A wrong face material pairing does not just wear faster — it fails through mechanisms that have nothing to do with wear. Nickel binder leaching out of tungsten carbide in chemical service. Free silicon dissolving from reaction-bonded SiC in caustic. Carbon blistering in heavy hydrocarbon. Every one of these failures started with someone specifying materials from a catalog without considering how those materials interact under real operating conditions.
References that catalog face materials individually — carbon, silicon carbide, tungsten carbide, ceramic — leave engineers to figure out pairings on their own. That approach misses the point. A seal face does not operate alone. It operates against another face, in a specific fluid, at a specific temperature, pressure, and speed. The combination determines performance, not the individual material; this page picks up where the per-material property pillar leaves off and walks through the pairing decision.
Hard-vs-Soft and Hard-vs-Hard Pairings
Carbon graphite running against silicon carbide handles the majority of standard pump sealing applications. The carbon face generates a microscopically thin graphite transfer film that manages heat and friction at the interface — but only when absorbable vapor or moisture is present. Without it, the self-lubricating mechanism fails entirely.
The decision between hard-vs-soft (carbon against SiC or TC) and hard-vs-hard (SiC/SiC or TC/TC) is not a preference or a premium upgrade. Three factors determine which configuration you need:
- Media viscosity and lubrication regime. Low-viscosity fluids (water, light solvents) operate in boundary or mixed lubrication. Carbon’s self-lubricating film compensates for the thin fluid film. High-viscosity fluids generate a full hydrodynamic film that supports hard-vs-hard operation — and actually require it, because viscous shearing forces can destroy carbon through blistering.
- Abrasive content. Particulate-laden media erode carbon faces rapidly. Two hard faces (typically SiC/SiC) resist abrasive wear and maintain flatness far longer than any carbon-vs-silicon-carbide combination in slurry or dirty service.
- Chemical compatibility with carbon. Strong oxidizing agents, certain acids, and solvents attack carbon’s binder system. If the media eliminates carbon as a viable face material, hard-vs-hard is the only option regardless of viscosity or cleanliness.
If none of these three factors eliminates carbon, stay with carbon/SiC. Hard-vs-hard pairings generate far more heat at equivalent conditions and have much lower PV limits — they solve specific problems, not general ones.
PV Limits by Material Pairing
PV (pressure times velocity) limits set the thermal envelope each face material combination can dissipate before the fluid film breaks down. The values below sort pairings from most thermally tolerant to least:
| Material Pairing | PV Limit (MPa*m/s) | Relative Ranking |
|---|---|---|
| Carbon / SiC | 147.1 | Highest |
| Carbon / TC | 122.6 | High |
| SiC / SiC | 85.8 | Moderate |
| Carbon / Ceramic | 24.5 | Low |
| Carbon / Cast Iron | 24.5 | Low |
| TC / TC | 24.9 | Low |
Carbon/SiC outperforms every other pairing by a wide margin — nearly 6x the PV capacity of TC/TC. That 6:1 ratio explains why hard-vs-hard is not an upgrade: you are trading thermal capacity for abrasion resistance.
For lubricating fluids (oils, glycols), multiply these values by approximately 1.5x. Non-lubricating services (water, light hydrocarbons) use the base values.
One critical caveat about these numbers: standard PV data derives from 100-hour tests on 3-3/8 inch unbalanced seals at 1,800 rpm, extrapolated to a two-year service life assuming 1/8 inch of wear. That extrapolation starts from roughly 0.00071 inches of measurable wear. In actual long-term service, mechanical seals rarely wear out. Failures come from thermal shock, chemical attack, and installation errors — not from exceeding a PV limit. Use PV values to rank pairings comparatively, not as absolute operating ceilings.
How Material Properties Drive Pairing Selection
Three material properties dictate how a face pairing performs at the seal interface: thermal conductivity, hardness, and surface finish capability.
| Property | Carbon (Resin) | TC | SiC | Ceramic (Al2O3) |
|---|---|---|---|---|
| Thermal conductivity (W/mC) | 6 | 100 | 145 | 35 |
| Hardness (Vickers) | ~100 | 1,500 | 2,400 | 1,800 |
| Surface finish (um Ra) | 0.10 | 0.01 | 0.04 | 0.07 |
| Max temperature (C) | 275 | 400 | 1,650* | 1,000* |
*Material limit only. Actual seal operating temperature limited by elastomers and seal design — typically 260-315 C maximum.
Thermal Conductivity
SiC conducts heat 24x faster than carbon. That single property explains most of SiC’s advantage in demanding services. At the seal interface, heat generation from friction must be removed through the face material. Carbon’s 6 W/mC thermal conductivity becomes the bottleneck in high-speed or high-pressure applications — not because carbon wears out, but because it cannot move heat away from the interface fast enough to maintain a stable fluid film.
TC at 100 W/mC handles heat well, but SiC at 145 W/mC is the clear winner for thermal management. When you see a seal failure with heat checking or thermal distortion on the faces, the first question should be whether the face material’s conductivity matches the application’s heat generation rate.
