23rd International Colloquium Tribology - January 2022 477 Improved design process of dry-running radial plastic plain bearings by coupling laboratory tests and component simulation Marc Fickert Leibniz-Institut für Verbundwerkstoffe GmbH, Erwin-Schrödinger-Straße 58, 67663 Kaiserslautern, Germany Corresponding author: marc.fickert@ivw.uni-kl.de Andreas Gebhard Leibniz-Institut für Verbundwerkstoffe GmbH, Erwin-Schrödinger-Straße 58, 67663 Kaiserslautern, Germany 1. State of the art and daily practice While comprehensive standards, design guidelines [1] and professional calculation tools [2-4] are available for hydrostatically and dynamically operated radial plain bearings, the design process of dry-running radial plain bearings made of plastic for demanding applications has so far not been possible without cost-intensive component tests. The reason for this is the currently available calculation methods which are of a highly approximate nature and cannot be readily applied to other materials due to the necessary transfer functions in the form of diagrams. There are significant methodological gaps, particularly with regard to the consideration of the sliding surface temperatures occurring in the respective installation situation, since the natural interdependence between the temperature-dependent coefficient of friction and the temperature dependent on the coefficient of friction, in combination with the heat conduction dependent on the installation situation, cannot be solved without further ado. But even with the use of individual and high-resolution calculation methods, such as the FEM method, a lot of material characteristics are required, the determination of which in the form of tribological material maps by means of model tests such as the block-on-ring wear test involve an enormous amount of work. 2. New concept “aBoR” Because of the above reasons, a novel method for the design of dry-running radial plain bearings is currently being developed at Leibniz-Institut für Verbundwerkstoffe GmbH. By building a computer-aided calculation model of a plain bearing and coupling it with a block-on-ring wear test rig, a control loop (“hardware-in-the-loop”) is created that simulates the real behavior of a plain bearing. The plastic block used in this process corresponds to a segment of a plain bearing at its most heavily loaded point. On the basis of the test rig measurement variables of sliding friction coefficient and block height, which are continuously transmitted to a simulation during an ongoing test, the simulation calculates the current operating state of a virtual plain bearing on the basis of all the measurement data transmitted to date. A thermal calculation model is used to calculate the temperature of the virtual shaft from the coefficient of sliding friction and the geometric and material-specific boundary conditions entered. The wear-related change in block height is used to calculate the wear-related change in geometry of the virtual bearing which in turn is used to calculate the current surface pressure distribution in the virtual bearing. These two calculation results more precisely, the representative normal force resulting from the surface pressure and the counter-body temperature are transmitted to the test rig and adjusted accordingly. This procedure is carried out iteratively and continuously. Thus, both the surface pressure prevailing in the further course of the aBoR test and the temperature of the ring test specimen no longer result, as before, only from the frictional power of the block-ring test specimen pair and the heat dissipation dependent on the respective test rig design, but are set according to the results of the simulation of a plain bearing running during the test. A patent application has been filed for the “aBoR” (advanced Block-on-Ring) method (DE 10 2021 109 854), which is intended to significantly optimize the informative value and processing time of a plain bearing design. 3. First results of early-stage prototype The described hardware-in-the-loop control loop, i.e. the communication between the wear test rig and the separately running simulation, was prototypically set up using a blockon-ring test rig. A thermal network model was used to calculate the temperature development and distribution of a virtual plain bearing on the basis of the measured coefficient of sliding friction (see Figure 1). In the example shown the coefficient of friction starts at a high level of almost 0.3, but quickly drops to a value below 0.1 after running-in. With the time-delayed increase in the friction surface temperature (shaft temperature) it then rises again rapidly to a maximum value of 0.31 and then falls to a steady-state value of 0.16. The shaft temperature then rises again to a maximum value of 0.31 and then falls to a steady-state value of 0.16. The shaft temperature rises temporarily to up to 170 °C. Knowledge of the operating temperatures of the individual plain bearing components was used to optimize the clearance of the plain bearing application. Based on the component temperatures, their thermal expansions are continu- 478 23rd International Colloquium Tribology - January 2022 Improved design process of dry-running radial plastic plain bearings by coupling laboratory tests and component simulation ously calculated (see Figure 2). The difference between the bearing inner diameter and the shaft diameter is the operating clearance (see Figure 3) which is set to its smallest value of 2.2 ‰ in the range of the highest shaft temperature. Figure 1: Simulated temperature curves based on the measured coefficient of friction (COF) Figure 2: Shaft and bearing diameter Figure 3: Operational clearance If this value is now reduced to the smallest possible radial clearance of, for example, 0.5 ‰ [5], the plain bearing clearance can be reduced by 52 µm, thus extending the service life accordingly. Table 1: Results of clearance optimization 4. Summary and outlook The currently available design methods for dry-running radial plain bearings are inaccurate with regard to the prediction of sliding surface temperature, dimensioning and lifetime, and inflexible with regard to the influence of the installation situation on the temperature distribution. As a remedy, a hardware-in-the-loop process called “aBoR” was developed. A first prototype, which uses a thermal network model to calculate the transient component temperatures, is already capable of continuously calculating the operating clearance and thus optimizing the geometric design of a bearing. The next development goals are the correlation of the model with the results of component tests, the implementation of FEM-based simulation methods for heat conduction and distribution and the development of a wear progress model. References [1] VDI-Fachbereich Produktentwicklung und Mechatronik, „Gleitlagerberechnung - Hydrodynamische Gleitlager für stationäre Belastung,“ in VDI-Handbuch Produktentwicklung und Konstruktion, Berlin, Beuth Verlag GmbH, 1992, p. 59. [2] Tribo Technologies GmbH, „Software for higher efficiency - Radialgleitlager,“ o. J.. [Online]. Available: https: / / www.tribotechnologies.com/ de/ tribox/ basismodule/ radialgleitlager. [access on 30.06.2021]. [3] TU Clausthal - Institut für Tribologie und Energiewandlungsmaschinen, „COMBROS R - Simulation des Radialglleitlagerbetriebsverhaltens,“ o. J.. [Online]. Available: https: / / www.itr.tuclausthal.de/ forschung/ tribosimulation/ combros-r. [access on 29.06.2021]. [4] KISSsoft AG, „KISSsoft Spezifikationen - Wellen und Lager,“ o. J.. [Online]. Available: https: / / www. kisssoft.com/ de/ products/ productoverview/ kisssoftr-elements. [access on 30.06.2021]. [5] M. Mäurer, Tribologische Untersuchungen an Radialgleitlagern aus Kunststoffen, TU Chemnitz, 2002.