Transient Rotordynamic Modeling of Rolling Element Bearing Systems

Author(s):  
A. Liew ◽  
N. S. Feng ◽  
E. J. Hahn

Non-linearity effects in rolling element bearings may arise from the Hertzian contact force deformation relationship, the presence of clearance between the rolling elements and the bearing races, and the bearing to housing clearance. Assuming zero bearing to housing clearance and ignoring rolling element centrifugal load effects, it has been shown in earlier work that Rotor Bearing Systems (RBSs) with deep groove ball bearings can give rise to non-linear behavior such as chaotic motion and jump. This paper extends the bearing model to include rolling element centrifugal load, angular contacts and axial dynamics. The effect of more sophisticated bearing models is illustrated in both a rigidly supported rigid RBS and a flexibly supported flexible RBS, the latter being a model of a test rig designed to simulate an aircraft mounted accessory drive unit. Results are presented on the effect of bearing preload on the unbalance response up to a speed of 18,000 rpm.

2002 ◽  
Vol 124 (4) ◽  
pp. 984-991 ◽  
Author(s):  
A. Liew ◽  
N. Feng ◽  
E. J. Hahn

Nonlinearity effects in rolling element bearings arise from Hertzian contact force deformation relationships, clearance between rolling elements and races, and the bearing-to-housing clearance. Assuming zero bearing-to-housing clearance, a simplified earlier analysis showed that rotor bearing systems (RBSs) with deep groove ball bearings can give rise to chaotic motion and jump. This paper extends the bearing model to include rolling element centrifugal load, angular contacts and axial dynamics; and illustrates their effects in a rigidly supported rigid RBS and a flexibly supported flexible RBS, the latter modeling an existing test rig. Results are presented on the effect of bearing preload on the unbalance response up to a speed of 18,000 rpm.


Author(s):  
N. S. Feng ◽  
E. J. Hahn

Non-linearity effects in rolling element bearings arise from two sources, viz. the Hertzian force deformation relationship and the presence of clearance between the rolling elements and the bearing races. Assuming that centrifugal effects may be neglected and that the presence of axial preload is appropriately reflected in a corresponding change in the radial clearance, this paper analyses a simple test rig to illustrate that non-linear phenomena such as synchronous multistable and nonsynchronous motions are possible in simple rigid and flexible rotor systems subjected to unbalance excitation. The equations of motion of the rotor bearing system were solved by transient analysis using fourth order Runge Kutta. Of particular interest is the effect of clearance, governed in practice by bearing specification and the amount of preload, on the vibration behaviour of rotors supported by ball bearings and on the bearing load. It is shown that in the presence of positive clearance, there exists an unbalance excitation range during which the bearing is momentarily not transmitting force owing to contact loss, resulting in rolling element raceway impact with potentially relatively high bearing forces; and indicating that for long bearing life, operation with positive clearance should be avoided in the presence of such unbalance loading. Once the unbalance excitation is high enough to avoid such contact loss, it is the bearings with zero or negative clearance which produce maximum bearing forces.


2002 ◽  
Vol 124 (3) ◽  
pp. 468-473 ◽  
Author(s):  
Har Prashad

The diagnosis and cause analysis of rolling-element bearing failure have been well studied and established in literature. Failure of bearings due to unforeseen causes were reported as: puncturing of bearings insulation; grease deterioration; grease pipe contacting the motor base frame; unshielded instrumentation cable; the bearing operating under the influence of magnetic flux, etc. These causes lead to the passage of electric current through the bearings of motors and alternators and deteriorate them in due course. But, bearing failure due to localized electrical current between track surfaces of races and rolling-elements has not been hitherto diagnosed and analyzed. This paper reports the cause of generation of localized current in presence of shaft voltage. Also, it brings out the developed theoretical model to determine the value of localized current density depending on dimensional parameters, shaft voltage, contact resistance, frequency of rotation of shaft and rolling-elements of a bearing. Furthermore, failure caused by flow of localized current has been experimentally investigated.


1982 ◽  
Vol 104 (3) ◽  
pp. 283-291 ◽  
Author(s):  
S. H. Loewenthal ◽  
D. W. Moyer ◽  
W. M. Needelman

Fatigue tests were conducted on groups of 65-millimeter bore diameter deep-groove ball bearings in a MIL-L-23699 lubricant under two levels of filtration. In one test series, the oil cleanliness was maintained at an exceptionally high level (better than a class of “00” per NAS 1638) with a 3 micron absolute barrier filter. These tests were intended to determine the “upper limit” in bearing life under the strictest possible lubricant cleanliness conditions. In the tests using a centrifugal oil filter, contaminants of the type found in aircraft engine filters were injected into the filters’ supply line at 125 milligrams per bearing-hour. “Ultra-clean” lubrication produced bearing fatigue lives that were approximately twice that obtained in previous tests with contaminated oil using 3 micron absolute filtration and approximately three times that obtained with 49 micron filtration. It was also observed that the centrifugal oil filter had approximately the same effectiveness as a 30 micron absolute filter in preventing bearing surface damage.


2013 ◽  
Vol 569-570 ◽  
pp. 497-504 ◽  
Author(s):  
An Bo Ming ◽  
Zhao Ye Qin ◽  
Wei Zhang ◽  
Fu Lei Chu

Spalling of the races or rolling elements is one of the most common faults in rolling element bearings. Exact estimation of the spall size is helpful to the life prediction for rolling element bearings. In this paper, the dual-impulsive phenomenon in the response of a spalled rolling element bearing is investigated experimentally, where the acoustic emission signals are utilized. A new method is proposed to estimate the spall size by extracting the envelope of harmonics of the ball passing frequency on the outer race from the squared envelope spectrum. Compared with the cepstrum analysis, the proposed procedure shows more powerful anti-noise ability in the fault size evaluation.


1995 ◽  
Vol 117 (1) ◽  
pp. 166-170 ◽  
Author(s):  
Rao S. Zhou ◽  
Fukuo Hashimoto

Plasticity indexes are usually used to describe the elastic-plastic deformation properties of a rough surface. The contact surfaces of any new rolling element bearing always change during the bearing run-in period until arriving upon a relatively stable surface texture. In this paper, a new random isotropic surface with very low plasticity index is proposed and created for the surfaces of rolling elements bearings. From both the calculated surface parameters and the run-in torque test results, the special rolling contact surfaces demonstrated “No Run-In” characteristics. Lower running torque is obtained from bearing tests and longer fatigue life is also expected by the subsurface stress analysis.


Author(s):  
P Brossier ◽  
D Niel ◽  
C Changenet ◽  
F Ville ◽  
J Belmonte

In the present study, some measurements have been conducted on a dedicated test rig to investigate rolling element bearing thermal behaviour. This test rig makes possible the determination of the tested rolling element bearing power losses through the resistive torque measurement. Some thermocouples are located on fixed parts of the system (housing, rolling element bearing outer ring) and others on rotating parts (rolling element bearing inner ring and shaft) via a telemetry system. A deep groove ball bearing, whose pitch diameter is equal to 85 mm, has been tested under oil jet lubrication for different operating conditions. Measurements of the penetration ratio, defined as the proportion of oil actually entering the rolling element bearing versus the oil injected, have also been conducted. An extended thermal network of the test rig has been established to enable a closer understanding of the rolling element bearing inner thermal behaviour. Based upon the first principle of thermodynamics for transient conditions, the studied system is divided into lumped elements at uniform temperature connected by thermal resistances which account for conduction, radiation and convection. Convection within the rolling element bearing depends on the amount of oil in the oil–air mixture known as the volume fraction. At specific test conditions, the developed model found good agreements with experiences for a given oil volume fraction of 15%. This value of volume fraction leads to an adapted formula for volume fraction in the case of jet lubrication which includes the measured penetration ratio.


2014 ◽  
Vol 592-594 ◽  
pp. 1099-1103 ◽  
Author(s):  
T. Govardhan ◽  
Achintya Choudhury ◽  
Deepak Paliwal

External load in a bearing is transferred from one race to another race through the rolling elements. In the present work, an investigation has been made to estimate the load on a rolling element in a bearing subjected to dynamic loading. The dynamic loading, in the present study, included harmonic and periodic loadings which are deterministic functions of time. The roller load is also investigated under random loading with known statistical values of mean and variance. Numerical values have been obtained for NJ204 bearing with known radial clearance. These results show the variation in the spectra obtained for different nature of external loadings. These results can be expected to satisfy the difference in theoretical and experimental spectra obtained by earlier researchers.


2015 ◽  
Vol 137 (2) ◽  
Author(s):  
L. Houpert

Analytical relationships for calculating three rolling element bearing loads (Fx, Fy, and Fz) and two tilting moments (My and Mz) as a function of three relative race translations (dx, dy, and dz) and two relative race tilting angles (dθy and dθz) have been given in a previous paper. The previous approach was suggested for any rolling element bearing type, although it has been recognized that the assumption of a constant rolling element-race contact angle is not well supported by deep groove ball bearings (DGBB) or angular contact ball bearings (ACBB). The new approach described in this paper addresses the latter weaknesses by accounting for the variation of the contact angle on the most loaded ball and also shows that misalignment effects on spherical roller bearing (SRB) loads are negligible. Comparisons between the simplified approach (option 1) and the “enhanced” numerical approach (option 2, which requires a summation of the load components on each ball with the appropriate contact angle included) is made, showing a good correlation as long as the relative misalignment remains reasonable or occurs in the plane corresponding to maximum radial displacement. Option 2 can, however, be recommended since it is easy to program and quite accurate at any misalignment level. Other pros and cons of both options are described. As in the previous paper, a full coupling between all displacements and forces, as well as roller and raceway crown radii, are considered, meaning that Hertzian point contact stiffness is used in roller bearings at low load with a smooth transition toward Hertzian line contact as the load increases. This approach is particularly recommended for programming the rolling element bearing behavior in any finite element analysis or multibody system dynamic tool, since only two nodes are considered: one for the inner race (IR) center, usually connected to a shaft, and another node for the outer race (OR) center, connected to the housing.


Author(s):  
Wentao Huang ◽  
Zhenzhen Dong ◽  
Fanchao Kong ◽  
Qiang Fu

In this paper, a vibration model considering the impact force is developed based on Hertz contact theory. The model considered both structural parameters and kinematic parameters. In this model, the contact areas between balls and races are simulated as linear springs, and the contact areas between the inner race and shaft and the contact areas between the outer race and housing are considered as spring-damper systems. Unlike previous studies, in which the stiffness was simply calculated in one direction, the stiffness parameters were calculated separately in the horizontal and vertical directions in the present study. In this model, the impact process between rolling elements and defective races, which is affected by velocity, was studied in detail. The solution of the vibration equation was obtained using the Runge-Kutta method. The numerical results for 6204 deep groove type ball bearings with outer race defects have been obtained and compared with the experimental results, which validated the effectiveness of this model in both the fault mechanism analysis of rolling element bearings and fault diagnosis.


Sign in / Sign up

Export Citation Format

Share Document