Direct answer: when a large SLA or MSLA print separates at nearly the same model height on repeated attempts, inspect the sliced layers around the first split before adding exposure or stronger supports everywhere. A sudden increase in cross-sectional area, merging features, a weak support transition, or a suction cup can concentrate release load at one repeatable layer. If the defect stays in the same tank or screen location after the model is moved or reoriented, shift the diagnosis toward tank film, debris, optics, platform stability, or machine condition instead.
"Same height" is useful only after you define what stayed the same: the same model layer, the same XY location over the vat, or roughly the same elapsed time. Those patterns lead to different tests. Formlabs lists model orientation, layout, supports, long pauses, tank age, a loose platform, optical contamination, and tank debris among SLA delamination causes. Its orientation guide also documents warpage at a sudden cross-section increase and defects from progressively expanding areas. Those are manufacturer-scoped clues, not proof that every resin printer uses the same motion, film, or exposure system.
First classify what actually repeats
| Repeat pattern | Leading branch | Best next test |
|---|---|---|
| The split follows the same feature or sliced layer when the model moves on the plate. | Geometry, orientation, supports, or a cup. | Step through the slice before and after the split; compare cross-section, merging islands, support engagement, and enclosed cavities. |
| Different models fail over the same screen or tank area, or the defect stays put when the model moves. | Tank, film, debris, optics, or local machine condition. | Stop printing and follow the exact printer maker's tank, film, optical-surface, and failed-print inspection procedure. |
| The defect returns after a similar pause or elapsed runtime but at a different model height. | Pause, temperature, resin state, or process interruption. | Compare job logs and timing; remove avoidable pauses and verify the exact resin profile and operating range. |
| Only a hollow or cup-shaped region splits, bows, or ruptures. | Trapped-volume suction. | Use the hollow-resin suction-cup diagnostic; open, vent, fill, or reorient the cavity before unrelated tuning. |
| The print detached from the build plate rather than peeling between cured layers. | Non-adherence, not delamination. | Use the printer maker's plate-adherence path; do not treat a missing base as an interlayer split. |
Why a large cross-section can create a repeatable split
In a bottom-up resin process, each cured layer must separate from the tank interface. As the exposed area grows, the job can demand more from the part, supports, release system, and platform. The visible separation may appear exactly where a narrow supported stem becomes a broad body, several branches merge, a box wall closes, or a large face begins. That makes the sliced transition more useful than the final object's overall height.
Formlabs documents "H-bar shrinkage" when a layer has a sudden increase in cross-sectional area and recommends reorientation to reduce the change between layers. It also warns that progressively expanding symmetric areas can create suction forces and print defects. This does not establish a universal safe area, angle, lift speed, or force. A 45-degree example in one official Formlabs geometry is an example for that geometry, not a setting to copy blindly into another printer.
Compare at least 10 layers before and after the first separation. Record when support contacts begin carrying the main body, when isolated islands merge, how rapidly occupied area changes, and whether a cavity becomes closed. If the break starts just after a broad feature appears, redesign or reorient to spread that transition across more layers, regenerate supports, and recheck the entire slice. Do not merely raise exposure until an overloaded geometry happens to survive.
Use this diagnosis order
- Preserve the evidence. Keep the failed part, mark its plate orientation, measure the first separated plane, save the sliced project and exact profile, and note resin identity, resin age, temperature, tank identity, film condition, pauses, warnings, and any debris. A photograph without this context cannot distinguish a model event from a machine-location event.
- Inspect the sliced transition. Look for unsupported red regions or the equivalent warning in the current slicer, abrupt area growth, merging features, long unsupported spans, thin necks, support contacts that arrive too late, and cups. Formlabs specifically directs users to inspect support coverage and avoid unresolved suction cups in its delamination workflow.
- Run one orientation test, not five setting changes. Move or rotate the model so the suspect feature begins at another model height and occupies another tank area. Regenerate supports. If the split follows the feature, geometry and support become more likely. If it stays in one machine area, stop repeating full prints and inspect the machine path.
- Inspect the tank and film using maker instructions. Formlabs says damage or debris in or on a tank can block light locally and tells users to inspect for debris, scratches, wear, dirt, dust, cured resin, or fingerprints. It also warns against metal tools for tank inspection. Other printers have different films and service rules, so use the exact manual rather than borrowing a tool or cleaning method from another system.
- Check optical surfaces and platform stability. Contamination can weaken or diffuse exposure in one location; a loose or misaligned platform can disturb layers. Follow the manufacturer's cleaning and alignment procedure. Do not improvise solvent, abrasive wiping, disassembly, or platform adjustment on an optical stack.
