Servicenavigation

Project Details

Back

Projekt Print View

Model for determination of thermal loads of the lubricant due to the mechanical and electrical loading in roller bearings

Subject Area Engineering Design, Machine Elements, Product Development
Term from 2018 to 2022
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 407468812
 

Final Report Abstract

In this project the local effects of electrical current passage in mechanical roller bearings at low and medium speed, hence at low and medium lubrication, were investigated both simulative and by experiment. Two partner labs, one expert on mechanical bearing aspects (MEGT, TU Kaiserslautern) and one expert on electrical contact effects (EW, TU Darmstadt) cooperated. A mechanical contact model between bearing balls and raceway, based on the coupled Euler-Lagrange method (CEL) was developed at MEGT for a detailed analysis of the contact between rough surfaces of the contact partners without any lubrication. The model was enhanced by nonlinear material properties of incipient plasticization, thermal softening, and strain rate dependence. Using these models at EW, the electrical contact simulation resulted in much too low electrical contact resistance values, which had been determined experimentally at EW in previous researches. The calculated surface of the mechanical contact from MEGT, simulated electrically by EW with Finite Element Analysis in JMAG, concluded that the major parts of the mechanical contact are not conductive, being covered with an insulating oxide layer. The current passage contact cross section area is even at dry bearings without lubrication by a factor 1000 smaller than the simulated mechanical contact area. Hence the detailed CEL model for rough surfaces is not necessary for the investigation of the current transfer in the rolling contact. The electrical current obviously flows only on a very small sub-area of the calculated load-depending bearing roughness peaks. So the results of contact physics of mechanical switches and breakers were adapted to the bearing contact modelling. According to Holm´s contact theory the high mechanical pressures at the surface asperities and/or the electric breakdowns of the oxide layer (“A-fritting”) lead to metal-metal contacts (“a-spots”). From that theory it follows that the radii of the a-spots enlarge as they are exposed to higher electric current and their number increases also with contact force. The metal surface at the a-spots oxidizes again, as soon as the current is shut down, which makes it very difficult to count and measure them after dismantling the bearing surface for microscope investigation. Three test rigs with bearings under mechanical load and electrical DC and AC current flow were set up at MEGT and EW to obtain detailed knowledge on the electrical current passage in the bearing contact at dry bearing and at mixed lubrication at different load force, temperature and frequency. By that speed- and load-depending bearing impedance values were measured as an input to further simulative investigation. Experimental parameter studies on the bearing force, bearing speed, bearing voltage and bearing current were performed to characterize the electrical contact in the bearing. A refined model for one selected electric circular contact a-spot was developed at EW in COMSOL Multiphysics, which by varying the parameters like a-spot radius allowed determination of local losses, impedance and temperature rise with the use of the measured voltage and current signals from the experiments. The local temperature rise due to the current passage in the bearing electric contact point between the balls and raceway may be very high, reaching the softening or even melting point of the material. The FE-model of the a-spot indicates, that the temperature in an “a-spot” of radius 1 μm has a rise-time lower than 1 μs, so the rise is fast enough even at rotating bearings to reach high values. The a-spot steady-state temperature depends only on the contact voltage and is independent from aspots shape and size, which confirms the findings of Holm´s theory. For a contact voltage of 0.4 V, the steady-state temperature of the a-spot is 1055°C, at an ambient temperature 20°C of the bearing body. As the electric contact area is very small, the average temperature rise of the whole bearing body due the current passage is negligible, and is dominated by friction heating. The high local temperature can damage the bearing surface at the electric contact points and in case of big current flow may also crack up the C-H-molecule chains of the surrounding oil. These impacts are far above the possibilities of simulation, as the number and size of a-spots is statistically distributed in the contact area and very difficult to be optically measured, as noted above. Therefore, at MEGT, first correlations to the incipient damage as a result of the combined mechanical and electrical load were obtained via 170 hours-tests at the axial bearing (type 51208) load test rig. All balls are under the same mechanical loading conditions and are all electrically operating in parallel under switched DC voltage at 20 kHz. The voltage was adjusted to get a discharge current, which - taking the whole mechanical contact area as “apparent” current passage area and not the 1000-times smaller total a-spots area - resulted in an “apparent” peak current density of ca. 0.1 A/mm2. From prior field tests at EW it is known, that this value usually does not destroy the lubrication and does not harm the bearing surface. Varying the bearing force from 4 … 8 kN, and comparing without and with current flow, even at this low “apparent” bearing current density discharge pits or even fluting was observed with current flow. Without current flow only at high load force some pits were observed at the balls, obviously due to increased mechanical wear (Fe-particles). The lubricant was not deteriorated.

Publications

 
 

Additional Information

© 2024 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung