planetary bearings in wind turbine (WT) gearboxes were published in 2015 [1]. Since then journal bearings play a crucial role in new generation wind turbine drivetrains. The transition from rolling element bearings to journal bearings enhances gearbox torque density due to the compactness of journal bearings. Furthermore, journal bearings promise a higher turbine reliability due to their potentially unlimited fatigue lifetime [2]. Especially in the light of great expenses for operation and maintenance (around 33 % of the levelized costs of electricity for offshore wind energy in 2021 [3]) the increase in drive train reliability remains a priority for the wind industry. The reliability of journal bearings is strongly influenced by the risk for a spontaneous failure. This can be prevented through a targeted running-in by abrasive wear. Wear is Science and Research 14 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 1 Introduction and Motivation In recent years, there has been a trend in wind energy technology to replace rolling-element bearings as planetary bearings in wind turbine gearboxes with journal bearings. The first groundbreaking investigations by M EYER et. al. on the suitability of journal bearings as Derivation of running-in procedures for planetary journal bearings in wind turbine gearboxes by means of abrasive wear simulation Thomas Decker, Georg Jacobs, Julian Röder, Timm Jakobs, Jan Euler* submitted: 20.09.2024 accepted: 28.12.2024 (peer review) Presented at GfT Conference 2024 * Thomas Decker, M.Sc. Prof. Dr.-Ing. Georg Jacobs Julian Röder, M.Sc. Timm Jakobs, M.Sc. Jan Euler, M.Sc. Chair for Wind Power Drives (RWTH Aachen University), Campus-Boulevard 61, 52074 Aachen The implementation of journal bearings as planetary bearings has enabled higher torque densities of planetary gear stages in wind turbines. Since journal bearings are generally more compact than rolling element bearings they enable more planetary gears to be fitted into a planetary gear stage compared to one of similar size with rolling element bearings. Additionally, when designed and operated correctly, journal bearings operate reliable with potentially unlimited fatigue lifetime. These advantages have led to a technology shift in recent years from rolling element bearings towards journal bearings as planetary bearings in wind turbine gearboxes. At the beginning of their service life journal bearings are subjected to a running-in phase either on a test rig or directly in the field. During the running-in phase the contour and surface roughness of the bearing adapt mainly due to abrasive wear in dependence of the loads the bearing experiences. The bearings must operate within their associated system environment (e.g. gearbox) so that the bearing properties can be established ideally. For optimal running-in of the bearings special load procedures can be executed in a controlled environment. Running-in procedures often in- Abstract clude gradually increasing loads until nominal load is achieved. For journal bearings in wind turbines this approach is not always applicable, since a deliberate running-in is a time-consuming process and therefore not necessarily part of standard end of line tests of wind turbine gearboxes. Wear simulation tools enable a derivation of ideal running-in procedures for example in terms of a reduced procedure duration and energy input. Such a simulation enables the assessment of changes in the microgeometry and surface roughness due to abrasive wear under the influence of operation under mixed friction conditions. With a methodically defined running-in process, the desired contour is created quickly without exceeding the bearings fatigue limits. Thus, this work presents a method for the derivation of running-in procedures for journal bearings based on abrasive wear simulations. Keywords journal bearings, wear simulation, wind turbines, planetary bearings, running-in the loss or degradation of material from the sliding surfaces of the journal bearings due to tribological stresses [4]. Among the known wear mechanisms [5], abrasion is best understood in terms of a simulative representation. With abrasive wear simulations the wear behavior of journal bearings can be investigated prior to expensive prototype testing. Therefore, wear simulations have the potential to play an important role in the industrial design process of journal bearings and the derivation of running-in procedures. S CHERGE et. al. demonstrated with an experimental study on tribometers that the long-term wear behavior of tribological systems can be improved by intentionally applying high stresses during the running-in process at the beginning of the bearing’s service