A gap between a printed wall and the infill usually means one of two things: the infill path did not reach far enough into the shell, or the two paths touched but did not fuse into one structure. First rule out broader under-extrusion. Then inspect wall-infill overlap, thin-feature path planning, speed and temperature, and material shrinkage in that order.
Do not begin by raising global flow. That can hide a narrow handoff problem while making walls, dimensions, corners, and top surfaces worse. The useful diagnosis comes from checking whether the gap is local to the wall-infill boundary or appears throughout the print.
Check these six things first
- Confirm the defect after the part cools. Look for a real open channel where infill stops short of the inner wall, not a shadow or ridge visible through a continuous shell.
- Check the whole print for weak extrusion. Thin walls, sparse infill, open top lines, clicking, or an inconsistent purge make feed stability the first branch.
- Inspect the sliced handoff. Compare the infill path with the inside perimeter and verify that the current profile gives them a deliberate overlap.
- Check model thickness and line planning. Narrow ribs and awkward wall widths can leave no clean landing zone for infill.
- Compare slow and fast regions. A path that reaches the wall but peels away can point to melt-rate demand, low fusion temperature, or aggressive cooling.
- Run one controlled coupon. Change only the branch supported by the evidence, then repeat the exact file.
Match the wall-to-infill gap to the likely cause
| What you see | Likely first branch | Best first check |
|---|---|---|
| A consistent channel follows the inner wall while walls and infill otherwise look full | Insufficient wall-infill overlap | Inspect the boundary in layer preview and review the profile's overlap control |
| Gaps appear with thin walls, sparse infill, open roofs, or an uneven purge | Under-extrusion or feed restriction | Stabilize the spool path, hotend output, nozzle, and material before overlap tuning |
| Only narrow ribs, thin boxes, or one model fail | Geometry and line-planning compromise | Compare model thickness with generated wall and gap-fill paths |
| Paths touch in preview but separate more on fast or long sections | Weak fusion at the handoff | Run one modest speed, heat, or cooling test from a known-good profile |
| The gap grows near lifted corners or on large high-shrink parts | Warping or contraction stress | Fix the visible curl, draft, or temperature instability before adding overlap |
| The shell is continuous but infill appears as a shadow or repeating ridge through it | Not a wall-infill gap | Inspect the cross-section and diagnose infill show-through instead |
Check 1: confirm that this is an open boundary gap
Cut or inspect a failed test only when it is safe and appropriate. A true wall-to-infill gap is an open channel between the inner perimeter and the internal structure. The infill terminates short, barely touches, or separates after printing. It is different from infill lines visible as shadows through a thin wall, raised bands caused by excessive overlap, a seam defect, or a rough roof supported by weak infill.
If the outside wall is continuous but the internal pattern telegraphs through it, use the infill show-through guide. If the visible failure is mainly on the top face, use the rough top-surface diagnosis. Keeping those symptoms separate prevents the overlap setting from becoming a catch-all fix.
Fix: photograph the cooled defect, note its layer and location, and inspect the same area in preview. Continue only when the boundary itself is open or weak.
Check 2: rule out under-extrusion before changing overlap
A mild feed problem can make infill lines land short even when the slicer generated the correct path. Look beyond the boundary: are walls thin, infill sparse, top lines open, or line width inconsistent? Does the extruder click? Does the spool tug, the filament path bind, or a steady purge vary in width? Those clues mean the wall gap is one result of a broader delivery problem.
Follow the under-extrusion check order before editing the boundary. Spool drag, a restricted path, an overloaded hotend, a partial clog, filament condition, and an incorrect material profile can all reduce output. Do not compensate for unstable delivery with a large overlap value.
Fix: establish a repeatable purge and full lines throughout the print, then rerun the unchanged coupon. If both the general extrusion and boundary recover, keep the correction in the feed or material branch.
