Ring crack propagation in silicon nitride under rolling contact

Silicon nitride has been found to have the optimum combination of properties which are suitable for rolling element bearing applications to withstand high loads, severe environments, and high speeds. Surface ring cracks are difficult to detect but are found on the surface of silicon nitride balls. These ring crack defects decrease the rolling contact fatigue life considerably. This paper presents an experimental study and numerical analysis of ring crack propagation in rolling contact. The contribution of this study is to provide understanding of ring crack propagation behaviour and life prediction in rolling contact. Rolling contact tests are performed on the silicon nitride/steel elements. Silicon nitride ball surfaces are examined before testing using a dye-penetrant technique and optical microscopy. The surfaces are examined using optical microscopy and scanning electron microscopy during testing and after failure. The numerical calculations are based on a 3D model of ring crack growth. The rolling contact loading is simulated by a repeated Hertzian surface load with normal pressure and tangential traction. Fracture mechanics analysis is utilised to determine the stress intensity along the crack front and the stress intensity factors are analysed using a 3D boundary element model. Life predictions from the present calculations are in line with the experimental observations.     

Materials failure mechanisms of hybrid ball bearings with silicon nitride balls

A modified Shell four-ball apparatus was used to determine failure mechanisms and estimate time to failure of hybrid ball bearings using silicon nitride and zirconia balls. The machine was set up to hold a hybrid bearing containing three ceramic balls, instead of the usual 14 in order to increase contact pressure. Five kinds of silicon nitride ceramics, which differed in terms of surface roughness, porosity as well as the amount and chemical composition of additives were investigated. The goal of this study was to establish a quality criterion for silicon nitride for industrial use in hybrid ball bearings. Lifetime and failure mechanisms varied between the five bearings with silicon nitride balls and a dependance was found on the porosity and chemical composition of the materials, whereas surface roughness did not seem to influence their performance.     

Surface strength of silicon nitride in relation to rolling contact performance measured on ball-on-rod and modified four-ball tests

Silicon nitride (Si₃N₄) has been used in various rolling contact applications in turbomachinery, automotive and power industry. It is favoured to replace conventional steel due to its low density, low friction, corrosion resistance and good performance under extreme condition. However, a major limitation of its wider application is its high material and machining cost, especially the cost associated with the finishing process. In the present study, a low cost sintered and reaction bonded silicon nitride (SRBSN) is used to study the surface machining effects on its rolling contact performance. Attempt has been made to link the surface strengths of Si₃N₄ derived from half-rod and C-sphere flexure strength specimens to the rolling contact lifetimes of Si₃N₄ rod and ball specimens. The rolling contact fatigue tests are carried out on ball-on-rod and modified four ball machines. Three types of surfaces with coarse, fine and conventional finishing conditions are examined. Flexure strength tests on half-rod and C-sphere show an increasing surface strength from specimens with coarse, fine to conventionally machined conditions. During rolling contact fatigue test of as-machined specimens, there are no failures observed on both ball-on-rod and four ball tests after 100 million stress cycles. However, there is a trend of decreasing wear volumes measured on the contact path of rods and balls with coarse, fine and conventional conditions. In four ball test, spall failures are observed on pre-crack specimens. There is a trend of increasing rolling contact fatigue lifetime from pre-cracked specimens with coarse, fine to conventional machining conditions.     

The influence of oil additives on spread cracks in silicon nitride

The properties of ceramic materials (high hardness, high-temperature capability, low coefficient of thermal expansion) are of interest to rolling element materials. One of the most interesting materials is silicon nitride.The paper presents an experimental study of the influence of oil additives (Cl, S, P, cerium dioxide (CeO₂)) on spread cracks in silicon nitride. The additives Cl, S, P are bound in molecules in liquid form soluble in the base oil. The CeO₂ is purely in powder form in suspension. The use of CeO₂ powder was made based on the good results of polishing of silicon nitride. A ceramic angular contact ball bearing was modelled using a four-ball machine. Silicon nitride, -in diameters balls were artificially damaged with pre-cracks. Ring formed pre-cracks were propagated on the ball surface by a blunt impact load using a tungsten carbide ball under five various impacts.Rolling contact fatigue failure modes were studied under high contact stresses and speed. The surface of silicon nitride balls before and after failure was examined using scanning electron microscopy.The research of the present paper shows that fatigue failure under rolling contact loading is markedly sensitive on the size of pre-cracks and oil additives. Propagations of surface pre-cracks takes place under fatigue load. Primary and secondary surface crack propagation are described.     