Hardness and Wear Resistance
SiC at 2,400 Vickers is the hardest commonly used face material — approaching diamond-level hardness. TC at 1,500 Vickers offers less hardness but much greater fracture toughness. That tradeoff matters: SiC survives corrosive media and abrasive conditions that destroy TC, but TC survives mechanical shock and vibration loads that would crack SiC.
For applications dominated by chemical attack or abrasive wear, SiC wins. For applications with water hammer, cavitation-induced vibration, or frequent thermal cycling, TC’s toughness advantage becomes the deciding factor. Hardness alone is a misleading selection criterion.
SiC Variant Selection: RB-SiC vs SSiC
Specifying “silicon carbide” without identifying the variant is like specifying “steel” without identifying the grade. Reaction-bonded SiC (RB-SiC) and sintered SiC (SSiC) have fundamentally different performance characteristics that directly affect seal life.
Reaction-Bonded SiC (RB-SiC)
RB-SiC contains 10-12% free silicon remaining from the manufacturing process. For most water-based, hydrocarbon, and moderate-chemical applications, this free silicon is irrelevant — RB-SiC performs excellently and costs less than SSiC. It is the standard SiC grade for the majority of pump sealing applications.
The problem appears in caustics and strong acids. Free silicon is attacked by these chemicals, weakening the SiC matrix and creating a degraded surface that acts as an abrasive against the opposing face. I’ve pulled RB-SiC faces from caustic soda service that looked sandblasted — the free silicon was completely dissolved, leaving a porous skeleton that destroyed the carbon mating ring within weeks.
Sintered SiC (SSiC)
SSiC contains no free silicon or graphite. It is the hardest, most chemically resistant seal face material available — virtually immune to the chemical attack mechanisms that destroy RB-SiC. The tradeoff: SSiC is more brittle than RB-SiC and requires flat drive pads rather than dowels to minimize stress concentrations. Specifying SSiC for chemical resistance may force a different seal geometry to prevent cracking during assembly or thermal cycling.
For any application involving caustics above 5% concentration, strong mineral acids, or fluids where SiC-vs-TC selection matters, always specify SSiC explicitly. A purchase order that says “SiC faces” leaves the decision to whoever fills the order — and the default is almost always the cheaper RB-SiC.
What the Spec Sheet Doesn’t Tell You About Face Material Performance
Spec sheets give you hardness, thermal conductivity, and chemical resistance charts. They do not tell you how materials actually fail in service. Three failure mechanisms that I’ve seen repeatedly — and that never appear in material selection guides:
Tungsten Carbide Leaching
TC’s achilles heel is its binder. The nickel or cobalt that holds the carbide grains together is vulnerable to chemical attack from ammonia, caustics, and high-pH fluids. The binder dissolves selectively, leaving a porous, weakened surface that accelerates wear on the opposing carbon face. The seal does not fail suddenly — it leaks progressively, and by the time you pull it, both faces are destroyed. Galvanic corrosion between the passivated stainless steel shaft sleeve and the active nickel in the TC face compounds the damage in high-pH services.
The API 682 standard specifies face materials for good reasons. When it calls for SiC in chemical service instead of TC, this leaching mechanism is exactly why.
Carbon Blistering
Carbon graphite is susceptible to blistering in hydrocarbon service with cyclic temperature conditions. Hydrocarbons absorb into the carbon matrix, and when temperature cycles cause expansion, subsurface bubbles form and rupture the face surface. High-viscosity fluids make it worse — the viscous shearing forces cause localized binder failure that accelerates blister formation.
If your application involves heavy hydrocarbons with temperature swings, carbon is not viable regardless of what the chemical compatibility chart says. Switch to hard-vs-hard (SiC/SiC) or consider duplex carbide.
The Duplex Carbide Alternative
Not every application fits cleanly into hard-vs-soft or hard-vs-hard. Duplex carbide (approximately 70% reaction-bonded SiC, 30% graphite) fills the gap for marginally lubricating fluids where carbon would blister but full hard-vs-hard generates excessive heat without adequate lubrication. It balances hardness with lubricity and enables thicker fluid films than a pure SiC face. For applications involving light hydrocarbons, dry-running risk, or slip-stick behavior at startup, duplex carbide deserves evaluation before jumping to hard-vs-hard.
Making the Right Pairing Decision
The face material pairing decision cascades through every other aspect of seal performance — from flush plan selection to expected service life to failure mode. Yet most selection errors I encounter are not exotic. They follow a pattern: someone specified “SiC” without choosing a variant, or selected TC for chemical service without checking binder compatibility, or defaulted to hard-vs-hard as a “better” option without understanding the PV penalty.
Start with the operating conditions, not the material catalog. Identify the fluid, its temperature range, its abrasive content, and any chemical attack vectors. Match those conditions to a pairing using the three-factor framework: viscosity, abrasives, chemistry. If carbon survives all three factors, carbon/SiC is your answer. Only eliminate carbon when the operating conditions demand it — and when they do, make sure you specify which SiC variant or TC grade goes in the seal.
