Limiting PcV and Tribological Performance of Mechanical Seals with Hard–Hard and Soft–Hard Tribopairs Under Water Lubrication
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Materials |
| Published | 2026-09-08 |
| Authors | Xiaoming Ren, Hui Song, He Li, Yuanren Zhou, Muqing Li |
| DOI | 10.3390/ma19183810 |
Polished Graphite-MCD seal faces reached a limiting PcV of 90.73 MPa·m·s−1 in deionized water, about 81% above Graphite-SiC. The same pair produced the lowest rotating-ring wear rate and the lowest stationary-ring wear-step height. The results show why low steady-state friction alone does not define seal reliability.
Limiting performance under water lubrication
Section titled “Limiting performance under water lubrication”Three seal-face configurations were compared under the same limiting PcV protocol. The SiC-SiC pair failed first, with severe friction-torque fluctuations near 300 r·min−1. Graphite-SiC failed by seal leakage at 5200 r·min−1. Graphite-MCD showed no pronounced fluctuations until 9400 r·min−1.
Graphite-MCD therefore operated at a higher speed and load than Graphite-SiC before reaching its failure criterion. Steady-state friction coefficients were approximately 0.032 for Graphite-SiC and 0.041 for Graphite-MCD. Near its limit, Graphite-SiC friction rose abruptly from about 0.04 to 0.11. That increase indicates loss of the previously stable interfacial condition. Low average friction therefore did not predict limiting performance.
| Seal-face pair | Limiting PcV (MPa·m·s−1) | Observed failure behavior |
|---|---|---|
| SiC-SiC | 3.40 | Severe friction-torque fluctuations near 300 r·min−1 |
| Graphite-SiC | 50.19 | Seal leakage at 5200 r·min−1 |
| Graphite-MCD | 90.73 | Pronounced friction-torque fluctuations at 9400 r·min−1 |
Test method and PcV calculation
Section titled “Test method and PcV calculation”The limiting test used deionized water at 0.2 MPa chamber pressure and 2000 N axial load. The rig began at 200 r·min−1 and increased speed by 100 r·min−1 each minute. Testing stopped at failure or 10,000 r·min−1.
Three independent tests were run for each pair. The median limiting PcV represented each pair, while minimum and maximum values defined the error range.
Seal failure was any of three conditions: a leakage alarm, persistent abnormal noise, or friction torque above 20 N·m. The stepped face had a nominal contact area of 753.98 mm². This area used outer and inner diameters of 64 mm and 56 mm. Relative sliding velocity used the 60 mm mean track diameter. Nominal contact pressure was fixed at 3.0723 MPa for the limiting calculation. PcV therefore represented pressure multiplied by velocity at the stepped sealing track.
MCD coating preparation
Section titled “MCD coating preparation”The seal included phenolic-resin-impregnated M106F graphite and pressureless-sintered SiC with 98.5% purity. The MCD counterface had grains approximately 4 to 8 µm across and was deposited on SiC by hot-filament chemical vapor deposition. Reaction gas was CH₄:H₂ at 45:1500 sccm.
Tantalum filaments were 0.5 mm in diameter and positioned 10 mm from the substrate. Adjacent filaments had 15 mm spacing, each received 1 kW, and the filament temperature was approximately 1300 °C. Chamber pressure was 4 kPa, and deposition lasted 4000 min.
The deposited layer was dense, columnar, and free of obvious pores, cracks, or delamination. Mechanical polishing used the SP40 system at a constant 3.5 kg load for 6 h.
Polishing preserved the diamond structure
Section titled “Polishing preserved the diamond structure”Polishing removed protruding grain tops and reduced coating thickness from 10.4 µm to 4.9 µm. Flatness deviation fell from 5.876 µm to 0.097 µm. AFM arithmetic roughness Ra fell from 139 nm to 0.756 nm. Root-mean-square roughness Rq fell from 175 nm to 0.975 nm. No crack propagation, localized spallation, or interfacial failure was detected after polishing.
The Raman diamond peak shifted from 1334.7 cm−1 to 1333.5 cm−1. Its full width at half maximum decreased from 7.53 cm−1 to 6.01 cm−1. The authors interpreted these changes as partial stress relaxation and improved near-surface crystalline integrity.
XRD retained diamond (111), (220), and (311) reflections and detected no graphitic carbon or secondary phase. The polished surface therefore retained the diamond structure while reducing topographic relief.