Wear properties of silicon nitride in rolling contact

The wear properties of silicon nitride were examined in dry rolling contact.

The wear coefficient of silicon nitride in pure rolling was of the order of 10⁻⁶ at the initial stage of wear and of the order of 10⁻⁸ at the steady stage of wear under hertzian pressures of 1.06, 1.30, 1.50 and 1.83 GPa.

The wear coefficient of silicon nitride in rolling-sliding was of the order of 10⁻³ under hertzian pressures of 1.06, 1.50 and 1.83 GPa.

The original grinding marks were decreased by the initial wear. Then a very smooth surface appeared in the steady state and its centre-line average roughness R_a was 0.02 μm.

In contrast, pitting and the adhesive accumulation of thin film debris on the surface started to occur in the steady stage of wear.

Three typical types of wear debris were distinguished. One of these, which was a glassy film, was confirmed to have an SiO₂ structure.     

Influence of grease composition on rolling contact wear: Experimental study

This paper presents an experimental study on the influence of grease composition on rolling contact wear (RCW). Experimental tests for three greases and correspondent base oils were carried out on a twin-disc machine under pure rolling conditions. The following parameters were varied: base-oil viscosity, percentage of soap concentration and the presence of additives. The second aim of these tests is to analyse the use of artificial dents as a technique to evaluate RCW: prior to the tests, artificial dents were printed on one of the contacting surfaces. These dents were used to calculate the wear volume that is removed from the surfaces, based on the assumption that when the material is removed, the diameter of these dents diminishes. For each stop, images obtained using video-microscopy were analysed and the diameter of the dents, at a given number of cycles, was calculated. It was found that tests with different lubricants resulted in different damaged areas around the dents. Based on this, a comparison of this area growth throughout the fatigue cycles was used as a way to compare the influence of grease composition on RCW.     

Rolling contact fatigue of alumina ceramics sprayed on steel roller under pure rolling contact condition

The rolling contact fatigue of sprayed alumina ceramics with a nominal composition of Al₂O₃–2.3 mass% TiO₂ was studied with a two-roller test machine under a pure rolling contact condition with oil lubricant. The influence of undercoating of sprayed Ni-based alloy on the rolling contact fatigue was investigated. The failure mode of all sprayed rollers was spalling caused by subsurface cracking. The undercoating did not contribute to the improvement of the rolling contact fatigue life. The elastic modulus of the alumina sprayed layer evaluated with the nano-indentation method was around 85 GPa. The depths of the observed subsurface cracks corresponded approximately to the depths where the orthogonal shear stress or the maximum shear stress calculated with two-dimensional FEM became maximum.     

Effect of surface texturing on rolling contact fatigue within mixed lubricated non-conformal rolling/sliding contacts

The rolling contact fatigue (RCF) life of highly loaded machine components is significantly influenced by the surface roughness features so that there is a continuous effort to design the topography of rubbing surfaces to enhance lubrication efficiency and prolong the operation of machine components. It can be suggested from the recent experimental results that lubricant emitted from shallow micro-dents could effectively lift off the real roughness features and reduce the asperities interactions within rolling/sliding mixed lubricated contacts. Thereby the additional supply of lubricant from surface features could help to reduce the risk of surface damage through the reduction of the interaction of rubbing surfaces during start-up or starvation. However, the introduction of such roughness features into the rubbing surfaces of highly loaded non-conformal contacts should consider not only the effects on lubrication film thickness but also on RCF.That is why this study is focused on the effects of surface texturing on RCF within non-conformal rolling/sliding contacts operated under mixed lubrication conditions. The principal task has been whether possible beneficial effect on film thickness is not accompanied by the reduction in RCF life. Textures with various sizes of micro-dents and their arrangement within the contacts have been considered. It has been found that results obtained with textured surfaces have exhibited no obvious reduction in RCF. Conversely, some increase in RCF using textured surfaces was observed that could be attributed to the positive contribution of micro-dents working as lubricant micro-reservoirs that reduce asperities interactions. Nevertheless, further experiments are necessary to confirm this possible beneficial contribution of surface texturing on RCF.     