- Then verify the exact resin profile and operating state. Confirm printer model, resin product and batch, supported wavelength/process, current profile, layer height, temperature range, mixing, and any maker-specific rest or lift behavior. Change one bounded variable at a time. There is no evidence-backed universal exposure increase that fixes all repeated delamination.
Support transition or undercure?
Both can leave a horizontal-looking separation, but they predict different evidence. A support transition usually follows a feature: the part broadens, several bodies merge, a heavy region begins, or contacts are sparse on the load side. Reorienting or redistributing supports changes the failure layer relative to the model. A light-path or profile problem is more likely when broad regions across the same machine area are weak, cured fragments appear in the vat, multiple unrelated models fail locally, or the exact resin/profile combination is unsupported.
Do not diagnose underexposure from a single clean split and respond with an arbitrary percentage. Overexposure can damage detail and dimensional intent without solving orientation, cup, film, debris, or platform problems. Start from the resin and printer maker's current profile, then use their calibration method. When dimensional acceptance matters, route the repaired job through the resin tolerance and inspection gate rather than treating a successful-looking reprint as qualification.
When stronger supports help, and when they hide the problem
Supports help when the model lacks adequate early contact, free-hanging sections appear, or the broad region is asking a few touchpoints to carry the entire release event. Formlabs' advanced-support record explains that each model has individual needs and exposes touchpoint size, density, placement, pillar geometry, and raft controls in PreForm. That supports a deliberate support review; it does not justify maximum density everywhere.
More supports can waste resin, scar surfaces, complicate removal, and make cleaning harder. They also do not fix a sealed cup, contaminated tank, weak light path, wrong resin profile, or loose platform. Change support placement around the measured transition, preserve critical surfaces, and compare one representative reprint. If the redesigned slice still creates a sudden broad face, change orientation or geometry before building a forest of supports.
Do not confuse this with the existing same-height filament diagnosis
The same-height 3D print failure guide owns filament-printer branches such as toolpath flow demand, spool drag, heat creep, nozzle collision, cables, and Z-axis binding. This page owns cured resin layers peeling apart in a vat process. The shared question - model layer, machine position, or elapsed time - is a diagnostic method, not proof that the mechanisms are interchangeable.
Likewise, the hollow resin blowout page owns a cavity-specific pressure differential. Use it when the slice forms a cup or the part has a ragged rupture, inward-bowed wall, splash, or trapped cavity. Use this page when broad solid or non-cupping geometry repeatedly separates between cured layers and you need to distinguish a sliced transition from a machine-local defect.
Safe failed-print stop before the next test
Delamination can leave cured fragments in liquid resin. Stop, protect skin and eyes as the resin safety data and printer maker require, and follow the exact failed-print cleanup and tank-inspection procedure before the next exposure. Do not reach into a vat, scrape film with metal, pour contaminated resin back into a bottle, or send liquid resin or solvent down a drain. Maintain the ventilation and wet-work controls described in the resin-printing ventilation decision.
Washing and curing do not repair a print that separated during printing. The wash-and-cure station guide covers the downstream cell, but failed fragments, contaminated solvent, trapped liquid, and damaged film must be handled under current manufacturer and local waste rules first.
A bounded reprint gate
- The first split plane is measured and matched to the saved slice.
- The job is classified as model-layer, machine-location, or elapsed-time repeatable.
- The revised orientation has regenerated supports and no unresolved islands or cups.
- The tank, film, optical path, and platform pass the exact maker's inspection procedure.
- The exact resin profile, temperature range, and layer choice are documented.
- One representative part or critical section passes before a full plate or large-volume repeat.
- The accepted part is inspected at the former split plane and at functional interfaces.
A large printer does not remove this proof burden. The Photon Mono M7 Max capacity review separates real large-part fit from resin inventory and failure exposure, while the printer review directory helps with a machine decision only after the failure mechanism is bounded. For an occasional or consequential part, compare the custom quote route or production-service path with another full-vat trial. A provider still needs the resin, geometry, inspection state, and use requirements; a quote is not engineering or safety approval.
Sources checked
- Formlabs: Delamination (SLA) - visual symptoms, documented causes, support and suction review, tank, optics, platform, pause, cleanup, and escalation boundaries.
- Formlabs: Model orientation best practices for SLA printing - abrupt cross-section changes, H-bar shrinkage, merging features, expanding symmetric areas, cups, orientation, and support tradeoffs.
- Formlabs: Editing advanced settings for supports in PreForm - model-specific support controls and the boundary against one universal support recipe.
Evidence boundary: GoodPrints did not inspect a reader's printer, measure peel force, test these settings, or establish a universal exposure, angle, support density, or safe cross-sectional area. The sequence above interprets current official Formlabs troubleshooting records into a bounded diagnostic. The exact printer, slicer, resin, safety data, and manufacturer instructions control.