Cemented carbide and technical ceramics are both hard, but they fail differently. Carbide polishing fights cobalt smearing and chemical attack; ceramic polishing fights subsurface damage and edge chipping. Treating them as one application is the usual reason a grade that works in one fails in the other.
Cemented Carbide: The Binder Is The Problem
Carbide is tungsten carbide grains held in a cobalt binder. The tungsten carbide is harder than most abrasives can scratch efficiently; the cobalt is soft and reactive. Polishing removes the carbide grains and smears the cobalt across the surface, producing a characteristic haze that is not a scratch and does not respond to a finer grade.
Two consequences follow:
- Carrier chemistry matters. Cobalt is attacked by acidic carriers, which accelerates binder removal and produces pitting. Neutral to mildly alkaline carriers are safer.
- Removal has to be balanced. A grade that removes carbide grains quickly can leave binder standing proud. The step after has to level both phases.
Typical carbide work includes seal rings, wire drawing die cores, wear parts and indexable inserts. Size ladders commonly run from 20 to 30 micrometres for roughing down to 1 to 2 for finishing.
Technical Ceramics: Damage Goes Below The Surface
Alumina, zirconia, silicon carbide, silicon nitride and boron carbide are brittle. A particle that is too coarse does not just scratch the surface; it drives a crack network into the material. The part can look acceptable and fail later under load or during coating.
So ceramic work is more sensitive to the PSD tail than to the nominal size. A W7 powder with a controlled D97 usually outperforms a W5 powder with an uncontrolled tail, because the tail creates the damage.
Applications include ceramic bearings, seal faces, ballistic plates, structural components and thin-section parts. For balls specifically, see ceramic and steel ball polishing.
Comparing The Two
| Cemented carbide | Technical ceramics | |
|---|---|---|
| Failure mode | Cobalt smearing, pitting | Subsurface damage, chipping |
| Key risk | Chemical attack on binder | Oversized particles in the tail |
| Carrier | Neutral to mildly alkaline | Wide choice, check dispersion |
| Size focus | Balance removal of both phases | Tight D97 and D90 |
| Grade direction | HDY mono for control | HDY, HDYP for finish |
| Typical finish size | 1 – 2 um | 0.5 – 2 um |
Why Poly Helps On Ceramics And Not On Carbide
Polycrystalline powder breaks into smaller cutting points, which limits the depth of any single cut. On ceramics that directly reduces the depth of the crack network. On carbide the limiting factor is chemical, so poly helps less and costs more.
A common compromise is mono through the ladder and poly only on the last step for ceramics, and mono throughout for carbide.
Edge Quality Is Its Own Specification
Both material classes appear in parts with functional edges: inserts, seal faces, nozzles, cutting blades. A polished face with a chipped edge is scrap, and edge chipping is driven by the same oversized particles that cause subsurface damage.
Two practical consequences:
- Chamfer before the fine steps. A sharp edge cannot survive fine polishing. Break the edge early with a controlled chamfer, then polish.
- Inspect the edge separately. Surface Ra on the face says nothing about the edge. Use the same magnification every time and compare against a reference part.
Where The Cost Actually Sits
| Cost driver | Carbide | Ceramics |
|---|---|---|
| Powder | Moderate, mono is sufficient | Higher at the finish, poly adds cost |
| Carrier chemistry | Critical, protects the binder | Secondary |
| Rejects | Pitting and haze | Subsurface damage found late |
| Rework | Repolish | Often scrap |
Ceramic rejects are usually more expensive than the powder difference, which is why tight PSD at the finish is worth paying for. On carbide the same money is better spent on carrier control and process stability.
Common Mistakes
- Chasing carbide haze with finer powder. The haze is smeared cobalt. Going finer polishes the smear instead of removing it.
- Ignoring the PSD tail on ceramics. D50 looks perfect and the part still chips, because D97 is out of control.
- Using an acidic carrier on carbide. It attacks the binder and produces pitting that is mistaken for abrasive damage.
How To Verify
For carbide, inspect under oblique light and check for pitting as well as scratching, then confirm the carrier pH. For ceramics, measure strength or run a dye-penetrant check after polishing, not just Ra, because subsurface damage can coexist with a good surface number.
Next Step
Applications: carbide and ceramic polishing, resin and vitrified bond wheels.
