Discussion: "Elastohydrodynamic Lubrication in an Instrument Ball-Bearing" (Kannel, J. W., and Snediker, D. K., 1976, ASME J. Lubr. Technol., 98, pp. 244–248)

E. P. Kingsbury
1976 Journal of Lubrication Technology  
The electrical conductivity across a bearing has been measured by many people over the years, and a series of traditional objections is usually invoked by discussers. Taking these as made and answered, it can be noted that lift-off speed has been found to be a function at least of surface finish (both as measured and as produced by different manufacturing techniques), of amount and type of lubricant available at the contacts, and of amount of running time on the bearing. It is perhaps
more » ... s perhaps oversimplified to attribute all effects to viscosity. A "commercial traction fluid" was found by the authors to have the same lift-off characteristics as a mineral oil of the same viscosity. Experiments at Northrop Corp., ("Experimental Parameters Controlling Hydrodynamically Starved Lubrication in Instrument Ball Bearings," by E. Zeigler, presented at the International Bali-Bearing Symposium, Cambridge, Mass., 1973) show that with sufficient discrimination large differences between these materials can be found. Driving torque also has been found to depend strongly on other lubricant properties than viscosity. In the above reference and in "Experimental Observations on Instrument Ball Bearings," by the discusser and presented at the Dartmouth Bearing Conference, 1968, quantity of available lubricant was found to be the controlling parameter. Many of the important conclusions in this paper have to do with the "percent film" formed in a small bearing. One can think of several different kinds of percentages which could be of interest 1 here: percent of area-of-contact formed by metallic (or other nonfluid) junctions, percent of transmitted load, percent of the time in steady state when there is metallic contact, etc. It is not clear exactly which percentage is being measured by the author's instrumentation. In particular, is it the same film percentage that is measured by the X-ray method (assuming that question has finally been settled)? V. Wedeven 3 The authors should have compared their results with the volume of work already completed on instrument bearing lubrication. A number of excellent studies [15-21] 4 on the role of the lubricant in instrument bearings have shown that the thickness and stability of the EHD oil films control the condition of lubrication. Oil film thickness measurements by capacitance [17] or dimensional changes [15, 17] were found to be most useful. The studies conclude that the best performance is achieved when the bearings run starved beyond a critical speed, and, that failure is associated with the reduction of EHD film thickness caused by depletion of the oil through evaporation, centrifugal forces, and insufficient movement of the oil within the bearing. Thus oil quantity as well as viscosity was found to be an important lubricant parameter. The authors use an electrical continuity technique to determine the EHD condition of an instrument bearing. While many bearing applications operate in a partial EHD film regime where percent film measurements are very informative, these measurements alone would not seem to be sufficient to investigate the EHD con--The Charles Stark Draper Laboratory, Inc., Cambridge, Mass. :i NASA Lewis Research Center, Cleveland, Ohio. 4 Numbers 15-21 in brackets designate Additional References at end of discussion. dition in an instrument bearing where life limitation is a concern. Long life is achieved with thicker EHD oil films than the percent film technique can "see." A means of measuring these thicker films is important, especially since prediction by EHD theory is difficult if starved conditions prevail. It does not appear that starvation occurred during the percent film measurements where the EHD films are very thin. It can be shown [20, 21] that the onset of starvation is a function of the speed-viscosity parameter (u ixo)-The value of this parameter is relatively small for the percent film measurements, but vary much larger for the torque measurements. The largest (u no) value is for the Castor oil. The torque for this fluid, which is not plotted in Fig. 3 , is lower than the trend established by the other fluids. The authors' X-ray film thickness data in Fig. 8 correlates, in general, very well with the film thickness determined by the percent film measurements, despite the necessary assumptions required to make the comparison. A notable and familiar feature shown in Fig. 8 is that the X-ray data indicate a rather large reduction of film thickness with pressure which is not predicted by theory. The percent film measurements, on the other hand, appear to follow the expected trend of theory more closely. Could the authors comment on the reliability of both methods? Additional References The quantity of lubricant was at times more than sufficient to maintain full films. (The lubricant quantity was chosen to approximate that which is standard for this type of bearing in spacecraft applications.) All percent film tests were carried out on new, freshly-lubricated bearings and all data was taken within 15-30 minutes of total running time. This was done to minimize interference by such ball-track depletion mechanisms as creep and evaporation. Furthermore, since the electrical technique cannot distinguish between EHD films and electrically insulating boundary films, the relatively short running times minimized the formation of permanent boundary films involving the TCP in the oil and air and H2O. With regard to interpretation of the percent film data, we should reiterate that our percent film measurement was simply a reading of voltage drop across the film from a fixed source. 0 drop indicated a complete short and a reading of 100 indicated full film. Further interference from these data is, of course, speculative. However, it is in our opinion, very interesting that some significant conclusions can be drawn from these types of measurements. With regard to other Journal of Lubrication Technology
doi:10.1115/1.3452811 fatcat:pqpu5jedwjahljlykzm4f6jzhi