PhD Defense | Essam SERAG | Tailored Coatings for Enhanced Long-Term Tribological Performance of Low-Speed Journal Bearings in Liquid Lead-Bismuth Eutectic

Tailored Coatings for Enhanced Long-Term Tribological Performance of Low-Speed Journal Bearings in Liquid Lead-Bismuth Eutectic

The objectives of this research are to develop optimized coatings aimed at enhancing the tribological performance of journal bearings operating in Lead Bismuth Eutectic (LBE), and to gain a comprehensive understanding of the corrosion and wear mechanisms of these coated bearings in LBE.

The research followed a systematic approach to achieve its objectives. Initially, an extensive literature review was conducted to survey recent advancements and innovations in the field of mitigating LBE corrosion, and mechanical wear. Various solutions have been introduced, with some addressing only LBE corrosion, while others offer effective protection against both corrosion and wear. Among these solutions, coatings emerge as a promising and effective means of protecting structural materials and mechanical components immersed in LBE. Based on the insights gained, two promising candidate coatings have been identified for further investigation: TiAlN and carbon-based coatings.

Subsequent efforts were directed towards optimizing these selected candidates. This optimization phase involved iterative testing and refinement to ensure the coatings could effectively mitigate wear and corrosion. It focused on coatings enhancements through the control of the deposition parameters of the magnetron sputtering technique, such as bias voltage and working pressure. The final goal was to develop coatings with specific properties tailored to the unique requirements of journal bearings operating in LBE, such as high hardness, strong adhesion to the substrate, low residual stress, etc. 

The results revealed that surface roughness, hardness and residual stresses of the TiAlN coatings increased with higher bias voltage and/or lower working pressure. Samples with high compressive residual stresses severely cracked under applied normal loads using Rockwell indenter. This finding highlights the importance of balancing hardness and residual stress to achieve optimal coating performance. The optimum deposition parameters identified were a bias voltage of -100 V and a working pressure of 5 mtorr, which resulted in coatings with high hardness, high adhesion to the substrate and relatively low residual stresses while maintaining high deposition rates. 

Additionally, the material and thickness of the adhesion layers were selected based on the acquired results. TiAlN coatings demonstrated the best adhesion to stainless steel substrates without requiring a TiAl intermediate layer. Furthermore, the adhesion of TiAlN coatings increased with the bias voltage up to -100 V, beyond which no further improvement was observed, indicating a plateau in adhesion enhancement. On the other hand, carbon-based coatings showed the best adhesion when a Cr+CrN double layer was deposited on the substrate, with a deposition duration of 2 minutes for each sublayer. 

Furthermore, coated flat substrates were tested using a Tribometer setup to study the coatings tribological properties. The findings showed that the uncoated substrate exhibited the highest friction coefficient and the highest specific wear rate, highlighting its poor wear resistance compared to the coated samples. Among the coatings tested, the Ti_0.46Al_0.54N coating displayed a higher friction coefficient and lower wear resistance relative to the carbon-based coatings. In contrast, the pure carbon coating deposited via the PA-CVD technique demonstrated superior wear resistance, evidenced by a significantly lower friction coefficient and specific wear rate compared to all other coatings.

Following the optimization process, a series of LBE corrosion tests were conducted. These tests aimed to assess the compatibility of the optimized coatings with LBE and to understand the underlying corrosion mechanisms under various conditions, such as differing temperatures (250°C ≤ T ≤ 410°C), oxygen concentrations (low and high), and exposure durations (500 h ≤ t ≤ 3000 h). This phase provided critical insights into how the coatings perform and degrade in LBE. 

The XPS depth profiles of the Ti_xAI_1-xN coatings (0.38 ≤ x ≤ 0.58), immersed in high oxygen concentration LBE, revealed minimal oxidation at 250 °C even after 1000 h. However, at 360 °C, a mixed oxide layer of (Ti, Al)O_x was formed on the coating's surface. Exposure to LBE at 410 °C for 500 h led to the creation of an oxide bilayer comprising TiO₂ (outer sublayer) and (Ti, Al)O_x depleted of Ti (inner sublayer). The coating with the lowest Al concentration (i.e., Ti_0.58Al_0.42N) exhibited the poorest oxidation resistance at this temperature, nearly entirely consumed to form a thick (from 2.6 μm to 2.8 μm) and fragile oxide bilayer, rendering it unsuitable for material protection within LBE. A model is proposed to illustrate the oxidation mechanism of these coatings in high oxygen concentration LBE. For tests conducted in low oxygen concentration LBE at 250°C for 3000 h, the nitride coatings showed no signs of surface degradation, indicating their effectiveness in protecting structural components immersed in oxygen-deprived LBE.

