1 Introduction

Tapered roller bearings are commonly used in drive trains or wheel bearing arrangements of commercial vehicles (see Fig. 1). In the German automotive industry, more than 10 million tapered roller bearings are used annually [1]. They can support comparably high loads due to the linear contact between tapered rollers and raceway. Due to their contact angle, these loads can act in radial, axial, and combined directions. In addition, they can support high tilting moments in small installation spaces without the need for support of additional bearings. Tapered roller bearings usually operate under both grease and oil lubrication. Radial tapered roller bearings consist of an inner and outer ring between which the conical rolling elements are located. The conical rolling elements are guided along the inner ring rib. In loaded tapered roller bearings, considerable frictional dissipation occurs between the rib and the roller ends, particularly at low rolling speeds. Startup and maneuvering are common cases in which low rolling speeds occur in wheel bearings of a truck. This can lead to a risk of wear and, in severe cases, failure. The increased frictional dissipation occurs due to insufficient lubricant film thicknesses in the roller and rib contact (Fig. 2).

Fig. 1
figure 1

Flange and raceway of tapered roller bearing 31213A

Fig. 2
figure 2

DMG Mori NTX 1000

In order to avoid contact of the functional surfaces under load, sufficient lubricant film thickness must be ensured during bearing design and throughout operation to guarantee full film lubrication in almost all operating conditions [2]. When a bearing is fully lubricated, the surfaces are completely separated by the fluid film and the friction results solely from the viscous behavior of the lubricant. In case of rolling contacts, the viscosity of the lubricant is influenced by the lubricant film pressure while the surfaces of the roller and raceway experience deformation. This is referred to as an elastohydrodynamic (EHL) contact [3].

During bearing operation in the regions of boundary or mixed friction, which for motor vehicles primarily occur when starting or maneuvering, wear results from the roughness asperities contacting. The rate of wear increases with increasing load. Between the rollers and the rib there is a rolling motion with high sliding components. This makes these contacts particularly critical from a tribological point of view. Additional frictional dissipation occurs at the roller end—rib contact, due to mixed friction at low speeds and/or high temperatures, which reduce the lubricant viscosity and therefore its load carrying capacity in contact [3, 4].

The characteristics of the lubricant film in contact is significantly influenced by the surface topographies of the friction partners involved. With the aid of phase-corrected filters as described in DIN EN ISO 11562 [17], the surface profile of the rolling partners can be divided into core, peak, and groove areas [5]. The peaks interfere with the formation of the lubricant film, while grooves, in which lubricant can accumulate in some circumstances, can have a positive effect on the formation of the lubricant film. This effect can contribute to the reduction of frictional losses [6]. In the case of mixed friction, low peak heights favor an even distribution of the induced load.

Known from applications of the automotive industry, e.g. cylinder liners, micro-lubrication dimples are used to induce additional lubricant into the contact zone. Such modified surfaces show a great potential for friction minimization compared to the standard honed surfaces of cylinder liners. This is due to the fact that the lubrication dimples increase fluid film pressure, which extends the area of fluid friction in the Stribeck curve towards lower speeds [7].

For the manufacturing of microstructures, both cutting and non-cutting processes are used. Non-cutting processes include, for example, microforming [8] and laser material removal processing [9]. Among the metal-cutting processes, micromilling has proven to be advantageous, especially due to its high flexibility in terms of shape and position as well as the possibility of process integration [10]. In this context, Kästner investigated the effects of simultaneous turn-milling on the formation of microstructures. Tactile measurements show that the dimple depth deviates from the set depth of cut by up to 6 μm (30%) due to tool displacement. The reduced effective depth of cut also leads to smaller structure dimensions in the longitudinal and transverse directions. This has a significant influence on the tribological effect of these structures [11]. Despite known individual findings and their high potential for friction reduction, turn-milling of micro dimples has not yet been implemented for the production of hardened component sections such as tapered roller bearings. The reason for this is that there is neither sufficient knowledge of the prevailing relationships between friction dissipation and the necessary lubrication dimple geometry nor of suitable machining concepts. For this reason, the investigations presented here aim to implement a suitable machining strategy for the production of defined microstructures for tribologically optimized applications and to identify relationships between geometry formation and process parameters.

