S. Fouvry

Quantification of fretting damage

Abstract The fretting behaviour of high speed steel SC 6-5-2 uncoated and coated with a TiN coating against an alumina ball was studied based on a fretting map approach. Cracking and material loss were observed depending on the contract loading. To quantify the damage, a methodology is proposed based on an elastic Hertzian-Mindlin contact description. The sliding regimes are clearly defined, applying some fretting sliding criteria. First appearing under partial slip conditions, the crack nucleation conditions were precisely identified for different number of cycles and normal forces. Comparison with various fatigue approaches indicates that, if the local friction coefficient as well as the local aspect of contact stressing are considered, the maximum principal and the equivalent Von Mises stresses permit a rather good quantification of cracking. Wear and debris formation mainly observed under gross slip conditions lead to a wear volume which is shown to present a linear evolution with the cumulated dissipated energy. This interfacial work approach is completed by a local description. Axial and lateral sliding axis linear energies are deduced allowing a rapid identification of the wear behaviour through different energy wear factors. Experimental comparison of the TiN coating with the steel substrate contact shows that, as long as the coating is present in the contact, the important compressive residual stresses prevent cracking whereas the energy wear analysis indicates a wear resistance of the coating up to 10 times superior to the studied steel.

An energy description of wear mechanisms and its applications to oscillating sliding contacts

Abstract To quantify wear rates, the Archard approach is classically applied. It relates the wear volume to the product of the sliding distance and the normal load. A wear coefficient is then extrapolated and is supposed to establish the wear resistance of the studied material. This synthesis shows that this approach does not work when the friction coefficient is not constant. It appears to be much more relevant to consider the interfacial shear work as a significant wear parameter. This approach is applied to study the wear response of different steels and then extended to different hard TiN, TiC coatings under reciprocating sliding conditions. By identifying wear energy coefficients the wear quantification can be rationalized and the wear resistance of the studied tribosystems can be classified. This also appears to be a convenient approach to interpret the different wear mechanisms. Metallic materials involving plastic strain are analyzed from FEM computations. The energy balance confirms that a minor part of the dissipated energy is consumed by plasticity, whereas the major part participates in the heat and debris flow through the interface. When a load energy approach is introduced an accumulated density of the dissipated energy variable is considered to quantify the tribologically transformed structure (TTS) formation. A wear “scenario” of metallic structures is then discussed. This energy wear approach is applied to analyze hard coating wear mechanisms focusing on abrasion and oxidation phenomena. The local wear energy analysis is transposed, thus allowing the lifetime of hard coatings to be quantified.

Analysis of sliding behaviour for fretting loadings: determination of transition criteria

Abstract The determination of the sliding transition is of primary importance in describing the loading conditions in a contact. Various damages have been observed depending on the sliding regime. Assuming Mindlin hypotheses, an analysis for an elastic ball on flat contact submitted to a constant normal force and a varying tangential force is made. Both the energy and the sliding ratio are studied in order to quantify the fretting behaviour. They present a theoretical constant at the transition but their experimental determination is dependent on the tangential accommodation of the testing apparatus. Nevertheless, by combining the two former variables it is possible to deduce a system free criterion. This analysis is completed by considering the discontinuity of the transition through differential calculations. Finally an expression of the boundary between the partial and gross slip is given to describe the transition in fretting maps (RCFM). These theoretical results are compared with experiments obtained for two different tribo-systems HSS-Cr-steel and HSS-TiN (5μm)-Cr-steel and various loading conditions. Good correlation is observed between theoretical and experimental results if the tangential accommodation of the apparatus is taken into account.

Wear analysis in fretting of hard coatings through a dissipated energy concept

Abstract Fretting situations are becoming increasingly important for industrial applications. Numerous studies have been undertaken to understand the damage created by low-amplitude reciprocating motion, particularly in the case of a ball on flat contact in dry conditions. The loading condition of materials can be characterized by a “running condition fretting map” which describes, in a graph of normal load versus displacement amplitude, the various domains of sliding; partial or gross slip. By considering the time evolution of sliding conditions, various regimes can be defined: a partial slip regime when partial slip is observed throughout the test, a gross slip regime for gross slip,regime for gross slip, and a mixed regime when both partial and gross slip are observed. The response of the material is, of course, dependent on the sliding regime: crack initiation and propagation are related more to the mixed regime, while surface transformations and loss of matter are related more to the gross slip regime. This paper analyzes the fretting gross slip situation and describes the loss of matter in the case of hard coatings deposited on high-speed steel. Two aspects are considered: (i) the wear volume measurements are usually performed using 3D topography, whereas modeling of the wear scar morphology indicates that valuable estimates of the worn volume can be obtained by 2D profilometry, which is much less time-consuming than 3D topography; and (ii) the wear volume observed in experiments is related to the energy dissipated in the contact at both the global scale and the local scale where the wear depth is an important parameter. With this energy approach, the wear behaviour of a TiN-coated and uncoated high-speed steel can be compared quantitatively.