Check 3: inspect wall-infill overlap in layer preview
When walls and infill are both full but a clean channel follows their boundary, overlap moves to the top of the list. Slicers may express this as a percentage, distance, or related wall-infill bonding control. The exact label and behavior vary, so use the current slicer documentation instead of copying a number from an unrelated nozzle, line width, printer, or profile.
Zoom in on several layers. Confirm that infill deliberately enters the inner perimeter enough to form a bond. Check whether the gap exists in the generated path or appears only in the physical print. Also verify that the nozzle size and line-width assumptions match the installed hardware and intended profile.
Fix: if preview shows the path stopping short, increase the relevant overlap in one modest step and re-slice the same coupon. Stop when the joint closes reliably. Too much overlap can create ridges, wall show-through, heavy dimensions, or an over-extrusion-like surface even when global flow is correct.
Check 4: examine thin features and path planning
A profile can work on a broad box and fail on a narrow rib because the model does not provide a clean combination of perimeter, gap-fill, and infill widths. Curved shells, tapers, wall thicknesses close to awkward multiples of the chosen line width, and tiny internal islands can force short or fragmented paths. That is a geometry-specific failure, not proof that the entire printer needs more flow.
Compare a simple coupon with the failing model. In preview, watch how walls, gap fill, and infill change through the problem area. If only one feature fails, test a deliberate wall thickness, another wall count, a suitable line-width plan, or a small geometry revision. Do not scale the whole model unless every dimension is wrong and measurement supports that decision.
Fix: give the slicer a stable landing zone for the internal path. Confirm that the change solves the narrow feature without making the external dimensions, fit, or wall strength worse.
Check 5: separate path reach from weak fusion
Sometimes preview shows adequate overlap and the printed lines touch, yet the joint peels apart. That points to fusion rather than reach. Infill may be running much faster than the walls, the hotend may be near its output limit, the material may be effectively too cool at that speed, or aggressive cooling may firm the boundary before the next path bonds.
This becomes more likely when slower regions close cleanly, fast regions separate, or the part also shows weak layer bonding. Start from a known-good material profile and the manufacturer's current guidance. Test only one modest change: reduce the relevant speed, restore thermal headroom, or adjust cooling in a controlled way.
Fix: keep the smallest process change that creates a sound joint without causing sagging, stringing, weak cooling, or lost dimensional control. A line merely touching the wall in preview does not prove it will fuse under the chosen process conditions.
Check 6: look for warping and contraction stress
If the gap grows toward lifted corners, long walls, or cold-facing regions, the shell may be moving away from the internal structure as the part contracts. Increasing overlap can make the first coupon look better while leaving the underlying thermal stress untouched.
Use the warping diagnosis when corners lift, walls curl, or the failure follows drafts and large cross-sections. Check build-surface preparation, first-layer stability, enclosure or room conditions, material guidance, and geometry before treating the boundary as a simple slicer setting.
Fix: stabilize the part first, return overlap to the known baseline, and repeat the coupon. If the boundary closes once the shell stops moving, record the result as a shrinkage or warping correction.
Run this controlled wall-infill test
- Save the failing project. Record nozzle, layer height, wall count, line widths, overlap control, infill, speed, temperature, cooling, and material.
- Slice a small repeatable coupon. Use enough wall and infill area to expose their boundary quickly.
- Print the unchanged baseline. Let it cool, photograph the cross-section or top view, and match it to the symptom table.
- Inspect the same layer in preview. Decide whether the generated path stops short or the physical paths fail to fuse.
- Choose one branch. Feed stability, overlap, geometry, fusion, or shrinkage.
- Change one variable. Do not raise flow, overlap, temperature, and wall count together.
- Repeat the exact coupon. Compare the gap, external walls, dimensions, top closure, and joint strength.
- Confirm on the real part. Keep only a change that survives one representative full print.
Use limits: when a closed wall-infill gap still is not a passing result
A visually closed boundary proves only that the infill reached and fused to the inner wall on that print. It does not prove the real part kept its dimensions, carried its intended load, survived repeated cycles, or remained sound after heat, moisture, chemicals, or weather exposure.