2 Experimental setup

2.1 Materials

The machined specimens were tapered roller bearing inner rings of type 31312-A (FAG, Austria) made of hardened 100Cr6 with a hardness of 62 ± 2 HRC.

2.2 Cutting tools

Solid carbide thread milling cutters with three cutting edges of the type MicroMill 05 053 010 (WNT, Luxembourg) were used to produce the microstructures. In order to use the thread milling cutters for structuring, two of the three cutting teeth were removed with a tool grinding process. The tools have a diameter of d = 5.8 mm.

2.3 Manufacturing process

The lubrication dimples are milled into the taper roller bearings on an NTX1000 turn-mill center (DMG Mori, Germany). In addition to the turret for turning, the turn-mill center has a swiveling milling head. This allows the milling spindle to be orientated at the same angle as the rib. The cutting parameters used are shown in Table 1.

Table 1 Parameters of the experimental setup

2.4 Surface topography measurements

The measurement of the induced microstructures was carried out on a tactile surface and contour measuring device of type MahrSurf LD130 (Mahr, Germany) and with a digital microscope of type VHX 600 (KEYENCE DEUTSCHLAND GmbH, Germany) The evaluation of the recorded surfaces was carried out with the software “µsoft analysis premium 7.11”® (NanoFocus AG, Germany). An optical 3D measurement system Infinite Focus XL200 G5 (Alicona Imaging GmbH, Austria) was used for measurement of the tools and for the generation of STL models for the material removal simulation. Within the scope of the investigation, a 20 × objective was used, which provides sufficient measurement quality for the investigated cutting edge geometries.

2.5 Simulation

The simulative design of the microstructure milling process was carried out with the dexel-based material removal simulation software IFW CutS®, which was developed at IFW [12]. For the investigation of the resulting structure geometry, a multi-dexel model was set up, in which an axial section of the inner ring of the taper roller bearing was discretized. To determine the theoretically optimal geometry, orientation and density of the lubrication dimples, a process model of the microstructure milling process was built up using the IFW CutS software. For this purpose, a kinematic model of the machine tool NTX1000 was created in the CutS environment so that the axis movements of the turn mill machine tool are digitally replicated. The core of the material removal simulation is a dexel model of the workpiece, in this case the tapered roller bearing to be structured. The dexel-based simulation was used to determine the material removal achieved by the machining process. The result of the simulation is the final geometry of the machined workpiece. Using an evaluation routine developed in Matlab® (MathWorks, USA), the topography of the inboard and the resulting geometries were exported and analyzed.

3 Microstructure milling process

3.1 Design of the microstructure milling process

The design of the microstructure milling process is based on the kinematics of turn-milling. In turn-milling, turning and milling are usually combined by a driven tool DIN8589 [18]. Turn milling is divided into orthogonal and parallel turn milling [13]. In addition to orthogonal and axial-parallel, tangential turn-milling was first introduced in 2001 by Dietzel in [14]. By introducing additional inclination angles in the X–Y, as well as in the X–Z plane, additional process strategies can be developed [15].

In [11], Kästner derived a wide variety of process strategies for microstructuring, starting from the process strategies of turn-milling and applying them to the machining of cylindrical components. According to Kästner, parallel microstructure milling was applied for both internal and external machining. Kästner used the orthogonal process strategy to structure cylindrical components orthogonally on the outside and orthogonally on the faces of workpieces. With Kästner’s orthogonal face process, tangential and radial structures can be created (Fig. 3). With regard to the development of a suitable process strategy for machining the tapered rib surface of the tapered roller bearing, the machining strategies developed by Kästner for the orthogonal external structuring process and the orthogonal end-face process can be combined here with the process strategies of tangential turning. The introduction of an additional tilt axis expands the spectrum of process strategies for turn-milling.

Fig. 3
figure 3

Axis-parallel and orthogonal turn-milling kinematic [11]

In order to produce large-area microstructures, such as in cylinder liners, continuous process strategies are required in order to achieve cost-effective production. Due to the comparatively small surface area of the inboard wall, discontinuous process strategies are also relevant, since only a small number of structure rows can be distributed over the circumference without overlapping the structures.