Wear resistance at fixed conditions
Section titled “Wear resistance at fixed conditions”Separate fixed-condition tests ran for 180 min at 0.2 MPa, 500 N, and 1000 r·min−1. Rotating-ring wear came from the median of three random profiles multiplied by the 60 mm track circumference. Each reported error range covered three independent tests.
| Seal-face pair | Rotating-ring wear rate (mm³·N−1·m−1) | Stationary-ring wear-step height (µm) | Main surface observations |
|---|---|---|---|
| SiC-SiC | 8.3 × 10−10 | 64 | Relatively flat faces; rapid adhesive instability |
| Graphite-SiC | 9.5 × 10−10 | 689 | Severe graphite roughening, local spallation, and SiC grooves |
| Graphite-MCD | 0.15 × 10−10 | 63 | Largely intact MCD and uniform shallow graphite grooves |
SiC-SiC showed little material loss because adhesive instability ended the test before sustained wear developed. Graphite-SiC produced deep graphite depressions, local spallation, continuous SiC grooves, and substantial material removal. Graphite-MCD retained a largely intact MCD surface with shallow grooves, small protrusions, and little debris. The graphite face wore uniformly in this pair, with much less height loss than graphite against SiC.
Failure mechanisms at the sliding interface
Section titled “Failure mechanisms at the sliding interface”Surface analyses separated three failure pathways. After water-film breakdown, SiC-SiC asperities contacted directly and formed adhesive junctions. Continued shear produced seizure, torque fluctuations, and early failure.
EDS on Graphite-SiC detected abundant carbon, silicon, and oxygen signals with correlated distributions. TEM identified SiC fragments, silicon oxides, and graphitized carbon in a continuous near-surface layer. The authors propose that brittle SiC fragments acted as hard third bodies and intensified graphite abrasion. The graphite-SiC transfer film reduced shear resistance, but it did not prevent reaction-layer disruption.
The MCD surface after Graphite-MCD sliding remained comparatively smooth, with no substantial debris accumulation. EDS showed low silicon and oxygen levels without clear enrichment. TEM found an ordered diamond lattice and no continuous interfacial layer. XPS showed a dominant sp³ carbon signal and only a weak O 1s signal. These measurements support the proposed synergy of low-shear graphite lubrication, stiff MCD load support, and interfacial stability.
Implications for seal design and test limits
Section titled “Implications for seal design and test limits”Material selection should account for the wear mode that develops after water-film breakdown. SiC-SiC showed low material removal after early adhesive seizure. A soft-hard pair can lower friction but still lose material rapidly if its hard counterface sheds debris. The study supports balancing lubricity, load support, chemical stability, and resistance to third-body abrasion.
Replacing SiC with polished MCD raised limiting PcV from 50.19 to 90.73 MPa·m·s−1 within the Graphite pair comparison. The measured result applies to deionized water, 0.2 MPa chamber pressure, and the stated loads and speeds. The paper reports no seawater, corrosive-fluid, or temperature variation, so those service conditions remain unresolved. The 180 min wear test also does not reproduce a full service life.
For seal development, coating adhesion, post-polish phase integrity, and debris control deserve verification on the selected geometry. Graphite-MCD provides a measured basis for screening polished diamond counterfaces in high-pressure water-lubricated seals. The design supports further evaluation for deep-sea seals, seawater pumps, and high-speed water-lubricated bearings. Those applications still require validation under their actual fluids and duty cycles.
Materials and support
Section titled “Materials and support”Mechanical seals in this study require a polished microcrystalline diamond hard counterface carried by SiC. 6CCVD’s diamond materials catalog lists MPCVD-grown polycrystalline diamond, plates and wafers up to 125 mm, and polishing below 5 nm for inch-size polycrystalline material. The study used a 4.9 µm HFCVD coating, so engineering support should confirm whether the MPCVD catalog route fits the required coating and seal geometry.
View Original Abstract
Water-lubricated seal interfaces are particularly vulnerable to fluid-film instability, which accelerates interfacial friction and wear and compromises the reliability and service life of underwater equipment. Here, a novel graphite-microcrystalline diamond (Graphite-MCD) soft-hard seal-face tribopair was proposed and systematically benchmarked against SiC-SiC and Graphite-SiC tribopairs under high-pressure water lubrication. Their sealing capacity, friction and wear behavior, and interfacial evolution were evaluated. The polished Graphite-MCD pair achieved the highest limiting PcV (90.73 MPa·m·s−1) and the lowest wear rate, markedly outperforming the SiC-SiC and Graphite-SiC pairs. Multiscale interfacial characterization showed that the SiC-SiC pair failed primarily through adhesive instability after breakdown of the water film. Although the Graphite-SiC pair reduced interfacial shear resistance, sliding generated a continuous triboreaction layer accompanied by third-body wear. In contrast, the Graphite-MCD pair preserved interfacial structural integrity and suppressed the formation of an unstable continuous reaction layer and abundant abrasive debris. This behavior enabled a synergistic combination of a low-shear lubricating interface and a stiff load-bearing counterface. These results demonstrate that improving the limiting PcV of mechanical seals cannot be achieved by reducing the friction coefficient alone; rather, it requires simultaneous control of lubricity and interfacial stability. The findings provide a mechanistic basis for designing soft-hard tribopairs for high-pressure, water-lubricated mechanical seals.