life. Without a specific running-in process, the wear rate during the first hours of operation is lower than in the test with the running-in process. However, the wear rate of the specimen that previously has been run-in under high stress has proven to be lower. Thus, in the long run the bearing that was run-in shows lower wear [6]. In [7] an experimental study was presented showing the effect of a gradually decreasing sliding speed on the running-in behavior of journal bearings. It was demonstrated that a cyclic repetition of a certain running-in procedure leads to a reduction of measurable friction moment over time. Although the running-in of journal bearings has been well researched experimentally, the simulative evaluation of running-in procedures with varying loads is rarely discussed in the literature. This work presents a method for the derivation of running-in procedures for journal bearings based on abrasive wear simulations. Different running-in procedures are simulated and compared in terms of the generated wear pattern, asperity contact and generated friction. The aim of this analysis is a method that will allow for the derivation of an optimum running-in procedure in the long term. To achieve this an existing simulation tool for the calculation of abrasive wear is extended by the calculation of time-varying loads and a thermal model. The method is showcased using the model of a component test rig. A qualitative validation is presented using experimental results. Lastly a transfer of the simulation method to a full-size WT gearbox model is presented. 2 State of the art The calculation of abrasive wear on hydrodynamic journal bearings operating under mixed friction conditions has been subject of research and development for several years and is addressed in numerous publications [8, 9]. An overview of the established method for the simulation of abrasive wear used in the investigation in this paper is shown in Figure 1. Typically wear simulations consist of an iterative loop between an elasto-hydrodynamic (EHD) simulation (1) in combination with a contact model (2) and a wear calculation model (3) [9, 10]. The oil film height h i,j and the asperity contact pressure p ai,j are calculated using a commercially available EHD/ MKS tool comparable to the work by König et. al. [9]. With every iteration step i of the wear simulation a time increment t acc,i of the wear process is calculated, which in turn depends on the calculated wear rate dh W ⁄dt in this increment [11]. The wear height h W,i,j is calculated in (3) and looped back into the EHD simulation, where it is considered as a change in the bearing’s contour c B,i,j . This results in a change in asperity contact pressure p ai,j and hydrodynamic pressure p h,i,j . Due to the high computational effort required for EHD simulations, common wear simulations are based on calculating only one or a few revolutions of the bearing under static load in the EHD simulation until a steady state in terms of pressure and oil film height h i,j is reached. This steady state is then assumed to be constant over the entire duration of Science and Research 15 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 Termination EHD simulation Asperity wear model 𝑡 𝑎𝑐𝑐 ≥ 𝑡 𝑚𝑎𝑥 Oil film height ℎ 𝑖,𝑗 Asperity contact pressure 𝑝 𝑎,𝑖,𝑗 Wear height ℎ 𝑊,𝑖,𝑗 Surface roughness 𝑅 𝑎,𝑘 Wear height ℎ 𝑊,𝑖,𝑗 Contour wear model Contact model Wear height Hydrodynamic pressure Start 1 2 3 4 Figure 1: Multi-scale simulation flowchart for abrasive wear of journal bearings based on [11] and [9] Science and Research 16 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 the iteration step i of the wear simulation. After a certain number of iterations n i , a steady state in terms of the bearing profile c B and pressure is achieved. The method presented in Figure 1 contains a termination criterion t max which leads to the finalization of the wear simulation once a predefined accumulated wear time (Eq. 1) is reached. Hereby, the wear occurring within a specified duration can be calculated. Eq. 1 K ÖNIG et. al. introduced the additional representation of asperity wear (4) at each increment representing the smoothing of the bearing’s surface. An experimental validation of this approach on a component test rig for journal bearings is presented in [9]. H AGEMANN et. al. demonstrated the importance of structural deformations (e.g. gear ovalization, bending of the planetary pin) in a WT gearbox on the wear behavior of planetary journal bearings. To capture those deformation effects it is crucial to represent structural deformation by means of a finite element model or multi-body simulation (MBS) models coupled with the EHD