3.2 Kinematics of the microstructure milling process for tapered roller bearings

In contrast to Kästener, where only cylindrical components were machined, the structure alignment angle β is introduced for the application case of tapered roller bearings (see Fig. 4). This angle can be used to describe the deviation of the structure alignment from the tangential position. The structure angle can be described by a deviating angle from a tangential structuring position of the tool, which produces a concentric structure on the inner rib surface, to an orthogonal tool position, which results in a radial alignment of the structures and has a structure alignment angle of 90°. By using this additional tilt angle α in the microstructure milling process, the symmetry of the resulting microstructures can be influenced. An orthogonal alignment of the symmetry axis of the tool cutting edge to the structured surface results in a symmetrical microstructure. Deviation from this can produce asymmetrical structures.

Fig. 4
figure 4

Influence of the structure alignment by the tool position for synchronized and reverse rotation

Further necessary technological variables for the design of a process strategy for microstructure milling are the rotational speed of the tool nWZ, the rotational speed of the workpiece nWS, and the resulting speed ratio λ. These, along with the newly introduced structure alignment angle β, have a significant influence on the geometry and spacing of the lubrication dimples. In the following, the calculation of the dimple length for an orthogonal face-side process strategy is described and it is shown how the structure alignment angle affects this. With an increasing angle of the structure position, there is an increasing inclination of the structure in the contact area with respect to the direction of rotation. An orientation of 90° results in radial alignment of the microstructure in the contact area. However, a previously executed collision analysis had determined that only a maximum alignment angle of − 70° to + 70° is feasible due to the kinematics of the machine tool. With an alignment of 0°, synchronized and reverse rotation has to be considered due to the compression or expansion of the dimple (see Fig. 4, right). At an alignment angle β of 90°, this distortion no longer occurs. In the positions in between, the effect decreases with increasing structure angle.

The speed ratio λ describes the ratio of the tool rotational speed and the workpiece rotational speed. This ratio also corresponds to the number of tool engagements per tool revolution.

$$\uplambda =\frac{{n}_{wz}}{{n}_{ws}}.$$
(1)

The spacing of these dimples can be described as the tangential spacing stan for the orthogonal face machining strategy with the tangential structure alignment. The tangential spacing stan is calculated from ratio of circumference of the rib diameter UWS and the tangential distribution ttan.

$${s}_{tan}=\frac{{U}_{ws}}{{t}_{tan}}.$$
(2)

The tangential distribution ttan can be determined via the speed ratio λ as follows:

$${t}_{tan}=\frac{2*\pi }{\lambda } \mathrm{with} \lambda =\frac{{n}_{wz}}{{n}_{ws}}.$$
(3)

The calculation of the dimple length lMS consists of two terms. The first term describes the length that the tool is engaged during cutting of the dimple in the workpiece without rotation. The second term describes the compression or stretching of the dimple due to the rotation of the workpiece.

$${l}_{MS}=2*\sqrt{{a}_{p,max}*{(D}_{WZ}-{a}_{p,max})} \pm {D}_{WS}*\mathit{sin}\left(\frac{{\varphi }_{s}}{2}*\frac{1}{\lambda }\right).$$
(4)

Kästner also describes the influence of the machining mode and shows that, with smaller speed ratios, the difference between synchronized and reverse rotation is clearly noticeable. It results in stretching during reverse rotation machining and in compression during synchronized rotation machining. As the speed ratio increases, the difference asymptotically approaches zero and becomes negligibly small from a speed ratio λ > 200 for the application case of the cylinder liner [11]. For the application case of the tapered roller bearing, the tool has a substantially smaller diameter DWZ in relation to the component. For this reason, the compression or stretching of the micro-lubrication pocket due to the workpiece rotation has to be taken into account.

