24th International Colloquium Tribology - January 2024 153 Development of a Digital Twin through Simulation of PVD/ PACVD Coatings Analysis of Both Dry and Lubricated Conditions Vincent Hoffmann 1* , Emanuel Tack 2 , Nick Bierwisch 3 1 Tribo Technologies GmbH, Barleben, Germany 2 Oerlikon Surface Solutions AG, Balzers, Liechtenstein 3 SIO - Saxonian Institute of Surface Mechanics, Ummanz, Germany 1. Introduction A better understanding of coatings and, in particular, their behavior in tribological systems is crucial for the optimization of coated components with regard to their area of application. The development of a digital twin in combination with laboratory and functional tests helps to accelerate the selection of a suitable coating for an application more quickly. In addition, such a model, which includes the coating, the substrate and all existing interfaces, helps to better understand the mechanical-technological behavior of coating systems. Simulations can help to uncover optimization potential within the substrate and layer architecture and to derive a targeted optimization of the coating architecture. 2. Methodology First the material parameter including the Young’s modulus, yield strength and tensile strength of the materials are determined to gain necessary input data for the simulation models. As a first simulation a dry scratch test shall be modelled with the help of analytical models. The work is illustrated using the example of a DLC coating produced by PVD/ PACVD processes. The method is described in detail in [1] and has already been successfully applied in [2] for PVD coatings and in [3] just as successfully for DLC coatings. As a second simulation a lubricated SRV test of the same coating system is analyzed with the help of thermo-elastohydrodynamic (TEHD) simulations. The results are validated by SRV model tests. Both the hydrodynamic and the solid-state load component within the contact, their prevailing temperatures and the lubrication film thickness can be calculated. The generated frictional losses can be evaluated in combination with the stress distribution within the layer architecture and in the substrate, allowing the interaction with the opposing body to be visualized.This approach paves the way for the prediction of the wear mechanism and rate, as well as the locally prevailing friction coefficient for the selected load collectives in the lubrication area of mixed friction. The introduction of additional features of the coating system, such as the surface finish or the properties of the oil in the lubricated system, can help to predict possible failure mechanisms before selecting a coating for application. The underlying equations and implementation has been validated successfully eg. in [4, 5 and 6]. 3. Determination of Material Parameters In order for the layer model to be analyzed from bottom to top, i.e. to be built up in the digital twin, a calotte is ground into the coated sample. The structure of the model starts with the substrate, continues with the adhesive layer, the intermediate layers and ends with the top layer. To measure Young’s modulus and hardness, a classical nanoindentation with a load of 2 mN is performed. Together with the Poisson’s ratio and the extended standard method, the individual Young’s moduli and yield limits of the individual layers are determined. The influence of the underlying layer structure is taken into account. Table 1: Determination of material parameters Calotte Grinding Nanoindentation Create access to individual layers Determine hardness, yield and tensile strength 4. Simulation of Dry Contact In order to determine the mechanical-technological limits of the coated substrate, critical load cases are modeled by increasing the normal force in a scratch test (preliminary test). As the load increases, the substrate deforms plastically and microcracks appear on the surface in the a-C: H top layer. For the simulation of the dry contact two different surfaces are analyzed. Whereas the coating is identical for the two brushed specimens, only the substrate of one specimen is polished for this analysis. Therefore, the rough specimen has a roughness depth of 0.3 µm and the smooth specimen of 0.05 µm. Both specimens will be tested in a scratch test with 4.16-N. In addition, the material stress at defined points of time and positions are calculated. Due to the differences in the surface structure, the stress in the material varies. 154 24th International Colloquium Tribology - January 2024 Development of a Digital Twin through Simulation of PVD/ PACVD Coatings Table 2: Results for Scratch Test Rough Specimen Smooth Specimen Microscope Images of Scratch Calculated von Mises Stress A critical value of 6 GPa was found within the surface layer for the von Mises stress, where local layer spalling occurs. For values smaller than 6 GPa micro cracks are created in the top layer. When the stress exceeds 6 GPa, chipping cracks are created behind the indentor. In this analysis the critical value is not exceeded for the smooth specimen but for the rough specimen. 5. Simulation of Lubricated Contact To investigate wear behavior and rate under lubricated conditions, one can carry out SRV tests. If hydrodynamic effects in combination with a contact of rough surfaces shall be analyzed, one can apply TEHD simulation in combination with a mixed friction model. Bases on the resulting pressure distribution it is possible to evaluate resulting material stress. To investigate this method two different SRV tests are carried out, which are summarized in table 3. Table 3: Test Parameters of the SRV Test Parameter Decreased Hydrodynamics Test Increased Hydrodynamics Test Ball Diameter 10 mm 10 mm Stroke 4.6 mm 4.6 mm Frequency 20 Hz 30 Hz Load 200 N 5 N Temperature 150 °C 35 °C Whereas one test is a typical SRV test where the effect of hydrodynamics is mostly avoided. The second has increased frequency and dynamic viscosity due to lower temperature of the lubricant. In addition to this the load is reduced significantly to provoke a visible effect of the hydrodynamics in the test. Table 4: Results for SRV Test Decreased Hydrodynamics Test Increased Hydrodynamics Test Microscope Images of Wear Mark Stress Distribution in Coating Layers The simulation could show that for the test with increased hydrodynamics up to 33% of the load is carried by hydrodynamics, hence reducing the degree of the roughness contact which in turn leads to a significantly reduced wear in this area. This could be proofed by the respective wear mark which shows a significant width reduction in the center of the sliding length. In comparison the “decreased hydrodynamics test” shows a uniform wear progression along the complete wear mark. Simulation could show that only a maximum of 0.5% of the load was carried by hydrodynamics. Based on the resulting contact pressure distribution resulting from hydrodynamic and solid contact pressure the stress distribution for both tests could be evaluated as well. Here one can clearly identify the formation of maximum stress values in the intermediate layers. 6. Conclusion With the help of simulation a coating system under dry and lubricated conditions could be simulated successfully. The derived distribution of the material stress correlates with the test data and hence paving the way to establish simulation as a tool to improve coating layers or apply as a digital twin for lubricated applications. References [1] N. Schwarzer, “Scale invariant mechanical surface optimization applying analytical time dependent contact mechanics for layered structures”, 2017. [2] N. Schwarzer, N. Bierwisch et al.: ”Optimization of the Scratch Test for Specific Coating Designs”, Surface and Coatings Technology, volume 206, 2011. [3] M. Zawischa, S. Makowski et al.: “Scratch resistance of superhard carbon coatings”, Surface and Coatings Technology, Volume 308, 2016. [4] M. Kroneis, L. Bobach et al.: „Calculation of the actuator system in swash plate axial piston machines by a coupled multibody and TEHL simulation”, J Engineering Tribology, volume 236(5), 2021. [5] L. Bobach, J. Mayer et al.: „Reduction in EHL Friction by a DLC Coating”, Tribology Letters, volume 60, 2015. [6] L. Bobach, R. Beilicke, D. Bartel: „Transient thermal elastohydrodynamic simulation of a spiral bevel gear pair with an octoidal tooth profile under mixed friction conditions”, Tribology International, volume 143, 2020.