Wear mechanisms in oil-lubricated and dry fretting of silicon nitride against bearing steel contacts

The wear and friction behaviour of silicon nitride against bearing steel was investigated under lubricated and dry fretting conditions as a function of amplitude and test duration. Tests were performed on a high frequency fretting tester. Silicon nitride bearing balls were used as the upper oscillating specimens while the lower stationary flats were standard specimens of bearing steel. Amplitudes in the intermediate 5 to 50 μm range and a test duration from 10 to 360 min were studied. In lubricated conditions a commercial lubricant. ISO VG 220, was used. Light microscopy, scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), Auger spectroscopy (AES) and transmission electron microscopy (TEM) were employed to determine the wear mechanisms.

Under lubricated conditions transition from high to low wear volumes was recognised with increasing amplitude. At lower amplitudes and in the early stage of fretting tests at moderate amplitudes, mechanical wear dominated. Cracks on the stick-slip boundary and spalling of a thin tribolayer was observed. Under these conditions the highest wear in lubricated fretting was obtained. In the final stage of fretting tests at moderate amplitudes, and from the beginning at higher amplitude, tribochemical wear is suggested as the dominant wear form. A 0.2 μm thick tribolayer was observed on the contact, containing inclusions with different Fe and Si contents. A very high concentration of carbon, formed by oil degradation, was also determined in this layer, confirming the critical influence of oil on the wear behaviour.

Quite a different wear mechanism is proposed for dry fretting conditions. Results of AES analysis showed a layer an order of magnitude thicker than in lubricated fretting, also having a remarkably different chemical composition. TEM analysis confirmed that the reaction layer consisted of a silica-rich amorphous phase containing small inclusions of Fe₂O₃ and Fe₃O₄. In contrast to lubricated conditions, where the layer created was ductile, in the case of dry fretting the layer was brittle. The continuous process of forming and spalling the brittle tribolayer caused much higher wear rates and wear losses than under lubricated fretting conditions. No transition in wear behaviour was observed as was the case in lubricated fretting.     

Influence of coating thickness on rolling contact fatigue of alumina ceramics thermally sprayed on steel roller

The influence of the sprayed layer thickness on rolling contact fatigue of a thermally sprayed alumina ceramics with a nominal composition of Al₂O₃–2.3 mass% TiO₂ was investigated using a two-roller test machine under pure rolling contact condition with oil lubricant. The influence of undercoating of sprayed nickel-based alloy on rolling contact fatigue was also investigated. Thicknesses of the ceramics-sprayed layer of tested rollers were 0.2, 0.5 and 1.0 mm. The failure mode of sprayed rollers was spalling caused by subsurface cracking. In the case of sprayed rollers without undercoating, rolling contact fatigue strength of rollers with 0.2 mm thickness sprayed layer was the smallest. Rolling contact fatigue strength of sprayed roller with 0.2 mm thickness sprayed layer was improved by undercoating. In case the failure depth was small as compared with the thickness of sprayed layer, effect of undercoating on the rolling contact fatigue strength was little. The depths where the maximum values of subsurface shear stresses occurred, almost corresponded to the observed depth of spalling cracks.     