The engagement angle φs describes the angle of a single tool rotation at which contact occurs between the cutting edge and the workpiece. This angle can be determined via the tool diameter DWZ and the chord length of the dimple during a stationary structuring process without additional feed motion.

$${\varphi }_{s}=\frac{2*\sqrt{{a}_{p,max}*{(D}_{WZ}-{a}_{p,max})} }{{D}_{WZ}}.$$
(5)

The procedure described here can be used to calculate the length of the microstructures resulting from an orthogonal face machining strategy. This results in structures distributed tangentially over the circumference on the inner side wall or rib. If the speed ratio is increased for this process strategy, the tangential distance stan between the structures is reduced. If the dimple length lMS corresponds to the tangential spacing stan of the microstructures, the structures overlap and a groove is formed. The speed ratio limit λlimit describes the point from which on the created dimples are overlapping. If the process is carried out below the limit speed ratio, the process generates dimples that do not intersect with each other.

The distortion term of the dimple length lMS is influenced in its sign by whether the machining is performed in synchronized and reverse rotation. This difference also influences the l speed ratio limit λlimit with otherwise constant process parameters. For the machining of dimples the depth of cut ap in a range of 0.01–0.05 mm is relevant. In this range, the difference of the speed ratio limit for synchronized and reverse rotation is Δλlimit = 32.6 for a depth of cut ap = 0.01 mm and Δλlimit = 30.1 for a depth of cut ap = 0.05. For the relevant range of process variables, the difference can therefore be assumed to be constant. The depth of cut ap has a more significant influence on speed ratio limit λlimit. With a depth of cut ap = 0.05 mm, an overlap of the dimples already occurs at a speed ratio limit λlimit = 287 with synchronized rotations. A depth of cut ap = 0.05 mm results in a on speed ratio limit λlimit = 612 for synchronized rotations. Another influence on the limiting speed is the tool diameter, which is assumed to be constant for this investigation.

For a successful microstructure milling process, the speed ratio λ must be kept below the limiting speed ratio λlimit. If the relationship lMS < stan is maintained, separated dimples can be obtained without any overlap.

3.3 Simulative investigation of the microstructure milling process

In order to be able to reproduce the process in detail, the components from the test setup were implemented as CAD models into the CutS simulation environment. Figure 5 shows the setup of the material removal simulation. In the right of Fig. 5, the analytical calculation of the dimple length lMS is plotted against the maximum depth of cut. The figure shows that at constant depth of cut ap, there is a reduction in the dimple length when the structural angle β is increased. This distortion is caused by the tool position. When the tool is positioned radially, the velocity vectors of the tool and workpiece motion are orthogonal to each other, so they do not add or subtract depending on the rotation. For comparison, the dimple length lMS was also plotted for structure alignment angles β = 0° and 90°. The tangential structures with a structure angle of β = 0° can be described by the analytical solution for the orthogonal face and there is a high agreement with the results of the material removal simulation. For the analytical consideration of the dimple length at a structure alignment angle of 90°, the idealized dimple length was calculated, since neither compression nor stretching of the dimple length occurs here. Since Kästner investigated only ideally flat end face, his approach cannot be used for the inner rib of the tapered roller bearing. This analytical solution is not able to consider a deviating structure alignment angle. For this reason, the material removal simulation (Dexel density: x = 100 mm−1, y = 100 mm−1, z = 300 mm−1) was used at this point to predict the resulting lubrication dimple geometry.

Fig. 5
figure 5

Simulated and analytically calculated dimple length lMS depending on the depth of cut with reverse rotations

For the simulative design of the microstructure milling process, a wear analysis of the tools used is carried out. Wear must be taken into account to a significant extent in microstructure milling processes, since initial wear results in a considerable reduction of the tool radius. For this purpose, flat surfaces of the tapered roller bearing were milled with approx. 26,500 microstructures. The process parameters vc = 200 m/min, ap,max = 50 µm were kept constant.

3.4 Tool wear behavior

The results of the wear analysis are shown in Fig. 6. The wear was quantified by the measurement of the tool radius, because dimple depth and subsequently the dimple geometry is depending on the tool radius. It can be seen that the tool diameter decreases by 53 µm after 26,500 microstructures, which corresponds to approximately one row of structures on the inner rib. The initial wear of the cutting edge exceeds in this case the depth of the dimple. Over the tool life of the milling cutter there is a further reduction of the radius by 24 µm. This change can be assumed to be a linear relationship when machining the inboard wall and must be taken into account to compensate for the resulting deviation in the depth of the dimple.