representation of the sliding contact [10]. In [11] it was demonstrated by Decker et. al. that the wear simulation method shown in Figure 1 can correctly represent the effect of running-in of journal bearings. The simulation yields decreasing asperity contact as the wear progresses over time and converges towards zero when a moderate load is applied. This selfhealing effect forms the basis of running-in simulation. A useful application for the aforementioned type of abrasive wear simulation was showcased by JAITNER et. al. They simulated the occurring wear of planetary journal bearings in a WT gearbox during a targeted running-in process using an EHD/ MBS. For the running-in procedure they chose a step wise increase in input torque and demonstrated, that the wear volume V W converges towards a steady state [8]. L INJAMAA et. al. suggest running-in of the bearing using a step wise decrease in sliding speed while keeping the specific pressure constant [7]. Both approaches target running-in but with a different concept to increase the bearings load. In reality, there are load limits for journal bearings that must not be exceeded in order to avoid damage (spontaneous failure) [12]. Therefore, the running-in process cannot be carried out under arbitrarily high loads, but must be conservative and gradual. There is therefore potential for optimization between the best possible running-in in terms of the wear generated, the process duration and the friction energy applied. In summary, simulative methods for calculating the abrasive wear on journal bearings are well established and experimentally tested running-in procedures can be found in literature as well. Currently, there is no standardized method for the derivation of running-in strategies for journal bearings without the need for experiments. This paper presents such a method based on the abrasive wear simulation tool chain. The running-in ef- = , fect is evaluated in terms of achieved wear amount, remaining asperity contact and friction energy during the procedure. The following chapter will elaborate on the necessary implementations to the wear simulation for the purpose of this work. 3 Model and Method The method used in this work is based on the works by K ÖNIG et. al. [9] and D ECKER et. al. [11]. It was developed for the calculation of abrasive wear (contour and asperity/ roughness wear) on planetary journal bearings for wind turbines and qualitatively validated by means of experiments [11]. The wear rate: Eq. 2 is calculated according to the wear law by F LEISCHER [13] as a function of the local asperity contact pressure p a,i,j , friction coefficient μ a,i,j [14], relative sliding speed of the journal bearing v S and the so-called friction energy density e R . e R is the model and material specific wear coefficient introduced by F LEISCHER [9]. The EHD/ MBS model used in this work represents a small journal bearing test rig. The test rig and the simulation model are shown in Figure 2. The MBS/ EHD model consists of two flexible bodies (housing with the journal and the shaft). The model parameters are listed in Table 1. The wear simulation method [11] is adapted to feature the simulation of time-variant operating conditions enabling the simulative investigation of different runningin procedures with changing load conditions over time. This is done by introducing varying input load values (specific pressure p̅ i and sliding speed v S,i ) to the EHD/ MBS model at each iteration step i of the wear simulation. Furthermore, a thermal modelling approach is implemented to the wear simulation. The methods presented in [11] and [9] have in common that the bearing temperature Θ B is assumed to be constant over time. In reality, the bearing temperature can vary over time as a result of fluid and asperity contact friction. As the viscosity of the lubricant is temperature-dependent, the temperature has a considerable effect on the bearing’s operating behavior. Therefore, in [15] a numerical modelling approach for calculating the temperature distribution Θ i,j of a planetary journal bearing is suggested based on a heat balance of the oil film. Based on the suggestions in [15] a simple thermal model is introduced for this work which determines a global mass temperature of the journal bearing Θ B . Heat transfer into the bearing due to friction and a dissipation term modelling heat flow from the bearing to the surrounding are considered. The heat balance implemented to the wear simulation is based on Eq. 3: , , = , , , , Parameter Value Bearing width 30 Bearing diameter 120 Radial clearance 70 Bearing material 12 2 ( = 100 ) Planet gear material 42 4 ( = 210 ) Lubricant viscosity class ISO VG 320 (PAO) Eq. 3 with α B being the heat transfer coefficient between the bearing and its surrounding structure and m B · c p being the thermal capacity of the bearing. The friction energy entry into the bearing over the area of the sliding surface W R is modelled according to [15]: Eq. 4 The time integration of Eq. 3 yields the temperature curve Θ B (t) over time. Compared to other thermal models, this approach represents a major simplification, but offers the advantage of a low computational effort. 