Acoustic emission of rolling bearings lubricated with contaminated grease

In elastohydrodynamic lubrication (EHL), an extreme smoothness of the contacting surfaces is essential for the formation of a thin lubricant film, which separates the moving surfaces from each other. Sharp asperities, in particular, are very detrimental for the sliding surfaces. Because of the limited thickness of the lubricating film, the contact zone is very vulnerable if the lubricant contains contaminants, especially if hard particles are present in the lubricant.The present paper focuses on the acoustic emission (AE) measurement method for the monitoring of the lubrication situation in a grease lubricated rolling bearing. The aim of the investigation was to clarify how the contaminants in the grease influence the acoustic emission of the rolling bearing. In the paper, the results of tests with clean greases and with deliberately contaminated greases, and the influence of the cleaning and re-greasing of the bearing are discussed. The results showed that the AE measurement indicated very clearly the lowest contaminant concentration included in the study that was as low as 0.02 weight-%. Small size contaminant particles generated a higher AE pulse count level than large size particles. The AE time signal analysis method proved to be a suitable method to indicate the hardness of the contaminant particles. Cleaning the bearing of contaminants and re-greasing with a clean grease reduced the AE level of the bearing.     

A microwedge model of sliding contact in boundary/mixed lubrication

A refined friction model of sliding contact in boundary/mixed lubrication regime is developed. In addition to the well-known asperity flattening and roughening effects, significant deformation of asperities can be incited by the elastic microwedges on the tool surface. A model of asperity deformation, which includes the effects of smoothing, roughening, and microwedge, is proposed for processes where smooth tool and rough workpiece are used. A finite element formulation incorporating the microwedge effect in the Reynolds equation for lubricant flow is also derived. Numerical results showed that the inclusion of microwedge in the analysis provides a good agreement with the experimental measurements, especially with many interesting phenomena that can be neither explained nor predicted by the other friction models.     

Microstructure and abrasive wear in silicon nitride ceramics

It is well known that abrasive wear resistance is not strictly a materials property, but also depends upon the specific conditions of the wear environment. Nonetheless, characteristics of the ceramic microstructure do influence its hardness and fracture toughness and must, therefore, play an active role in determining how a ceramic will respond to the specific stress states imposed upon it by the wear environment. In this study, the ways in which composition and microstructure influence the abrasive wear behavior of six commercially-produced silicon nitride based ceramics are examined. Results indicate that microstructural parameters, such as matrix grain size and orientation, porosity, and grain boundary microstructure, and thermal expansion mismatch stresses created as the result of second phase formation, influence the wear rate through their effect on wear sheet formation and subsurface fracture. It is also noted that the potential impact of these variables on the wear rate may not be reflected in conventional fracture toughness measurements.     

Residual stresses and retained austenite evolution in SAE 52100 steel under non-ideal rolling contact loading

Residual stresses are introduced and modified during manufacturing as well as by normal use under rolling contact loading. Operations such as heat treatments, shot peening, grinding, etc., are known to alter the magnitude and distribution of residual stresses. Our work revolves around the changes in magnitude and distribution of residual stresses, as they relate to deformation and the strain induced transformation of retained austenite. The residual stresses and retained austenite measurements were carried out using X-ray diffraction techniques. The rolling contact fatigue lives of different variants of SAE 52100 bearing steel were evaluated in a 5-ball-rod rolling contact fatigue machine under testing conditions leading to surface nucleated failure, i.e. non-ideal rolling contact. The tests were accelerated by applying well controlled micro-indentations on the wear track. The contribution of the residual stresses and amount of retained austenite to the rolling contact fatigue life were analyzed.     

Numerical analysis for predicting the rolling contact fatigue crack initiation in a railway wheel steel

This paper concerns the modelling of the rolling contact fatigue of a railway wheel steel, which is simulated with moving Hertzian contact pressure. Parametric studies are carried out with a two-dimensional elastic-plastic finite element model of a part of a wheel containing defects. Several parameters, namely the size and shape of material defects, the load magnitude and the friction coefficient are varied to investigate their effect on the railway wheel fatigue damage. Defects or small friction coefficient are a plausible explanation to the initiation of deep subsurface fatigue cracks.     