Fig. 6
figure 6

Wear analysis of the structuring tool. 1. working sharp tool 2. initial wear condition of the tool 3. worn tool

The real geometry of an initially worn milling tool, in order to depict the real process in the simulation, was measured with the Infinite Focus XL200 G5. In edge Fig. 6-2 the measurement is visualized. The point cloud was exported as an STL-file to implement the real tool geometry in the material removal simulation. Due to the corner radius of the tool, a running-in behavior occurs when the tool is used, especially at the beginning of its service life, during which the resulting structure geometry changes. After the initial wear of the tool occurs, a constant dimple geometry is produced until the tool fails. The results of the material removal simulation with consideration of the tool wear for different machining angles are shown in Fig. 7. The dexel orientation is shown on the left of the figure. The x-dexels are radial, the y-dexels are tangential and the z-dexels are axially oriented on the model of taper roller bearing 31312A. The calculation step Δφ describes the angle step width in which the machining process is discretized in the material removal simulation.

Fig. 7
figure 7

Material removal simulation with worn tool for the machining with reverse rotation of lubrication dimples

The topography simulation shows a dependency between the shape and geometry of the dimple and the machining position. A variation of the structure alignment angle causes a change in the dimple length, structure width, and structure volume while the dimple depth remains constant. The fact that the rib surface can be described by a truncated cone surface leads to a non-linear behavior of the structure geometry. The different structure alignment angles are realized by moving the tool circumferentially along the inner rib of the tapered roller bearing. This results in a tilt of dimples on the inner rib surface. The left part of Fig. 7 shows the positions that belong to the respective topographies of the simulated rib (right). An orientation of 0° allows a concentric structure to be generated in the contact area of the inner rib.

3.5 Displacement behavior of the tool

Hardened bearing material such as 100Cr6 poses a particular challenge for the milling of microstructure geometries to specifications. With a hardness of 62 ± 2 HRC, the present process can be classified as hard machining. The thin shank of the milling tool ensures that the inner rib can be reached in different alignment positions. In order to be able to describe the displacement behavior an empirical compensation approach was developed (Fig. 8). For this purpose, the milling tool was used on a planar surface with the diameter of the inner rib. These experiments were conducted to investigate the difference between the set depth of cut ap and the actual resulting dimple depth t.

Fig. 8
figure 8

Tool displacement depending on the depth of cut ap

In order to avoid that process parameters are examined without material removal, a low dimple depth of approx. 5 µm is already set at 0 µm depth of tool engagement. The investigation shows that with a set depth of cut of ap = 50 µm, there is a displacement of 41% in the resulting dimple depth t. To achieve a defined dimple depth of 30 µm, this relationship has to be taken into account accordingly. In the subsequent tests, the displacement values determined were used to compensate for the tool displacement during experimental microstructuring. The aim was to set the desired dimple depths with high accuracy or reproducibility.

4 Experimental validation

4.1 Validation of the microstructure milling process

The simulated surface topographies were directly compared with the experimentally generated surface topographies. Figure 9 shows the three investigated depths of cut of ap,max = 10, 20 and 30 µm. While the simulation is able to predict the dimple depth, there are deviations in the prediction of the dimple length. Although the real tool geometry was taken into account in the material removal simulation and the displacement of the tool was compensated, these deviations still occur. In addition to the tool cutting geometry and the tool displacement, there are other aspects that influence the geometry of the microstructures. Material effects cannot be easily taken into account with material removal simulation. This leads to an additional distortion of the dimple geometry. The dexel-based material removal simulation also cannot take into account the elastic deformation of the workpiece. Furthermore, material removal is represented in the simulation even if the required minimum chip thickness is not reached when the cutter is plunged into the material.