4 Results The simulation results of the different running-in procedures are discussed below. The procedures differ in the load profile (pressure and sliding speed). In order to achieve comparability between the procedures, the entire parameterization of the model is kept constant except for the procedure itself. All procedures have the same = = ( ) = duration (t = 10 h) and at the end one additional operating point (p̅ ̅ = 8 MPa, v S = 0.2 m/ s) is simulated as a reference in order to compare the values for the remaining contact pressure p a . Pressure variation: Two exemplary simulation results for running-in procedures are shown in Figure 3. First a step-wise increase in load p̅ ̅ at constant sliding speed v S (Figure 3 (a)) as suggested by J AITNER et. al. [8] and second a load ramp at constant sliding speed respectively (Figure 3 (b)). The procedures are comparable in terms of the time integral of the specific pressure p̅ int and sliding distance d according to Eq. 5 and Eq. 6. These two values are used as an equivalence criterion in this work. Eq. 5 Eq. 6 The results indicate an almost identical amount of wear in both procedures with roughly 0.73 mm 3 of wear volume V W . The asperity contact pressure p a shown in Figure 3 was averaged over all nodes of the EHD mesh of = ( ) = ( ) Science and Research 17 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 Figure 2: (a) Journal bearing component test rig, (b) MBS/ EHD model of the test rig Table 1: Model parameters of the MBS/ EHD journal bearing test rig model (a) (b) pattern can be achieved in a shorter time with a steeper ramp-shaped load increase than with a step function without exceeding a certain asperity contact or friction energy input. Speed variation: As an alternative to the variation in pressure (gradual load increase) shown above, runningin can be achieved by gradually reducing the sliding speed v S . In [7] L INJAMAA et. al. demonstrated experimentally that a step-wise reduction in sliding speed over several hours of operation leads to a reduction in friction. Science and Research 18 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 the sliding surface. For both simulations an equal initial temperature of 60 °C is assumed. The running-in procedure with the step-wise load increase results in a higher bearing temperature Θ B in the first 2.5 h of operation than the simulation with the load steps. In Figure 3 (a) it can be seen that each new load step results in a peak in asperity contact pressure p a and an increase in bearing temperature Θ B slightly delayed. This cannot be observed in the ramp-wise load increase, where the asperity contact and temperature curve are smoother. From this, the hypothesis can be derived that a comparable wear Figure 3: Results of the wear simulation for running-in procedures with a step-wise load increase (a) and a ramp wise load increase (b) (a) (b) Figure 4: Results of the wear simulation for running-in procedures with a step-wise reduction of the sliding speed (a) and a ramp wise sliding speed decrease (b) (a) (b) Similar to the procedures tested on journal bearings in [7] a wear simulation at constant specific pressure (p̅ = 8 MPa) is performed with a step-wise decrease in sliding speed from 0.2 m/ s to 0.1 m/ s over a duration of 10 h. The integral specific pressure p̅ int , sliding distance d and the reference operating point after 10 h of operation is chosen to be equal to the procedures shown above. Similar to Figure 3 a step-wise (see Figure 4 (a)) and ramp-wise decrease in sliding speed is simulated to compare both procedures (see Figure 4 (b)). Both simulations resulted in comparable amounts of wear (0.80 - 0.82 mm 3 ) It is noteworthy that the two procedures from Figure 4 (speed) generate about 10 % more wear than the procedures with pressure variation shown in Figure 3. Just like the load steps in Figure 3, the gradual reduction in sliding speed causes a peak in contact pressure at the beginning of each new operating point. A summary of the simulation results shown above is provided by Table 2. All procedures can be characterized by identical values for the sliding distance covered during the procedure d = 8640 m and the integral