The Ra target for a tungsten carbide seal face is 0.01 um. For carbon-graphite, it’s 0.1 um — ten times coarser. That single difference means the grit, the abrasive type, and the number of progression stages change completely depending on which face you’re holding.
Before you grab a tube of paste, check what face material you’re working on. That one decision drives everything else: which abrasive compound, which grit number, and how many stages to finish.
Surface Finish Targets by Face Material
Tungsten carbide (TC) needs a mirror: 0.01 um Ra, lapped to one light band of flatness. Silicon carbide (SiC) is slightly more forgiving at 0.04 um Ra, also one to two light bands. Ceramic (alumina) targets 0.07 um Ra with one to two light bands. Carbon-graphite is the most tolerant at 0.1 um Ra and two to three light bands.
These aren’t interchangeable. A carbon face lapped to TC specifications wastes time and can actually seal worse — the surface becomes too smooth for the carbon to develop its normal running film. A TC face left at carbon-grade finish will leak immediately.
For high-pressure applications above 40 bar, tighten every material to one light band maximum regardless of the standard range. Each helium light band represents roughly 0.3 microns of deviation, so the difference between one and three bands is the difference between a seal that holds and one that weeps from day one.
Hard Face Materials (TC, SiC, Ceramic)
Hard faces demand tighter finishes and more lapping stages. The general benchmark for hard seal face material combinations falls between 1 and 3 microinches (0.025 to 0.076 um). TC sits at the tight end of that range, ceramic at the loose end.
The most common mistake I see is treating all hard faces the same. SiC at 0.04 um needs roughly half the finishing effort of TC at 0.01 um. Lumping them together means either overworking SiC or underfinishing TC.
Soft Face Materials (Carbon-Graphite, Ni-Resist)
Carbon-graphite faces relax after lapping. You can lap a carbon face to under one light band, put it on the shelf, and measure three light bands the next day. This is normal — carbon is not dimensionally rigid like a hard face. Once the spring loads it against a flat hard face, it conforms back.
Stop chasing perfect flatness on carbon. Lap to two light bands, verify it, and move on. The real quality indicator for carbon is surface finish uniformity, not absolute flatness numbers.
How to Match Abrasive Type to Face Material
The abrasive in your paste must be harder than the face you’re lapping. That sounds obvious, but the practical choices narrow fast once you factor in cost and embedding risk.
Diamond Compound
Diamond and boron carbide are the only abrasives hard enough to efficiently cut silicon carbide and tungsten carbide. For these materials, diamond paste is not optional — conventional abrasives simply won’t produce the required finish in any reasonable time. Use polycrystalline diamond compound in the 1 to 6 micron range for finishing hard faces.
Diamond paste also works on carbon and ceramic, but it’s expensive overkill. A tube of 3-micron diamond compound costs many times what an equivalent silicon carbide compound costs. On a carbon face that only needs 0.1 um Ra, aluminum oxide gets you there at a fraction of the price.
Silicon Carbide Compound
SiC compound is the workhorse for ceramic (alumina) faces and for rough-stage work on softer materials. Clover-brand SiC compounds range from 80 grit (267 um) down to 1200 grit (3 um), covering everything from heavy stock removal to near-finish work.
One critical rule: never use silicon carbide abrasive on a silicon carbide seal face. The abrasive particles embed in the face material because they share the same hardness. Those embedded particles then act as fixed cutting points during operation, scoring the mating face and causing progressive grooving damage. If you’re working on SiC faces, use diamond — boron carbide is the only conventional alternative.
Aluminum Oxide Compound
Aluminum oxide (AlOx) compound handles carbon-graphite, Ni-resist, and stainless steel faces. It’s the gentlest of the common abrasives and the most affordable. Lapmaster AlOx compounds range from F360 (22.8 um) to F1200 (3.0 um).
For carbon faces, aluminum oxide is the right choice. Diamond is unnecessary, and SiC is too aggressive for final finishing stages.
Grit Selection by Seal Face Material
| Face Material | Abrasive Type | Starting Grit | Finishing Grit | Target Ra |
|---|---|---|---|---|
| Tungsten Carbide | Diamond | 6 um | 1 um | 0.01 um |
| Silicon Carbide | Diamond | 6 um | 1 um | 0.04 um |
| Ceramic (Alumina) | SiC or Diamond | F600 (9 um) | F1200 (3 um) | 0.07 um |
| Carbon-Graphite | AlOx | F600 (9 um) | F1200 (3 um) | 0.1 um |
| Ni-Resist | AlOx | F400 (17 um) | F1200 (3 um) | 0.1 um |
Starting grits assume a face in serviceable condition — minor wear, no deep scoring. If a face has visible grooves or chips, begin one or two stages coarser for stock removal before entering the finishing sequence. If the damage won’t come out within two extra stages, that face needs replacement, not more lapping.
