Ceramic & Technical Ceramics

Technical ceramic manufacturing for solid-state batteries and power electronics carries sintering yields below 50% in many pilot production programs — not because the sintering process itself is poorly understood, but because green body defects that are invisible before the kiln amplify catastrophically at 1,000–1,700 °C. Niobia AI inspects before the thermal budget is spent.

From Green Body to Sintered Part

Technical ceramic manufacturing begins with green body forming: tape casting (for thin sheets of LLZO solid electrolyte or AlN substrates) or die pressing (for thicker structural parts). Both processes embed density gradients, inclusion clusters, and micro-cracks into the green body that are not visible to the naked eye but are detectable by vision systems calibrated for the specific ceramic and forming method.

Binder burnout at 400–600 °C removes the organic binder that held the green body together, leaving a porous pre-sintered structure with approximately 40–50% open porosity. Sintering at 1,000–1,700 °C (temperature depends on the ceramic: Al₂O₃ at 1,600–1,700 °C, AlN at 1,700–1,900 °C, LLZO at 1,100–1,200 °C) densifies the part to 95–99.5% theoretical density. Cooling rate during and after sintering determines whether residual thermal stresses crack the part.

Defect Taxonomy: Green Body to Sintered Part

Green body defects divide into three classes: geometric (warpage from non-uniform tape-cast thickness, cracks from die ejection), compositional (binder distribution gradients that create density variation), and inclusion-based (hard agglomerates or foreign particles that act as stress concentrators). All three classes propagate to the sintered part, amplified by the anisotropic shrinkage that occurs during sintering (typically 15–25% linear shrinkage, non-uniform if green body density varies).

Post-sinter defects include surface cracks from thermal shock during cooling, delamination in multilayer structures where adjacent layers have different shrinkage rates, and porosity inhomogeneity that reduces ionic conductivity in solid electrolytes. For LLZO solid electrolytes, published data shows that porosity above 3–5% reduces ionic conductivity below 0.1 mS/cm — more than an order of magnitude below the 0.1–1 mS/cm target that makes solid-state cells competitive with liquid electrolyte designs.

Where Most Ceramic Programs Get This Wrong

Most ceramic scrap decisions happen after sintering, when the thermal budget has already been spent. The inspection logic is backwards. At 8–24 hours per sintering cycle with kiln loads of 50–200 parts, a green body defect rate of 20% means that 20% of the kiln's thermal capacity is occupied by parts that will be rejected after sintering — at a cost of energy, kiln time, and handling that dwarfs the cost of the green body inspection that would have caught them.

The second common failure is the absence of a correlation between green body density measurements and sintered porosity. That correlation is chemistry- and process-specific (it varies with the tape-cast film, binder system, and sintering profile) and must be built from production data. Without it, there is no basis for a green body accept/reject decision that accounts for how the specific defect type will behave in sintering.

What AI Process Intelligence Changes

Green body inspection before sintering — using vision systems calibrated for the specific ceramic, forming method, and defect morphology — provides the intervention point that costs the least. A green body rejected before sintering costs the green body forming time and material. The same defect caught after sintering costs the sintering energy, kiln time, and the handling of a part that must now be scrapped at higher temperature and with more brittle failure modes.

For solid-state battery electrolytes, the porosity-to-conductivity correlation built from 3–6 months of production data enables a non-destructive performance prediction from post-sinter dimensional measurement and surface inspection — the first step toward replacing destructive EIS measurements on 100% of parts with a model-based prediction that samples 5–10% for verification.

Technical ceramic manufacturing for advanced energy applications is one of the highest-leverage AI inspection opportunities available today, precisely because the yield is so low and the cost of the defect is so high. Solid-state battery programs running 40–50% sintering yield have more to gain from process intelligence than almost any other manufacturing process in the clean energy supply chain.