Probabilistic model for random uncertainties in steady state rolling contact

Wear phenomena involve a large number of physical and mechanical parameters which are not always well known or controlled during relative movement between two bodies. Numerous industrial applications necessitate an evaluation of technological component life time and wear modelling often fails to give accurate estimation.We use the classical Archard's wear model where wear is related to dissipated power. It appears that great dispersion can occur in the estimation of dissipated power related to a lack of knowledge of certain parameters.We present here a probabilistic approach of the contact problem resolution. We consider the specific contact problem in the case of steady state rolling. A wear apparatus has been used to test different materials and we use the simplified model Fastsim to evaluate slip and tangential traction in the contact zone. For each parameter of the simulation, we construct a probabilistic density function with the only information available.A Monte-Carlo method is implemented and the resolution of numerous cases allows the dissipated energy to be evaluated as a mean value and a confidence region for 95% viability.     

Effect of silicon nitride coating thickness on fatigue of silicon microbeams

In this paper, two silicon nitride layers with thickness, 0.2 and 0.4 μm, are coated onto single crystal silicon (SCS) in order to achieve Si₃N₄/Si cantilever microbeams. The effect of LPCVD silicon nitride surface coatings on fatigue properties of SCS cantilever microbeams is investigated. Fatigue testing is conducted at both 40 Hz and 100 Hz. Typical S–N (strain amplitude–fatigue cycle) curves of the beams are achieved and correlated fatigue failure modes are investigated. It is found that thinner Si₃N₄ coating of 0.2 μm results in better fatigue lives of Si₃N₄/Si beams than thicker Si₃N₄ coating of 0.4 μm. Both thinner and thicker coated beams have major fatigue crack planes along {1 1 1} planes; however, thicker coated beams possess specific failure mode of delamination, which is not found in thinner coated beams. Delamination reduces the reinforcing effect of thicker Si₃N₄ coating and leads to its shorter fatigue life. For thicker coated beams, fatigue life at 100 Hz is longer than that at 40 Hz. The mechanism for delamination and the effect of cyclic frequency is investigated, and factors for better fatigue life are proposed.     

A comparison of multiaxial fatigue criteria as applied to rolling contact fatigue

Under rolling contact fatigue (RCF) existing multiaxial fatigue criteria are not well validated and predict significantly different results. Results for simple typical Hertzian RCF pure rolling are shown as previously remarked by the authors, the Dang Van criterion applied to RCF gives over-optimistic fatigue limits, due to the large influence of the hydrostatic component of the stress, particularly under some conditions. It is here shown that the “simpler” Crossland criterion gives a more realistic fatigue limit of Hertzian peak pressure, and the more “elaborate” Papadopoulos criterion gives an even more conservative value, of about 3-3.5 times higher than the fatigue limit under pure shear. It is suggested that the multiaxial criteria per se do not give a reliable estimate of the fatigue limit, and perhaps an integration within Weibull-like theories should be attempted in the future, as well as a more “unified” approach and mix of criteria taken from gears design, rolling contact in railways, and in rolling bearings.     

Tribological behavior of selected engineering polymers under rolling contact

The rolling wear behavior of unfilled polyamide 6, polyoxymethylene and polyetheretherketone materials were investigated, using different wear testing rigs. In most of them, a steel ball rolls on the polymer plate surface in a rotational or linear motion. The results qualify the different polymers with regard to their resistance against rolling wear. The potential of various fillers and reinforcements for a possible improvement was also identified. But almost all showed an insufficient suitability when compared to the unfilled materials. Differences in wear mechanisms were illustrated by microscopic methods. But more works need to be done to understand all the interesting observations.     

Improvement of tribological performance of steel by solid lubricant shot-peening in dry rolling/sliding contact wear tests

The effect of shot-peening treatment with the particulate MoS₂ solid lubricant on the wear resistance of steel in the dry rolling/sliding contact wear tests was investigated. The duplex shot-peening treatment with ceramic balls and the particulate MoS₂ solid lubricant provided excellent wear resistance under a severe loading and sliding condition because the uniform and minute surface roughness given by shot-peening treatment with ceramic balls could keep shot-peened MoS₂ particles with a low friction coefficient on the sample surface. Furthermore, the sample surface was covered with shot-peened MoS₂ particles by a MoS₂ layer formed during the rolling/sliding contact wear test.