For diamond grits on TC and SiC, the progression typically runs: 6 um, 3 um, 1 um. Three stages. For conventional abrasives on ceramic and carbon, a two-stage sequence often suffices: F600 for shaping, F1200 for finishing. Ni-resist may need a third stage starting at F400 if the surface has corrosion pitting.
Verify the Ra after each stage with a profilometer or, at minimum, check flatness with a monochromatic light and an optical flat. If the finish isn’t progressing, the problem is usually contamination or insufficient time at the current stage — not a need to skip ahead. A complete lapping procedure with technique and motion patterns will make each stage more efficient, but the grit decision comes first.
Grit Progression and Stage Cleaning
Jumping from a coarse grit directly to a finishing grit leaves scratches that the fine abrasive cannot remove. Every intermediate stage must fully erase the scratch pattern left by the previous one before moving finer. Rush this, and you end up with a face that checks out on flatness but has deep subsurface scratches that open up under operating pressure.
Between Every Stage: Clean Everything
The single most damaging shortcut in lapping is skipping the cleaning step between grit changes. One particle of 6-micron diamond left on a face during a 1-micron finishing pass creates a groove that matches the failure pattern inspectors call “abrasive damage” — concentric or spiral scoring across the face.
Clean the face, the lapping plate, your hands, and any fixtures with solvent between every grit change. Compressed air alone is not enough. Wipe with a lint-free cloth dampened with isopropyl alcohol or acetone, then inspect under magnification before applying the next finer grit.
When to Stop
Here’s a trick that saves time: stop when you’ve hit the target Ra range, not when you’ve hit the finest grit you own. Joseph Boylan at Morgan Advanced Materials documented that overly smooth hard-face pairs can wring together, generating excessive torque and heat. Hard-on-hard combinations like SiC-against-SiC or TC-against-TC have a practical lower limit on surface finish. Polishing past the specification doesn’t improve sealing — it creates a suction effect between the faces that fights separation and accelerates wear on startup.
Target the specified range. A TC face at 0.01 um Ra is correct. The same face polished to 0.005 um is a problem waiting to happen.
Common Grit Selection Mistakes
Using Diamond on Everything
Diamond paste works on every seal face material. That doesn’t mean it should be used on every one. Lapping a carbon-graphite face with diamond paste is like cutting softwood with a carbide-tipped blade — it works, but you’re spending far more on consumables with no improvement in outcome. Match the abrasive to the material: diamond for hard faces, aluminum oxide or silicon carbide for soft ones.
Skipping Grit Stages
Jumping from a 6-micron diamond straight to 1-micron diamond seems like it saves a step. What actually happens is the 1-micron paste spends most of its working time trying to remove the 6-micron scratch pattern instead of refining the surface. The finishing stage takes much longer than it should, and the result is often a surface with residual deep scratches hiding under an apparently smooth finish. Those hidden scratches become leak paths under pressure.
Ignoring Grit Contamination Between Stages
I’ve pulled seal faces with grooving damage that looked like the seal ran in slurry service — but the pump was handling clean water. Every time, the root cause traced back to grit contamination during lapping. A coarse particle carried over from the previous stage cuts channels into the finished surface. Those channels trap fluid during operation and create the classic spiral grooving pattern that diagnostic references associate with abrasive particle damage.
If you skip this step, you’ll be back in a week pulling a seal with grooves that shouldn’t be there.
The Grit Decision in Practice
The abrasive type and grit number are the two decisions that determine whether a lapping job produces a sealing surface or a scrap part. Every other variable — pressure, speed, time, plate condition — matters, but they’re secondary to getting the right paste on the right face.
Carbon faces relax after lapping and recover flatness under spring load — so chase finish quality, not flatness numbers. Hard faces hold their shape but punish you for overpolishing or contamination. And no amount of technique fixes the wrong abrasive: SiC compound on a SiC face will embed, diamond on carbon is wasted money, and skipping grit stages just hides damage under a surface that looks acceptable until the seal pressurizes.
Check the material, pick the abrasive, follow the progression, clean between every stage. That sequence prevents more post-lapping seal failures than any other single practice change.
