Common Mistakes

7 Costly Mistakes in Advanced Technical Ceramics Production (and the Fixes)

The seven mistakes that quietly destroy yield and margin in technical ceramics production, each with the symptom, the root cause, and a numeric fix you can apply this week.

Technical ceramics punish small errors because yield losses compound across five or six process steps. A line running 97 percent yield at pressing, 95 percent at green machining, 96 percent through sintering, and 92 percent through grinding and inspection nets only about 81 percent overall, so a single wrong assumption anywhere in the chain can erase the margin on an entire order. Most troubleshooting calls trace back to the same handful of mistakes: shrinkage math applied wrong, scrap counted in the wrong place, kiln loading quoted at a fill rate the plant never achieves, and grinding time borrowed from metalworking intuition. Each mistake below lists the symptom, the root cause, and a fix with a number attached.

Symptom: fired parts run consistently 4 to 6 percent undersize even though the die was cut to the powder supplier's stated shrinkage. Root cause: linear and volumetric shrinkage got mixed up. An alumina body that shrinks 17 percent linearly loses roughly 43 percent of its volume, and plugging the volumetric figure into a linear allowance produces dies that are far too small, while the reverse error makes parts oversize and adds grinding stock. Fix: fire five test bars pressed at your own green density, measure linear shrinkage on each axis, and size tooling with the Sintering Shrinkage Allowance calculator. Requalify whenever green density drifts more than 0.02 g/cm3 from the qualified value.

Symptom: diameters land in tolerance but thickness misses by 0.3 to 1.0 percent, always in the same direction. Root cause: uniaxially pressed parts shrink anisotropically because density is higher near the punch faces than at mid height, so shrinkage along the pressing axis can differ from radial shrinkage by 0.5 percent or more on parts with aspect ratios above 1.5. A single shrinkage factor cannot satisfy both axes. Fix: measure and apply separate axial and radial allowances, and if the spread exceeds 1 percent, reduce fill depth, move to isostatic pressing, or adjust binder loading to flatten the density gradient. Verify the axial to radial split with fired test coupons every shift.

Symptom: the fired inspection report shows 92 percent yield, but powder purchases say the process consumes 30 percent more material than shipped product contains. Root cause: green scrap never gets counted. Cracked or chipped green parts go back into the reclaim drum without a transaction, so the loss hides in material variance instead of yield. Fix: log pressing losses with the Powder Press Yield calculator, handling and green machining losses with the Green Body Scrap Rate calculator, and post fire losses with the Fired Dimensional Yield calculator, then multiply the three. Value reclaimed powder honestly with the Ceramic Scrap Recovery Value calculator; zirconia reclaim is often worth only 40 to 60 percent of virgin.

Symptom: firing cost per part comes in at double the quoted figure on small lots. Root cause: the quote assumed the kiln fires at 80 to 85 percent of setter capacity, but the scheduler released the job as a partial load at 45 percent fill. A batch kiln burns nearly the same gas or kilowatt hours whether it holds 400 parts or 900, so cost per part scales almost inversely with fill. Fix: set a minimum release quantity per firing, combine compatible jobs on shared setters, and check every planned load against the Kiln Utilization Cost calculator before release. Make loads under 60 percent fill require a supervisor signature.

Symptom: the diamond grinding cell becomes the plant bottleneck and lead times slip two weeks. Root cause: grinding time was estimated from metal cutting intuition. Fully dense alumina and zirconia grind at specific removal rates around 1 to 5 mm3 per mm of wheel width per second, roughly a tenth of what steel allows, and every extra 0.1 mm of grind stock left by sloppy shrinkage control adds proportional wheel time. Fix: quote from measured removal rates using the Diamond Grinding Time calculator and attack stock, not feed. Cutting grind allowance from 0.5 mm to 0.25 mm per side roughly halves cycle time and doubles cell capacity without buying a machine.

Symptom: parts pass the lab quench test and still crack in service within weeks. Root cause: the test load does not represent the duty. A single water quench from 300 C says little about a burner nozzle cycling between 800 C and 150 C hundreds of times, and quench severity itself shifts with bath temperature and part thickness. Fix: define the test delta T from the worst measured service transient plus a 25 percent margin, and size the quench batch and cycle count with the Thermal Shock Test Load calculator. Run at least 10 cycles and check retained flexural strength; a strength loss above 20 percent after cycling predicts field failures far better than visible cracking does.

Symptom: finished parts wait three to five days at the CMM while downstream operations starve. Root cause: 100 percent dimensional inspection was carried over from qualification into production, even on characteristics running Cpk above 1.67 with no nonconformance in months. A second common cause is measuring parts still warm from grinding coolant; a 10 C temperature offset shifts a 50 mm alumina dimension by about 3 to 4 microns, enough to fail a tight tolerance falsely. Fix: model queue load with the Ceramic Inspection Bottleneck Risk calculator, move stable dimensions to sampling plans, and soak parts 30 minutes at 20 C before final measurement of anything tighter than 10 microns.

Published 2026-07-02.