Intro

Optimizing parts and processes is very important because even small efficiency developments can save a lot of energy and materials in the long run. An important part of optimization is to identify the source and influencing factors of loss. In our current topic, namely rolling bearings. During the operation of oil-lubricated rolling bearings, losses will also occur and can be divided into two groups. The first group is load-related losses or contact losses, which are mechanical friction caused by contact forces. To determine these losses, appropriate computational methods and multi-body dynamics simulation techniques are available [1]. The second group of losses is caused by lubricant displacement and lubricant shear on the component surface and is independent of the load. The purpose of our project is to investigate this group, commonly referred to as hydraulic losses. Compared to contact losses, the number of calculation methods available for hydraulic losses is limited. Only the same empirical models or numerical methods that require high power and computation are available. Experimental and numerical studies were therefore carried out within a collaborative project to gain a more detailed understanding of load-independent hydraulic losses. The experiment was carried out at Machine Elements, Gears & Transmissions (MEGT) and the simulation was carried out at the Institute of Tribology and Energy Conversion Machinery (ITR). The focus of the project is on full immersion and semi-immersion lubrication of tapered roller bearings, where the impact of the environment near the bearing is also considered as an influencing factor.

Test method

To investigate hydraulic losses, MEGT[2, 3] developed a test bench with a modular design that can be used to measure the total frictional torque of axially loaded rolling bearings with different designs, sizes, lubrication and axial bearing positions (Figure 1). The test unit consists of an angular contact ball bearing 7208 as a support bearing and a tapered roller bearing 32208 as a test bearing. The rolling bearings are installed in an X shape and are pre-tightened by a load unit consisting of a load cell, a dish spring and a load bolt. Since there is a separate oil chamber, the oil level can be adjusted independently at the bearing, and the flow formed in the oil chamber will not interfere with each other. In addition, the environmental impact of hydraulic losses in the second oil chamber when the bearings are tested at different axial positions can also be studied. The proposed test unit is driven by a DC motor with a belt with a maximum speed of 10,000 revolutions per minute. There is a shaft torque sensor between the drive and the test unit that measures the loss of torque generated in the test unit up to a maximum of 5NM.

Figure 1: Test bench for experimental study of hydraulic losses in horizontal bearing configurations [2]

During the measurement process, the outer ring temperature, oil temperature, rotational speed, axial load and shaft torque are observed and measured. With the measured torque, the hydraulic loss can be determined based on the change in oil level (Figure 2). This means that we need two measurements with the same boundary conditions, except for the amount of oil. One is the amount of oil investigated and the other is the minimum amount of lubrication. The difference between the measured lost torque is the pure hydraulic loss under the adjusted boundary conditions.

Figure 2: Example of determining hydraulic losses based on oil level changes under a fully submerged tapered roller bearing 32208

Numerical method

In addition to the experiments, CFD fluid simulations were carried out at the Institute of Tribology and Energy Conversion Machinery (ITR) to determine hydraulic losses. These 3D transient simulations are performed using ANSYS CFX and model the complex flow distribution in the test chamber. In the calculation, the oil properties and speed are set according to the measured value. The computational grid covers the oil chamber of the test bearing and contains approximately 5.5 million hexahedral units. The grid of this blocky structure is divided into three sections and has a finer resolution near the walls.

Result

Figure 3: Results of the reproducibility test and the first numerical calculation

Since the difference between each measurement point in the reproducibility test is less than 5%, we can say that the measurement has good reproducibility. The comparison shows that there is a good correlation between the experimental and numerical results, with only a small deviation in the case of higher viscosity. In general, we can say that lower oil viscosity (higher oil temperature) leads to lower hydraulic losses, which increase almost linearly with rotational speed. After the reproducibility test, further measurements were made at different oil levels and at different test bearing axial positions. The results show the influence of oil volume, oil temperature (i.e., oil viscosity), speed and bearing position (i.e., whether there is a wall near the test bearing) on the hydraulic loss value. In addition, the so-called pump effect of tapered roller bearings and the foam of oil can also be observed in the measurement results presented at STLE's 74th Annual Meeting & Exhibition

【 Reference 】

[1]Aul, V.; Kiekbusch, T.; Marquart, M.; Sauer, B.: Experimentelle und simulative Ermittlung von Reibmomenten in Wälzlagern. 51. Tribologie-Fachtagung GfT 09/2010

[2]Gonda, A.; Großberndt, D.; Sauer B.; Schwarze H.: Experimentelle und numerische Untersuchungen der hydraulischen Verluste in Wälzlagern unter praxisrelevanten Bedingungen. 59. Tribologie-Fachtagung (GfT) 2018, 24.-26.09.2018, Göttingen; pp. 35/1-35/10, Band 1

[3]J. Liebrecht, X. Si, B. Sauer und H. Schwarze, "Untersuchungen von hydraulischen Verlusten an Kegelrollenlagern,“ Tribologie und Schmierungstechnik, pp. 14-21, 5 2014.