Fig. 9
figure 9

Comparison of simulated and measured microstructures with a structure angle of β = 0°

In order to determine ideal microstructure geometries for tribological contact on the inner rib of the tapered roller bearing, the influence of the structure alignment was also investigated. The structure alignment is determined by varying the structuring position angle of the tool. Figure 10 shows structure orientations of 0°, 45°, and 70°. Here, the predicted surface from the material removal simulation and the measured structure on the inner rib are also compared. As in the investigation of the 0°-oriented structures, a reduced dimple length also occurs for the dimples with a structure alignment of 45° and 70°. For the face orthogonal process strategy with an angle of 0°, there are only deviations in the length. With an increasing structure alignment angle the lengths and depths of the dimple also deviate.

Fig. 10
figure 10

Comparison of the simulated and measured microstructure as a function of the structure angle

The determined values for tool displacement compensation of the 0°-oriented dimples can only be used to a limited extent for the 70°-oriented structures, since in addition to the orientation of the structure, the tool engagement situation also changes with the variation of the structure alignment angle. In Fig. 8 it is illustrated, that at a structuring position of 0° the tool rotation axis is orthogonal to the raceway and parallel to the inner rib orientated. This orientation can be achieved in the 0° position by the inclination of the swivel axis by 20°. This is the pitch of the truncated cone surface. With a deviating structuring position this orientation cannot be achieved. (Fig. 10, left). As the structuring angle increases, the influence of the taper slope increases. When generating approximate radial structures, this influence can no longer be compensated by the process strategy.

A significant problem that arises with hard microstructure milling is burr formation. At a cutting depth of more than 20 µm depending on the cutting speed, the burr formation on the dimple edges are less pronounced. However, if a cutting depth of less than 20 µm occurs, then there is a slight increase in burr formation. This can be attributed to the minimum chip thickness. Due to the higher proportion of microploughing during chip formation, there is a significant increase in burr height (Fig. 11). The unfavorable cutting conditions at 45° and 70° orientation of the structure result not only in increased tool wear, but also in more pronounced burr formation. However, this is undesirable since burr formation has negative effects on the operating behavior of the tapered roller bearings. The burr causes pressure spikes, meaning it can be worn off during operation and thus causes debris that is propagated through the bearing. The resulting debris can penetrate the lubricant, leading to more solid–solid contact and associated higher degrees of friction and wear.

Fig. 11
figure 11

Influence of cutting depth and cutting velocity on the burr formation

4.2 Validation of the tribological effects

A short excerpt of the results of tribological evaluation is shown in Fig. 12. The tribological assessment is presented in [16]. The test equipment is based on the FAG-FE8 setup that is standardized according to DIN 51819. The bearings ran for 30 min with an axial load of 40 kN and at 10 RPM, i.e. enough time to observe a steady state in the measured torque. There are two sometimes counteracting physical effects that affect the tribological system: the hydrodynamics effect of the dimples and an improvement of grease starvation conditions [16]. Elastohydrodynamic simulations show that a positive hydrodynamic effect does not occur for dimples deeper than approximately 1 µm depth [16]. For this reason, a very low grease quantity of 10% of the free volume was selected in the tests in an attempt to create starved grease lubrication conditions. A very low speed of 10 min−1 was chosen in order to cause mixed lubrication conditions for the flange-roller end contact. This testing procedure is intended to increase the positive effect on the lubricant supply from the dimples. The bearings were greased with LiX-PAO110 grease. In case of the dimples with a depth of 10 µm a positive effect can be seen for the first seconds of the tests. Here the grease and its thickener structure is apparently distributed more favorably by the dimples. Also, potential shows that with further reduced dimple depths, a more favorable frictional torque can be expected.

Fig. 12
figure 12

Comparison of torque curves between microstructured bearing pairs with a cutting depth of 10 and 20 µm and reference bearings without dimples for the grease LiX110 at 1 mm/s sliding speed

After the running-in phase, the measured friction torque of the structured bearings was higher than the reference. An increasing dimple depth results in a higher friction torque, corroborating the results of the simulation in [16]. In the case of the bearing presented here, no friction reduction after the running-in phase could be achieved for the tapered roller bearing through microstructure milling, though a friction reduction during the running-in phase was achieved. The inspection after the tribological testing showed that also the dimples with a low depth are retained despite wear on inner rib. The tests with structured tapered roller bearings showed a positive influence on the running-in behavior. This short-term effect can be attributed to a favorable influence on the grease distribution in the bearing. This behavior shows that the use of dimpled bearings is more suitable for low velocity applications such as lifting equipment, where the structures act as lubricant reservoirs to protect against deficient lubrication.