specific pressure p̅ int = 156 MPa · h. The running-in procedures are compared in terms of the remaining maximum asperity contact pressure after 10 h of running-in, the time integral of the friction power (cf. Eq. 4) and the generated wear. In general, all running-in procedures yield a significant reduction in asperity contact compared to the beginning of the procedure (cf. Figure 3 and Figure 4). Additionally, no striking difference can be observed between the results of the step and ramp simulations. Here the equivalence criterion from Eq. 5 and Eq. 6 appears to be applicable (asperity contact and generated wear are almost equal). In terms of generated friction energy, the ramp procedures are favorable (~10 % less friction energy). The speed variation (procedures 3 and 4) yields about 10 % higher amounts of wear which corresponds to the higher friction energy generated during the running-in procedure. The most significant difference lies in the remaining asperity contact which is smaller in the load variation simulation results (procedures 1 and 2). Procedures 1 and 2 yield smaller amounts of wear and a slightly smaller remaining asperity contact. In the light of these evaluation metrics the load varying with a load ramp function approach appears to be favorable over the speed varying approach. a Experimental validation In this work, as described above, the wear simulation method was extended by an analytical, thermal model and the option to calculate the abrasive wear under the influence of varying loads during running-in procedures. In this chapter a qualitative validation of these features is presented. For the sake of simplicity, the validation is only demonstrated for a running-in procedure with stepwise increasing load exemplarily. For the validation an experiment on a journal bearing component test rig was performed. A step-wise increasing load procedure is chosen exemplary and for the experiment the identical test setup was used as presented in [11]. In Figure 5 (a) the measurement results for the friction moment M Fr (t) and bearing temperature Θ B (t) over time are compared with the simulation results. In total 6 consecutive load steps were tested for a test duration of 2.1 h. During the experiment the sliding speed v S was kept constant at 0.2 m/ s. Science and Research 19 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 Procedure Variable Value 1 Load step function = . Contact pressure remaining after = : , ( = ) = 0.23 MPa , ( = ) = 0.79 MPa Integrated contact friction work: = 39.93 J Wear volume reached: = 0.72 mm 2 Load ramp function = . Contact pressure remaining after = : , ( = ) = 0.24 MPa , ( = ) = 0.80 MPa Integrated contact friction work: = 36.35 J Wear volume reached: = 0.72 mm 3 Speed step function = . Contact pressure remaining after = : , ( = ) = 0.36 MPa , ( = ) = 1.01 MPa Integrated contact friction work: = 56.09 J Wear volume reached: = 0.80 mm 4 Speed ramp function = . Contact pressure remaining after = : , ( = ) = 0.37 MPa , ( = ) = 0.98 MPa Integrated contact friction work: = 45.43 J Wear volume reached: = 0.82 mm Table 2: Summary of the simulation result for four different running-in procedures of the bearing’s axial contour c B before and after the experiment are shown in comparison to the contour generated from the wear simulation. Here the axial contour of the bearing is evaluated directly in the load zone. It can be seen that wear occurred during the experiment, since a significant change in the contour is measured at the bearing edges (x B = 0 mm and x B = 30 mm), which corresponds to the simulation result as well. The running-in procedure caused a recession of the contour of approximately 11 μm at x B = 0 mm. From the validation experiment it can be concluded, that the wear simulation of consecutively increasing load steps yields realistic and reasonable results. Although the approach for modeling the bearing temperature is comparatively simple and therefore more computational efficient compared to other state-of-the-art modeling schemes, it can qualitatively model the increase of the global bearing temperature. b Simulation of running-in procedures on system level The approach presented above was demonstrated on a small component test rig for journal bearings. The simulations showed that the gradual increase in load yields slightly more favorable results than the approach with decreasing sliding speed. The operating behavior of a journal bearing always depends a lot on its system environment. In the case of a planetary journal bearing the Science and Research 20 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 One important observation is, that in terms of the mean friction moment