewise, the overall range of particle sizes is largely independent of frequency and displacement amplitude, ranging from approximately 0.4 μm to 50 μm under all conditions tested, although some large pieces of the order of 100 μm are observed, proposed to be pieces of compacted debris beds formed in the contact which become detached and are subsequently ejected. Examining the morphology of ejected debris particles showed particles to consist of sub-micron crystallites of iron oxide, of the order of 0.1 μm in diameter, which sinter together to form larger coherent structures. The size of these crystallites is observed to be largely independent of displacement amplitude, and as such it is proposed that the impact of the examined parameters on the size of debris particles expelled from the contact is due to their influence upon the sintering of debris particles within the contact and their subsequent ejection, as opposed to the mechanism of detachment of particles from first body surfaces.
Two distinct wear regimes are identified, namely (i) at low frequencies and displacement amplitudes, an oxide debris bed covers most of the worn surface, preventing the development of subsurface damage and adhesive transfer but resulting in relatively high wear rates, and (ii) at high frequencies and displacement amplitudes oxide coverage of worn surfaces is relatively sparse, resulting in metal-to-metal contact and the development of significant subsurface damage and adhesive transfer. The significant change in the nature of debris retained within wear scars contrasts with the nature of debris ejected from the contact, which does not exhibit a strong dependence upon the examined parameters; as such it is proposed that wear occurs by the formation and removal of oxide debris in all of the cases examined, with any metallic debris formed being retained within the contact.
The effects of displacement amplitude and frequency in the mechanisms of fretting wear are found to exhibit a significant degree of interdependence, with the magnitude of either parameter having a significant impact upon the effect of the other on the wear rate and damage mechanism. It is proposed that the interacting effects of frequency and displacement amplitude arise from the impact of both parameters on three key processes, namely (i) ingress of oxygen to the contact; (ii) formation of oxide debris; (iii) expulsion of oxide debris from the contact. A model is outlined based on these processes to calculate wear rates and determine the operative wear mechanism under a given set of conditions; a key supposition of the model is that these three processes are in competition with one another, with the observed wear rate being governed by whichever of these rates is the smallest, accordingly termed the “rate-determining process”. Equations relating these key processes to fretting parameters are derived based on current understanding of the physics of fretting wear, enabling wear rates and mechanisms (i.e. the operative processes that control the overall wear rate) to be predicted.
The model is implemented using a simple time-marching method, calibrated against experimental data for the set of frequencies and displacement amplitudes examined in the experimental investigation. The alignment of predicted wear rates and mechanisms with experimental data is assessed over a broad range of fretting parameters, including observations reported in the literature employing the same specimen configuration and material combination. Relatively good agreement is observed between the mechanisms predicted using this model and those observed in laboratory experiments, although the accuracy of wear rate predictions is seen to vary considerably. The alignment of the model with experimental data provides a promising indication of the capability of the proposed physical framework to model fretting based on the physical processes of wear, although significant errors in the predicted wear rates at some conditions highlight the need for further investigation of the effects of parameters on debris behaviour, in order to improve the accuracy and applicability of the model going forward.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