The XPS depth profiles of the Cr_xC_1-x coatings (x = 0.2, 0.5, and 0.8), immersed in high oxygen concentration LBE at 360°C for 1000 h, showed different behaviour depending on the Cr content of the coatings. The coating with the lowest Cr content (Cr_0.2C_0.8) got completely removed, leaving the substrate exposed to LBE. This is attributed to the interaction between carbon atoms in the film and dissolved oxygen in the LBE, which resulted in the formation of CO or CO₂ that eventually escaped the system. In contrast, the other two coatings with higher Cr content developed a protective Cr₂O₃ layer, acting as a diffusion barrier for oxygen atoms. The SEM analysis of the Cr_xC_1-x coatings, immersed in low oxygen concentration LBE at 250°C for 3000 h, indicated that the Cr_0.8C_0.2 coating maintained good structural stability, while the Cr_0.5C_0.5 coating cracked and shattered, and the Cr_0.2C_0.8 coating experienced delamination.

The next phase involved applying the optimized coatings to the surfaces of journal bearing shafts. These coated shafts were then subjected to wear tests in LBE to evaluate their tribological performance. Additionally, chrome-plated shafts were also tested to compare their performance to that of the coated-shafts. Two bearing materials were used: DEVA 120, with larger carbon grains and higher carbon content, and DEVA 121, with finer carbon grains and lower carbon content. The wear tests provided valuable data on how the coatings withstand mechanical stress and friction, in addition to the failure mechanisms of the different families of coatings.

The test results indicated that the uncoated shaft tested with DEVA 120 under a low load of 500 N performed the worst, failing immediately due to adhesive and abrasive wear mechanisms. The PA-CVD carbon-coated shaft tested with DEVA 120 under a load of 1000 N failed shortly after the test began. This failure was attributed to the tribofilm wear mechanism where the carbon coating was accommodated within the irregularities of the bearing surface, followed with dissipation into LBE, exposing the shaft to the hard DEVA 120 bearing, leading to abrasive and adhesive wear, similar to the uncoated shaft.

The chrome-plated shafts showed excellent performance with DEVA 120 under a load of 1000 N, with no wear tracks on the shaft surface. This was due to the formation of a transfer layer between the mating surfaces, reducing torque and preventing abrasive and adhesive wear. However, when tested with DEVA 121, the chrome-plated shafts failed immediately with a high linear wear rate, resulting from abrasive and adhesive wear mechanisms. The lower tendency of DEVA 121 to form a protective continuous transfer layer compared to DEVA 120 led to the rapid deterioration of the chrome-plated shaft.

In contrast, the TiAlN-coated shafts displayed opposite behaviour. When tested with DEVA 120 under various loads, the coated shafts failed due to fatigue wear from the shear and normal stresses developed at the shaft surface during journal bearing operation. However, the TiAlN-coated shafts tested with DEVA 121 under 1000 N showed different results. One test failed due to poor coating adhesion, as the tested shaft was coated before the optimization of etching and cleaning processes, which influence coating adhesion. Despite this, the test lasted three times longer than those with DEVA 120 thanks to the lower average torque (lower shear stress). A repeat test with an optimized coating showed that the coating survived the entire duration, thanks to the good coating adhesion while maintaining the low average torque.

Multi-layer-coated shafts, with a bottom layer of TiAlN, an intermediate layer of both TiAlN and carbon, and a top layer of pure sputtered carbon, were tested with DEVA 120 under 500 N and 750 N. These shafts survived thanks to the lowest average torque compared to all the other RHAPTER tests, indicating the effectiveness of the top layers in reducing shear and normal stresses. The superior performance of the multi-layer coating was attributed to the lubricating effect of the top sublayer, the softer amorphous structure of the top layers reducing peak stress concentrations, and the inherent properties of multi-layer structures in suppressing crack propagation at interfaces.

It is crucial to recognize that wear is a system property rather than a material property, making it a complex phenomenon influenced by various factors. Consequently, there is no universal solution that fits all situations. Finally, a comprehensive coatings selection tree was developed. This tool synthesized the results obtained throughout the research, providing a practical framework for selecting appropriate coatings based on specific performance criteria and operational conditions.

Promoter:

• Stéphane Lucas (UNAMUR)

SCK CEN mentors:

• Paul Schuurmans
• Ben Caers