simulation and measurement are in good agreement. The typical running-in of the bearing at each load step can be recognized by the frictional moment M Fr converging from an initially high peak value to a stationary value. During the first four load steps this effect is significantly more pronounced in the simulation than in the experiment. One possible reason is a deviation in the transition point from pure hydrodynamic operation to mixed friction between the simulation and the experiment. Additionally, the measured friction moment shows higher peaks (around 10 Nm more at the last load step) than in the simulation and the measured friction decreases significantly faster than the simulation result. One possible explanation for this deviation is the constant wear coefficient used in Eq. 2. A possible remedy to correct this deviation is the use of an alternative wear law such as the wear law presented in [16] featuring a time dependent wear coefficient. In terms of the bearing temperature Θ B a qualitative agreement between the simulation and the measurement can be observed. The temperature measurement was performed with a temperature sensor 3 mm under the sliding surface in the load zone of the bearing. Both temperature curves from simulation and measurement increase from initially 20 °C to around 35 °C towards the end of the experiment. Slight deviations of the simulated temperature from the measurement can be observed (~ 5°). In Figure 5 (b) the measurements Figure 5: Validation results for a running-in procedure tested on a journal bearing component test: (a) time series of measured and simulated friction moment and bearing temperature, (b) comparison between the measured and the simulated bearing contour after running-in (a) (b) system behavior of the gearbox in terms of deformations plays a crucial role in the wear assessment (e.g. gear mesh influences, titling of the gear wheel and bending of the planetary pin). To showcase the simulation approach of running-in procedures from this work on WT gearbox level additional simulations are performed using an MBS/ EHD model of a planetary gear stage of a 850 kW WT gearbox from a Vestas V52 turbine. The model comprises flexible bodies of the housing, planet carrier, gears and the sun shaft (cf. Figure 6) [11]. The most important model parameters are listed in Figure 6. All the other parameters are kept equal to those described in Table 1. The MBS/ EHD model of the gearbox is coupled to the wear simulation according to Figure 1. Insights into the wear simulation with the gearbox model can be found in [11]. Two exemplary simulations are performed with an either step-wise (Figure 7 a) and ramp-wise (Figure 7 (b)) increase in input torque leading to an increasing pressure. Both procedures are equivalent in terms of p̅ int and d. The simulation results shown in Figure 7 indicate a comparable amount of friction in both procedures with approximately 9 mm 3 of wear volume V W . Contrary to the results from the journal bearing test rig shown above the initial asperity contact is much higher and the increase in temperature at the beginning of the running-in procedure is more pronounced. This can be explained by the strong edge loading of the journal bearing due to the tilting of the gear wheel. This results in higher initial asperity contact pressure and higher amounts of wear compared to the test rig model. Nonetheless both running-in procedures result in a significant reduction in asperity contact compared to the initial condition of the simulation. At only 4 %, the difference between the two procedures in terms of the friction energy generated in the gearbox model is lower than in the component model. This confirms the findings described in [10] that the journal bearing behavior is strongly system-dependent. 5 Discussion and Outlook Journal bearings are a driver of torque density and reliability of wind turbines. Therefore, they contribute significantly to the improvements in cost effectiveness of wind energy achieved in recent years. The performance of a journal bearing is highly influenced by the runningin behavior at the beginning of its service life. The intentional induction of abrasive wear during the running-in phase at the beginning of the bearing service life can be beneficial to the performance of the bearing. This work Science and Research 21 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 Figure 7: Exemplary results of the wear simulation for running-in procedures with a step-wise load increase (a) and a ramp wise load increase (b) Planet gear Planet carrier Gearbox housing Sun shaft Journal Bearing 𝐵 = 190 𝑚𝑚 𝐷 = ∅178 𝑚𝑚 Figure 6: MBS/ EHD model of a planetary gear stage of a WT gearbox (a) (b) [3] Stehly, T. u. Duffy, P.: 2021 Cost of Wind Energy Review. 