Reducing Dental Lab Remakes: A Root-Cause Framework for the Scan-to-Sinter Workflow

Reducing Dental Lab Remakes: A Root-Cause Framework for the Scan-to-Sinter Workflow

Blog Article

A crown that returns for poor fit, open contact, incorrect occlusion, shade mismatch, fracture, or distortion rarely carries one catastrophic error. It usually carries several small ones: a margin that was never clearly captured, a connector sized by habit rather than template, a bur that had quietly worn past its useful life, a drying step shortened to protect turnaround.

Reducing remakes therefore requires more than better final inspection. The laboratory has to identify where each error first became *detectable*, set an acceptance gate at that stage, and connect scanner, milling, material, and furnace records to the same case identifier. The following seven causes provide a practical framework for building that system.

## Start with a consistent remake definition

Before measuring anything, settle the definition. A remake is any restoration that must be reproduced because it cannot meet the documented prescription or the laboratory's acceptance criteria.

Record remakes separately from minor adjustments, elective shade changes, and cases altered by a new clinical prescription. Without that separation, the rate reflects clinical preference as much as laboratory performance, and the team ends up chasing problems that are not theirs to fix.

Use both case-based and unit-based rates. One failed full-arch framework is one case, but it may represent more production time and material risk than several single crowns combined. Add the responsible process stage, defect type, detection point, and estimated internal cost to every record. This shows whether the laboratory is fixing the earliest cause or repeatedly absorbing the same downstream loss.

## The seven root causes at a glance

| # | Root cause | Typical remake outcome | Where it can first be controlled |

| --- | ------------------------------ | ------------------------------------------------------ | ---------------------------------------- |

| 1 | Incomplete scan data | Poor margin, contact, or bite fit | Scan acceptance before CAD |

| 2 | CAD design inconsistency | Open contact, high occlusion, weak connector | Design checklist and peer review |

| 3 | Milling or bur variation | Chipping, poor internal fit, rough margin | Calibration and tool-life control |

| 4 | Wrong material or setup | Fracture risk, shade mismatch, shrinkage error | Material-indication and lot verification |

| 5 | Handling and drying errors | Cracks, contamination, uneven shade | Green-state and pre-furnace inspection |

| 6 | Sintering variation | Warping, discoloration, low translucency | Program, tray, and furnace controls |

| 7 | Weak final QC or communication | Defect reaches clinic or prescription is misunderstood | Release checklist and feedback coding |

## 1–2. Input data and design decisions

A digital file can look complete while hiding an unclear margin, a missing interproximal surface, an incomplete emergence profile, a distorted scan-body area, or an unreliable bite. If design begins without resolving the gap, the CAD technician must guess — and the remake appears later as poor fit or contact, even though the first preventable cause was the input record.

Create a scan acceptance screen before CAD. Check margin visibility, adjacent and opposing dentition, tissue interference, scan stitching, bite alignment, implant components, prescription, material request, and shade information. Return incomplete data before design rather than after milling. For model or impression digitization, apply the same rule to the desktop scanner and the physical source.

Equipment resolution alone cannot correct missing anatomy or an unstable capture protocol. Standard scanning paths, calibration, clean optics, staff training, and a documented rescan request are part of the result.

On the design side, margin selection, cement space, proximal contact, occlusal clearance, minimum thickness, connector dimensions, pontic form, and emergence profile all determine whether a restoration fits and survives. When each designer works from personal preference, the same prescription produces different outcomes across shifts.

Build indication-specific design templates and record approved exceptions. Use a second review for full-arch, long-span, implant, or unusually thin cases — and have the reviewer compare the design against the scan and prescription, not merely inspect the final shape. A design that cannot meet material thickness or connector requirements should be stopped before nesting.

Track which CAD variables change after try-in feedback. Repeated adjustment to the same spacer, contact, or occlusal setting is evidence for a controlled template update, not a reason for technicians to keep making case-by-case corrections.

## 3. Milling calibration, strategy, or burs drift out of control

Chipped margins, rough surfaces, incomplete anatomy, poor internal fit, and green-state cracks can all result from calibration drift, worn or incorrect burs, tool mapping errors, unstable fixtures, inappropriate milling strategies, or contamination between wet and dry workflows. Sintering reveals many of these defects — it does not create all of them.

Establish machine-specific checks for axis or spindle calibration, collet condition, fixture cleanliness, dust extraction, coolant where applicable, tool position, bur diameter, and remaining bur life. Run a standard control part after maintenance, a collision, a software change, or repeated fit failures. Record the machine, strategy, material lot, bur set, and operator with every case.

Each dry or dry/wet machine–material–strategy combination should be qualified separately. Follow a scheduled maintenance plan rather than waiting for visible damage or a cluster of remakes.

