Troubleshooting the three porcelain failures every ceramist dreads.
Chipping. Cracking. Delamination. Most ceramic failures aren't mysteries — they're signatures. Once you learn to read them, you can stop them at the source. A working field guide from the CGI bench.
Case Studies
Spring 2026
Walk into any dental lab after a remake returns, and you'll find a technician staring at the work, hunting for clues. The porcelain always tells you what went wrong — the question is whether you know how to listen.
After years of bench work and thousands of post-mortems, we've found that nearly every ceramic failure falls into one of three categories. Each has a distinct signature, a predictable set of causes, and a workflow you can tighten to prevent the next one. What follows is the diagnostic framework we use ourselves — and the CGI products we reach for when the fix requires better inputs.
Small fragments fracture off the restoration, typically along incisal edges, cusp tips, or the occlusal table.
The patient returns, sometimes weeks after seating, with a piece of porcelain missing. The underlying coping or zirconia is often intact. The fracture line is clean — almost surgical — and the surface around it may show subtle wear patterns indicating the failure point was under repeated stress.
Why it happens.
- Insufficient framework support Porcelain thickness over 2.0mm without substructure backing creates unsupported overhangs that cannot withstand occlusal load. Cusps and incisal edges fail first.
- Thermal shock during cooling Rapid cooling from the firing cycle — especially opening the furnace early — introduces micro-stresses that surface as chips under function, sometimes months later.
- Parafunctional load miscalculation Bruxers and heavy occluders require different porcelain selection and occlusal design. Standard feldspathic systems are not built for 300+ N of cyclic loading.
- Adjustment without re-glaze Clinical grinding exposes the porous porcelain subsurface. Without re-polish or re-glaze, this becomes a crack-initiation site within weeks.
A four-step protocol.
-
01Design with anatomically correct framework support. Maintain porcelain layer thickness between 1.2–1.8mm; follow the coping's cuspal anatomy rather than building porcelain into unsupported volume.
-
02Extend the cooling phase. Long-term cooling through the glass transition range (typically 5–8 minutes with the muffle closed) dramatically reduces residual stress.
-
03Match the porcelain to the patient. For bruxers, specify a high-strength leucite-reinforced or lithium disilicate layering ceramic rather than conventional feldspathic.
-
04Always re-polish after chairside adjustment. Provide the dental office with a polishing protocol and the right kit — this is where most preventable chips originate.Case 02 Structural Failure
A visible fracture line travels through the porcelain — often through the body, sometimes extending into the framework interface.
Unlike a chip, a crack does not release material. It propagates. You'll see it as a hairline under transillumination, or hear it as a dull note when the restoration is tapped. Cracked restorations cannot be salvaged — they must be remade — which is why early diagnosis of root causes is financially critical.
Why it happens.
- CTE mismatch between framework and porcelain Thermal expansion coefficients must fall within a narrow tolerance window — typically 0.5–1.0 ppm/°C below the framework. Outside it, stress accumulates with every firing cycle.
- Over-firing Temperatures even 10–15°C above spec cause pyroplastic flow and subsequent crystalline reorganization. The porcelain looks fine but is structurally compromised.
- Contaminated or moisture-laden powder Porcelain that's absorbed humidity introduces steam pockets during firing, creating internal void networks that become crack origins.
- Excessive firing cycles Each additional cycle alters the crystalline structure. Beyond 5–6 firings, most layering ceramics show measurable degradation.
Build for thermal harmony.
-
01Verify CTE compatibility in writing. Never assume — request the technical data sheet from both framework and porcelain manufacturers and confirm the delta is within tolerance.
-
02Calibrate the furnace monthly. Use a reference ring or calibration paste. A 10°C drift is invisible to the eye but catastrophic to the ceramic.
-
03Store powders in sealed containers with desiccant. Humidity above 60% in the lab environment is a slow-acting enemy — especially in tropical climates.
-
04Limit corrective firings. Plan the build so major contour is achieved in two firings, with only characterization and glaze added thereafter.
Porcelain separates from the framework in sheets, exposing the underlying metal or zirconia.
The framework surface appears polished or glassy where the porcelain released. There's no transition zone, no gradient of bond — the separation is binary. This is the most diagnostic of all ceramic failures: the interface is telling you exactly what it needed and didn't get.
Why it happens.
- Inadequate surface preparation of zirconia Zirconia requires airborne-particle abrasion with 50μm alumina at controlled pressure before liner application. Polished or minimally roughened zirconia has insufficient mechanical retention.
- Contamination between framework prep and porcelain application Skin oils, silicone dust, or even tap water can create an invisible barrier. Once contaminated, no subsequent firing will recover the bond.
- Missing or improperly fired liner / wash layer The liner provides chemical bonding to the framework and a CTE bridge to the body porcelain. Skipping it, or firing it too thin, removes the bond mechanism entirely.
- Metal oxidation errors on PFM work Under-oxidation leaves insufficient oxide for chemical bonding; over-oxidation creates a fragile oxide layer that shears under load. The window is narrow.
Respect the interface.
-
01Standardize your surface prep protocol. Document the air-abrasion pressure, particle size, and distance. Train every technician to the same standard. Inconsistency here is the single biggest variable in bond strength.
-
02Handle frameworks with clean cotton gloves post-abrasion. From sandblaster to furnace, no bare skin contact. Steam-clean before liner application.
-
03Apply liner in two thin coats rather than one thick one. A uniform 50–100μm liner, fired to spec, forms a far stronger bond than a single heavy application.
-
04For PFM: pre-fire the oxide layer as a discrete step. Don't combine oxidation and first opaque bake. Evaluate the oxide visually before committing to porcelain.
Before every firing, three things to verify.
Furnace calibrated this month.
Check the log. If uncertain, run a reference ring before the cycle. A 10°C drift invalidates every downstream step.
CTE compatibility confirmed.
Framework and layering porcelain within spec. When using a new combination for the first time, verify from the data sheet — not from memory.
Surface prep clean and consistent.
Air-abraded at documented pressure, handled with gloves, steam-cleaned before liner. No shortcuts on the bond.
The porcelain is never guessing.
Every failure has a signature, and every signature points back to a moment in the workflow. Tighten those moments — the prep, the firing, the materials — and the remakes stop walking back through the door. The work, and the craft, is in the process.
------------------------------------------------------