2022 [4] Deters, L., Fischer, A., Santner, E. u. Stolz, U.: GfT Arbeitsblatt 7 - Tribologie. Verschleiß, Reibung (Definitionen, Begriffe, Prüfung). 2002 [5] Sommer, K., Heinz, R. u. Schöfer, J.: Verschleiß metallischer Werkstoffe. Wiesbaden: Springer Fachmedien Wiesbaden 2018 [6] Scherge, M., Shakhvorostov, D. u. Pöhlmann, K.: Fundamental wear mechanism of metals. Wear 255 (2003) 1-6, S. 395-400 [7] Linjamaa, A., Lehtovaara, A., Kallio, M. u. Léger, A.: Running-in effects on friction of journal bearings under slow sliding speeds, Bd. 234. 2020 [8] Jaitner, D., Schmelzle, B. u. Fiereder, R.: Assessment of the Wear Behavior of Journal Bearings within a Planetary Gear Stage of a Wind Turbine Transmission. In: Bearing World Journal Vol. 7. Frankfurt am Main: VDMA Verlag 2022, S. 7-12 [9] König, F., Ouald Chaib, A., Jacobs, G. u. Sous, C.: A multiscale-approach for wear prediction in journal bearing systems - from wearing-in towards steady-state wear. Wear 426-427 (2019), S. 1203-1211 [10] Hagemann, T., Ding, H., Radtke, E. u. Schwarze, H.: Operating Behavior of Sliding Planet Gear Bearings for Wind Turbine Gearbox Applications—Part II: Impact of Structure Deformation. Lubricants 9 (2021) 10, S. 98 [11] Decker, T., Jacobs, G., Graeske, C., Röder, J., Lucassen, M. u. Lehmann, B.: Multiscale-simulation method for the wear behaviour of planetary journal bearings in wind turbine gearboxes. Journal of Physics: Conference Series 2767 (2024) 5, S. 52012 [12] Forschungsvereinigung Antriebstechnik e.V.: FVA755 - Gleitlagerverschleißgrenzen II. Projektabschlussbericht. 2022 [13] Fleischer, G., Größer, H. u. Thum, H.: Verschleiß und Zuverlässigkeit. Berlin: VEB Verlag Technik 1980 [14] Offner, G. u. Knaus, O.: A Generic Friction Model for Radial Slider Bearing Simulation Considering Elastic and Plastic Deformation. Lubricants 3 (2015) 3, S. 522-538 [15] Prölß, M.: Berechnung langsam laufender und hoch belasteter Gleitlager in Planetengetrieben unter Mischreibung, Verschleiß und Deformationen. Dissertation. 2020 [16] Lijesh, K. P. u. Khonsari, M. M.: On the Modeling of Adhesive Wear with Consideration of Loading Sequence. Tribology Letters 66 (2018) 3 Science and Research 22 Tribologie + Schmierungstechnik · volume 71 · issue 5-6/ 2024 DOI 10.24053/ TuS-2024-0034 presents a method for the derivation of running-in procedures for journal bearings based on abrasive wear simulation. Different load scenarios are simulated and compared in terms of friction during the running-in procedure, generated wear and the remaining contact pressure after running-in. A reference operating point was selected to compare the asperity contact pressures after the procedures. The simulated procedures are equal in terms of integral specific pressure and sliding distance. The results indicate that the approach using a gradual increase in load is favorable over the approach with decreasing sliding speed, since it generated less wear while resulting in a lower asperity contact. In terms of generated friction energy the ramp-wise increase in load is favorable over the step-wise load increase. A qualitative validation of the simulation model in terms of friction moment, bearing temperature and the generated wear contour was experimentally achieved on a journal bearing test rig. The running-in procedures with load variation were applied on a WT gearbox model and proved a transferability of the findings from the component level simulation to the system level. Overall, it can be concluded on the basis of the available results that the load variation generates a slightly better wear pattern than the speed variation and is also preferable due to the friction energy generated. A ramp-shaped progression of the load or sliding speed is to be preferred in any case, as peaks in the asperity contact are caused by the steps. Increasing the slope of the ramps offers the potential to save test time in the procedures with the same wear result. Potential for optimization can be derived from this. Further optimization of the runningin procedures will be presented in a future paper. In addition, the use of a simplified temperature modelling approach is particularly advantageous with regard to the high computational effort associated with the wear simulations of the running-in procedures. Future work will address how thermal models with a higher fidelity can be made applicable to the presented method. Literature [1] Meyer, T.: Validation of journal bearings for use in wind turbine gearboxes. inFOCUS: WINDPOWER (2015) [2] Thys, T. u. Smet, W.: Selective assembly of planetary gear stages to improve load sharing, Bd. 87. 2023