## 4. The zirconia does not match the indication or workflow

Zirconia products differ in strength–translucency balance, layer structure, shade system, approved thickness, shrinkage factor, and sintering schedule. Selecting by shade name or price alone creates avoidable risk: a highly translucent option may not suit every long-span or high-load case, and a high-strength grade may not meet every esthetic objective.

The material record should include product line, lot, disc identification, shrinkage data, nesting height for multilayer blocks, indication, and approved furnace program. Do not mix material names or program labels that look similar. When a new lot or product is introduced, qualify representative crowns and bridges before releasing it for high-risk cases.

## 5. Green-state handling, coloring, or drying creates hidden defects

Pre-sintered zirconia is vulnerable to edge damage, connector cracks, contamination, and uneven liquid uptake. Metal particles, dirty air, mixed brushes, excessive handling, or unsuitable support during drying can surface later as discoloration, cracks, or shade inconsistency.

Inspect each milled unit before coloring. Use zirconia-dedicated tools, clean oil-free air, controlled coloring-liquid application, and the drying method specified for the material. Record liquid identity, application method, and drying completion. Never load visibly damaged or damp restorations into the furnace simply to protect turnaround time.

A pre-furnace gate is inexpensive compared with repeating design, milling, coloring, and sintering — and it separates defects created before heating from true furnace or program problems.

## 6. The sintering program and load are not material-specific

Peak temperature alone does not define a zirconia cycle. Heating rate, holds, cooling, restoration geometry, tray construction, beads, covers, spacing, support frames, and load mass all affect the thermal result. An unapproved fast cycle may save time on the schedule while increasing remake risk if the zirconia or restoration type was never validated for it.

Assign every zirconia product a controlled program and loading standard. Those programs should reproduce the approved curve, not replace it with the fastest available settings. For bridges and full-arch frameworks, include support design and tray placement in the work instruction.

Review calibration, thermocouple and heating-element condition, chamber cleanliness, cycle history, and interrupted runs whenever defects affect several units or repeat by tray position.

## 7. Final QC and clinic–lab communication do not close the loop

A restoration may be technically sound and still be remade because the prescription, shade communication, occlusal expectation, implant component, delivery date, or approval responsibility was unclear. Final QC must compare the result with the case requirements, not only check surface appearance.

Use a release checklist covering fit verification method, margin integrity, contacts, occlusion, shade, surface, anatomy, component identity, and required photographs or files. When a case returns, code the defect *and* the earliest stage where it could have been detected. Share structured feedback with the clinic instead of assigning blame without evidence.

Monthly review should identify repeat patterns by client, indication, scanner, designer, machine, material, furnace program, and technician. Small patterns only become actionable when the same case identifiers follow the restoration through the complete workflow.

## Build a 90-day remake reduction plan

**Days 1–30.** Create a remake taxonomy and baseline the rate by case and unit.

**Days 31–60.** Select the two largest root-cause groups and install an earlier acceptance gate at each.

**Days 61–90.** Verify whether the rate, remake cost, turnaround loss, and overtime have changed — without simply shifting defects into another category.

- Track remake rate, first-pass yield, internal adjustment rate, and cost per remake.

- Audit a sample of accepted cases as well as failed ones.

- Assign one owner to each corrective action and a date for verification.

- Update scanner, CAD, milling, material, and furnace work instructions when evidence supports the change.

- Train all shifts before judging whether a new standard works.

Do not set an arbitrary zero-remake promise. The useful target is a sustained reduction in *preventable* causes, supported by stable quality and documented learning from every return.

## Frequently asked questions

**What remake rate should a dental lab target?**

There here is no universal benchmark, because case mix, indication complexity, and prescription quality vary widely. Baseline your own rate by case and unit first, then aim for a sustained reduction in preventable causes rather than an absolute number.

**Should remakes be tracked by case or by unit?**

Both. One failed full-arch framework is one case but may represent far more production time and material risk than several single crowns. Unit-based rates hide that difference; case-based rates hide volume.

**Where should the first acceptance gate be placed?**

At the scan. A file that looks complete can still hide an unclear margin or an unreliable bite, and every downstream stage then builds on that gap. Returning incomplete data before CAD is far cheaper than discovering it after milling.

**How long before a remake-reduction program shows results?**

Most labs see the first measurable movement within 60 to 90 days, provided the two largest root-cause groups are targeted with an earlier gate and the change is applied across all shifts before it is judged.

## About the author

This framework was prepared by the technical team at [Vsmile](https://vsmileglobal.com/), a CAD/CAM dental materials and equipment manufacturer supplying [zirconia blocks](https://vsmileglobal.com/product/Zirconia-Blocks), milling machines, scanners, and sintering furnaces to laboratories in over 120 countries. For a remake-reduction review covering scan files, design settings, machine and bur records, zirconia lots, furnace programs, and defect photographs, [contact the Vsmile technical team](https://vsmileglobal.com/contact-us) with representative cases and your current remake coding.

Report this page