5 Conclusion

Due to the supporting load predominately with line contacts, tapered roller bearings are suitable for applications with high loads and are therefore preferably used in the drive train or in wheel assemblies of commercial vehicles. In order to avoid contact of the functional surfaces under load and to minimize the resulting frictional dissipation, a sufficient lubricant film thickness is to be ensured during bearing operation. Particularly when motor vehicles are beginning to move, bearing operation is in the boundary and mixed friction regimes, resulting in wear due to surface asperity contacts. The lubricant film in this lubrication regime is significantly influenced by the surface topography of the friction partners. This dependency can be used for the inner side of tapered roller bearings. However, this technology has not been implemented for tapered roller bearings to date because there is neither sufficient knowledge of the prevailing relationships between load stress and the necessary lubrication pocket geometry, nor are suitable machining concepts available. The aim of this research was therefore to implement a suitable machining strategy for the production of defined microstructures for the optimization of tribological contact at the inner rib in hardened 100Cr6 tapered roller bearings.

To optimize the roller-rib contact in a tapered roller bearing, a possibility is the use of lubrication dimples. In this work, a microstructure milling process that was designed with the aid of a material removal simulation is presented. Here, the kinematic limits of the machine tool were determined under which it is possible to induce microstructures. Furthermore, tool wear and tool displacement were considered in the simulative design of the manufacturing process. Subsequently, the findings from the simulation were used to induce microstructures on tapered roller bearings made of hardened 100Cr6. The main findings of the simulative design and the experimental investigations are:

  • Due to the use of the selected tool concept, there is tool displacement during hard milling of lubrication micro dimples. At cutting speeds of vc = 200 m/min, this is up to 41% at a maximum depth of cut of 30 µm. This correlation was considered during process adjustment to compensate for tool displacement.

  • Despite the consideration of tool wear and tool displacement, deviations occur in the prediction of the dimple length. This can be attributed to material effects, which cannot be taken into account in the dexel-based material removal simulation. Furthermore, it is not possible to consider the elastic deformation of the workpiece in the dexel-based material removal simulation.

  • The determined values for the compensation of tool displacement for the 0° structures can only be used to a limited extent for the 70° structure because, in addition to the orientation of the structure, the tool engagement situation also changes during the micro milling process. Due to the inclination of the inner rib, the depth of cut is varying, which also increases the tool displacement. Furthermore, tool wear increases due to changed tool engagement.

  • A major challenge that arises with hard microstructure milling is burr formation. With increasing cutting speed, the burr formation on dimples edges becomes less pronounced. The burr formation on the microstructure can also be reduced by the structuring position. The unfavorable cutting conditions at 45° and 70° orientation of the structure result not only in increased tool wear, but also in more pronounced burr formation.

  • Furthermore, with this procedure, new tools for the microstructure milling of geometrically complex contact surfaces, such as the inner rib of the tapered roller bearing, can be tested virtually, flexibly and efficiently.

  • The characterization of the surface integrity in the dimpled functional surfaces by means of X-ray residual stress measurement cannot be performed reliably. For residual stress measurement using the sin2Ψ method, a sufficiently large volume of the workpiece must be recorded, which is why standard laboratory collimators often have diameters of 2 mm. For the measurement, the specimen should ideally have a flat surface. If a structured specimen is measured, a projection of the structures is recorded during inclination and the measurement is falsified by the geometric effects of the specimen. For this reason, further research is needed here.

The presented investigations show the feasibility for the manufacturing of microstructures for components with complex geometries. The shown process strategies for hard milling of dimples can be adapted for other complex tribological contact surfaces of several other applications. With regard to series production of these functional surfaces, an increase in productivity can be achieved by multi-bladed tools. For the preparation of a series process for milling of dimples on the inner rib of the taper roller bearings, the friction reduction must be verified and quantified in a next step.