The fretting sliding transition as a criterion for electrical contact performance

Abstract The fretting phenomenon in electrical contacts is a plague in electrical connector industries. It induces severe electrical distortions, high electrical contact resistance and micro cuts. A specific analysis combining different fretting tests on several coated and uncoated CuSn4 bronze substrates has been conducted. The experimental investigation showed a direct correlation between the fretting sliding condition and the electrical contact performance. The stabilized partial slip condition is sufficient to maintain a low and stable electrical resistance independent on the contact dimension and geometry. On the other hand, gross slip turns out to be disastrous for oxygen-sensitive materials. Noble material coatings, which prevent the formation of an oxidized debris layer, can only delay the electrical distortion. When the substrate is reached, high and unstable electrical resistance is again observed. Therefore, to predict the electrical performance of connectors it appears essential to determine the sliding transition amplitude between the stabilized partial slip condition and stabilized gross slip condition. FEM elasto-plastic analyses have been conducted on a 2D cylinder/plane CuSn4 contact. It was shown that the sliding transition and consequently the electrical performance of the contact can conveniently be predicted if the cyclic hardening behavior of the material and the friction law of the tribosystem are correctly identified.

A fretting crack initiation prediction taking into account the surface roughness and the crack nucleation process volume

This paper presents an experimental study of the fretting crack nucleation threshold, expressed in terms of loading conditions, with a cylinder/plane contact. The studied material is a damage tolerant aluminium alloy widely used in the aerospace application. Since in industrial problems, the surface quality is often variable, the impact of a unidirectional roughness is investigated via varying the roughness of the counter body in the fretting experiments. As expected, experimental results show a large effect of the contact roughness on the crack nucleation conditions. Rationalisation of the crack nucleation boundary independently of the studied roughnesses was successfully obtained by introducing the concept of effective contact area. This does show that the fretting crack nucleation of the studied material can be efficiently described by the local effective loadings inside the contact. Analytical prediction of the crack nucleation is presented with the Smith-Watson-Topper (SWT) parameter and size effect is also studied and discussed.

An energy description of hard coating wear mechanisms

Numerous hard coatings have been developed to improve the wear resistance of sliding components under dry or lubricated conditions. High microhardness is usually required to resist abrasion wear mechanisms. However, microhardness cannot be considered as a sufficient parameter to design a tribological application. Wear analysis and friction coefficient identification are systematically required. To quantify the wear kinetics, the Archard approach is classically applied. It relates the wear volume with the product of the sliding distance and the normal load. A wear coefficient is then extrapolated and is supposed to establish the wear resistance of the material studied. The present study shows that this approach does not work when the friction coefficient is not constant. It appears to be much more relevant to consider the interfacial shear work dissipated through the contact as a significant wear parameter. This approach is transposed in the most convenient way to study the wear response of a sintered steel and then extend it to the study of different coatings, TiN, TiCVC, and a high-speed steel substrate for gross-slip fretting conditions. The identification of wear-energy coefficients permits classification of the wear resistance of the tribosystems studied. This involves identifying the partition of the total dissipated energy consumed through the different damage processes, and relating that with physical variables, such as the binding energy, the energy activation for oxidation and mechanical energy.

Effect of shot peening on the fretting wear of Ti–6Al–4V

Abstract In this paper, we report on the fretting wear behaviour of polished and shot peened Ti–6Al–4V specimens. For fretting experiments, due to micro-displacements at the interface between two contacting surfaces, two types of damage can be observed: crack initiation and debris formation. Shot peening, which is already well known for improving fatigue resistance of titanium alloys, is shown to have a beneficial effect on the crack initiation and propagation under fretting wear loading, as cracks observed on specimens after cylinder-on-flat fretting tests are shorter in shot peened specimens than in polished ones. It is also demonstrated that shot peening decreases the friction coefficient only at the beginning of the test, as long as the asperities induced by shot peening are not worn-off. The effects of displacement amplitude, normal force and test duration on the wear volume have been investigated: in all cases, shot peening has no significant impact on the wear process. The same amount of debris are formed and ejected for both polished and shot peened specimens. Moreover, it is found that, for both types of specimens, the linear relation, developed for steels and hard coatings, between wear volume and cumulated dissipated energy is not valid in the present case as different wear volumes are measured for the same cumulated dissipated energy, depending on the experimental conditions (normal force, displacement amplitude). Using the test duration as the variable parameter, energy wear coefficients are calculated for different experimental conditions.

An elastic–plastic shakedown analysis of fretting wear

Abstract Wear induced by debris formation is studied under fretting and reciprocating sliding conditions. A sintered powder steel alloy is tested. Different compaction conditions inducing various surface porosities change the quantity of oil trapped during the final oil quenching and consequently modify the friction behavior. The study reveals a strong variation of the wear coefficient in relation to the friction coefficient. A low wear regime is observed for low friction coefficients, whereas high wear coefficients are observed for high friction values. This evolution is analyzed considering elastic, elastic shakedown and plastic shakedown conditions which are imposed on the material. The wear behavior of the studied tribosystem concludes that to better identify the intrinsic wear resistance of a material, the energy approach, which consists of a comparison of the wear volume to the cumulated dissipated energy, is more appropriate.

Theoretical analysis of fatigue cracking under dry friction for fretting loading conditions

Abstract The development of wear maps has enhanced a new approach to the wear behaviour of materials as it is a precise description of the mechanical loading (local stress and strain fields). This paper studies crack nucleation using fatigue criteria in the case of an elastically loaded dry contact. Fretting behaviour is analysed for partial slip conditions which have been identified experimentally and theoretically (transition criteria developed using Mindlins analysis). A Dang Van crack nucleation criterion is introduced for a ball-flat contact. Typical analytical expressions of the criterion were obtained on the surface and in the volume depending on the loading conditions and the contact system. They are expressed as a function of the mechanical properties, the friction coefficient, the contact geometries and the fatigue limits under both shear and tensile loadings of the material. The results obtained permit the investigation of the position and the direction of the first crack created by fretting conditions.