ABSTRACT

Quantitative X-ray diffraction methods, used for the analysis of the state of residual stress and microstrain in the raceways of tested gearbox bearings, allow the characterization of the microstructure and the changes thereof in response to the rolling contact loading conditions during the test. The state diagnosed is a measure of the severity of the running conditions and can be used successfully to guide optimisation of both the internal design of the bearing and the rigidity of the gearbox.

SYNOPSIS

The endurance life of ball and roller bearings tested under high loads and temperatures is ultimately determined by the occurrence of spalling fatigue failure. But even so the concept of clean steel in the manufacturing of bearings has lead to superior endurance performance as long as the bearings are tested under ideal running conditions that provide clean and adequate lubrication. Consequently, design and life performance model have been adjusted with the practical result of downsizing for many of the bearing types. Nevertheless, it is a difficult task to predict the life of bearings under all possible service conditions to guarantee their performance since bearing failure may occur by other modes then by spalling fatigue failure.

Spalling fatigue is characterised by dominant subsurface crack growth and will occur only then when the hardened steel is weakened by fatigue. The weakening follows in response to micro-plasticity effects first in the highest loaded region at some distance below rolling contact surfaces. The decisive parameters for weakening to develop are bearing load, operating temperature and microstructural stability. With progression of the weakening increases the probability for spalling fatigue to occur. This follows from both the presence of previously uncritical type of non-metallic inclusions in the "volume at risk" and from the gradually increase of the volume "at risk". The latter increases potentially the number of failure triggering sites involved.

However, other failure modes come into place in particular when the lubrication is contaminated. The so-called "surface distress" failure mode is then dominant and determinated for the length of the "service life" of bearings. This "service life" can be a fraction of the endurance life obtainable under clean lubrication conditions. "Surface distress" appears in the rolling tracks from running the bearing under contaminated lubrication conditions. The solid particle contamination of lubrication in service, is one of the most serious threats to high quality performance of rolling bearings. These particles originate from the environment in which the bearing has to operate. Especially in gearbox applications, bearings have to operate under conditions that allow debris to be brought into the bearing and into the lubrication inlet zone in front of the rolling contacts between bearing elements. Once entrapped and squeezed they cause damages in the rolling tracks by indenting the surfaces in contact.

The accumulation of these surface damaging events is a notorious problem and can be recognised from the increasing noise produced by the bearing. Soon after, the rolling contacting surfaces will suffer from wear, which in turn contributes to a further increase of noise and the generation of debris of bearing origin. Consequently internal bearing clearance will change, which, especially in gearbox applications is required to be maintained as close as possible to the design level, seen the adverse effects of a change in clearance on the shaft alignment, effecting the fine tuned gear contacts in the transmission of a torque. Disturbance of the alignment will contribute to wear on gear surfaces which will develop with a further contribution to the noise and debris produced in the gearbox.

Striving for smooth and soundless performance, over the entire life span of a gearbox, requires the control of all contributing parameters. However, with existing life predicting models, it is troublesome to calculate the allowable load for required " service life". Complexity arises from aspects of fluctuating dynamic load, misalignment as a result of less rigid design of the gearbox housing, temperature gradients in the gearbox, contamination levels in the lubricant and susceptibility of the hardened steel to absorb plastic deformation upon indenting events. Only by testing the complete gearbox under all possible combinations of torque, temperature and lubrication conditions assures sufficient information of performance in service of all the bearings in the gearbox application. Occurrence of bearing failure in trial tests is then corrected by changes in design following an approach or trial and error.

Alternatively a most efficient approach is to combine testing with bearing performance analysis. The latter is executed on bearing component terminated from the test. The methods applied are based on describing the changes occurring in the state of the microstructure at and below the surface of bearing components. Quantitative X-ray diffraction methods are used including analysis of the state of residual stress, of microstrain and of retained austenite present in the hardened steel. The analysis provides a measure of the "microstructural response" in the rolling tracks of bearing components to the particular loading conditions the bearing is exposed to the application., Bearing components that suffer from too high load and or temperature will show subsurface response. Bearings that suffer from surface distress will show alterations in the near surface region. It has been shown that these three microstructural parameters are sensitive indicators to the loading conditions in the rolling tracks and as such can be used to characterise the severity of the operating conditions in the applications. Testing time under realistic conditions can be a fraction of required "service life" and can be applied in principal to any component of any bearing in the gearbox. The observed response in combination with knowledge, of the mechanical behaviour of hardened bearing steel under cyclic rolling contact, allows then to judge the local loading condition the bearing component investigated has been exposed to. The occurrence of surface fatigue is connected with subsurface weakening of the hardened steel and the occurrence of surface microstructural change to the capability of the steel's microstructure absorbing and accommodating micro-plasticity effects. Extrapolation of the results allows for the prediction if the "microstructure" of the bearing component will be successful in reaching "service life" requirements.

Discrimination is possible between conditions that will lead to:

1. surface distress as developed from particle contamination,

2. surface distress as a results of starved lubrication,

3. spalling fatigue as a results of heavy loading or

4. smooth operation without any change in the microstructure indicative of well selected bearings and running conditions.

The approach of combining "testing under service conditions" with "material response analysis" carries the advantage of making much shorter/faster R&D efforts in developing new gearboxes. The approach stimulates close operation between bearing and gearbox manufacturers. Three examples will be discussed highlighting the impact of bearing performance analysis on gearbox performance.