How Accurate Is Dental 3D Printing, Really?
For decades, the stone model was the unquestioned foundation of dental lab work. Today, resin printers have taken over that role almost entirely — but only because they've closed a tolerance gap that plaster could never close. Here's what the accuracy numbers actually show, why printed models replaced stone in most labs, and how to keep a printer dialed in tightly enough to earn that trust.
What the tolerance actually looks like on the bench
Accuracy in a dental model has two components that matter: trueness (how close the model is to the actual digital design) and precision (how repeatable that result is from one print to the next). Stone models struggle with both. Impression material can distort on removal, gypsum expands slightly as it sets, and hand-trimming introduces operator variability that changes from technician to technician.
A well-maintained DLP or LCD dental printer removes most of those variables from the equation. The geometry comes directly from an intraoral or lab scan, and the printer's job is simply to reproduce that file, layer by layer, without adding its own error.
From plaster to pixels: why the stone model lost its job
Stone didn't lose its place in the lab because it was inaccurate on its own terms — gypsum has been a reliable material for over a century. It lost its place because the workflow around it accumulated error at almost every step, and each of those steps has since been replaced by something more direct.
- 01 Physical impression taken, subject to tray flex and material shrinkage
- 02 Impression shipped, risking distortion in transit
- 03 Gypsum poured by hand, with mix ratio affecting expansion
- 04 Model trimmed manually, introducing technician variability
- 05 Model stored — vulnerable to chipping, humidity, and wear with reuse
- 01 Intraoral or desktop scan captures geometry directly, no impression material
- 02 Digital file transmitted instantly, with zero transit distortion
- 03 Design software segments dies and articulates the model automatically
- 04 Printer reproduces the file layer by layer at a fixed, repeatable resolution
- 05 Model is durable, reprintable on demand, and never needs re-trimming
This is the real reason printed models displaced stone in most labs — not a single dramatic leap in resolution, but the removal of four or five separate points where error used to creep in. A printer that is out of calibration, however, reintroduces exactly that kind of error in a new form, which is why calibration discipline matters as much as the technology itself.
Keeping a printer accurate enough to trust with a margin
A printer fresh out of the box is usually at its most accurate. Everything after that — resin changes, vat wear, optical drift, ambient temperature swings — pulls it slowly away from spec. These are the checks that keep it close enough to genuinely replace a stone model rather than just resemble one.
Re-level the build plate and re-zero Z
An unlevel plate is the single most common cause of dimensional error across an arch — it shows up as a model that's accurate on one side and subtly warped on the other. Re-level after every resin tank swap, every plate removal, and any time the printer is moved.
Weekly, or after any plate removalRun a dimensional test print, not just a visual one
Print a known reference geometry — a calibration grid or a model with published dimensions — and measure it with digital calipers or a scanner. A model that looks clean can still be a few percent oversized or undersized in XY if the light engine has drifted.
Monthly, or after any firmware or resin changeRe-verify exposure times for each resin batch
Resin viscosity and pigment density shift slightly batch to batch, which changes how the material cures at a given exposure time. Run a light-cure test (an exposure matrix print) whenever a new resin lot goes into service, rather than assuming last quarter's settings still hold.
Every new resin lotInspect and clean the FEP film and optical window
A scratched or cloudy FEP film scatters light unevenly across the build area, which shows up as inconsistent cure depth from one region of the plate to another. Replace the film at the first sign of clouding rather than pushing it to failure.
Biweekly inspectionHold the room, not just the printer, to spec
Resin viscosity is temperature-sensitive, and a lab that swings from a cool morning to a warm afternoon will see exposure and cure results drift with it. A stable ambient temperature does more for repeatable accuracy than most single hardware upgrades.
Continuous — monitor dailyStandardize the wash and post-cure cycle
Under-washing leaves uncured resin on the surface; over-curing can introduce shrinkage and warping, particularly on thin die sections. Fix a wash time, solvent, and post-cure duration as a lab standard, and don't let it vary print to print.
Every print — keep it fixedPrecision upstream earns precision downstream
A printed model is only as good as the calibration discipline behind it — and the restorations built on top of it are only as good as the model. That's the same principle CrystalCeram® is built around at the ceramic stage: consistent, cadmium-free powders and stains formulated so that the accuracy your lab worked to capture digitally doesn't get lost in the layering and firing that follows.
Explore the CrystalCeram® system at cgiporcelain.com.