- Loaded bracket, fixture, handle, enclosure, or machine part: print the critical feature in its final orientation, inspect the boundary after cooling, and run a representative proof load and cycle test. Reject a setting that closes the gap by creating thin outside walls, distorted holes, weak layers, or a brittle wall-infill junction.
- Bearing seat, fastener boss, sliding fit, press fit, or mating shell: more overlap can move material toward the perimeter. Measure the actual bore, wall, and mating dimensions after the part cools; a cleaner cross-section is not a pass if the assembly binds, loosens, or loses clearance.
- Air, water, vacuum, fluid, or powder boundary: no visible channel does not establish leak tightness. Use a suitable material and process, inspect the complete path, and perform a bounded leak test appropriate to the real pressure and consequence. Do not use this coupon as pressure-vessel qualification.
- Heat, outdoor, chemical, vibration, or fatigue service: repeat the real feature under the expected exposure and duty cycle. A room-temperature coupon does not qualify creep, UV stability, chemical resistance, impact strength, or long-term fusion.
- Food-contact, medical, lifting, vehicle-retention, electrical-safety, fire-safety, or other consequence-heavy use: slicer overlap and a visual inspection are not a compliance or safety case. Use an appropriate validated material, process, inspection plan, safety factor, and qualified alternative where failure could injure someone or damage critical equipment.
Release sequence: repeat the controlled coupon, then print the actual critical feature in its intended orientation. Let it cool, inspect the wall-infill junction, measure the controlling dimensions, verify the real mate, and run the representative load, cycle, leak, or exposure check. Repeat the job before treating the profile as stable.
Stop changing overlap if the gap returns with clicking, a variable purge, thin walls, open top lines, or weak layers elsewhere. That is a feed, hotend, material, or melt-rate problem. Also stop escalating overlap when only one narrow feature fails; revise that feature or its line plan instead of distorting every wall in the model.
Evidence boundary: Prusa's current infill guidance notes that perimeter count is the main contributor to model strength while infill also contributes compression resistance. Its extrusion-multiplier guidance treats flow as a material-specific calibration. Closing one handoff by forcing global flow or overlap therefore does not qualify the complete part.
Common fixes that waste time
- Raising global flow first: it can crowd walls and corners while leaving a path-planning problem unresolved.
- Using extreme overlap: excessive overlap can telegraph infill through the shell, build ridges, and distort dimensions.
- Changing the infill pattern without checking the boundary: pattern can affect local paths, but it does not repair an unstable feed system or a deliberately short handoff.
- Adding more walls blindly: another perimeter may help one geometry, but it can also remove infill from a narrow feature without explaining the original failure.
- Copying a percentage from another profile: overlap behavior depends on the slicer, nozzle, line widths, and path plan.
- Testing on different models: geometry changes can mask whether the profile correction actually worked.
What to do next
If a modest overlap correction closed a clean boundary gap, save it with the exact nozzle and profile context and verify dimensions. If the whole print was starved, keep working in the extrusion branch until walls, infill, and roofs are stable. If paths touch but peel apart, confirm fusion and layer strength before increasing overlap further. If the failure follows one narrow feature, correct the geometry or line plan at that feature.
Use the 3D print quality troubleshooting hub when the pattern is not actually at the wall-infill boundary. The correct next page may be rough top skin, infill show-through, weak layers, over-extrusion, or warping; those defects need different evidence and different fixes.
Frequently asked questions
Do gaps between walls and infill always mean under-extrusion?
No. Under-extrusion is the first branch only when the rest of the print also looks short on material. A consistent boundary channel with otherwise full paths more often points to overlap or line planning.
Should I increase flow to close the gap?
Not until you prove the printer is under-delivering everywhere. Global flow changes affect walls, dimensions, corners, top skin, and infill, while a wall-infill gap may need only a boundary or geometry correction.
How do I know whether overlap is too low?
Preview shows the infill ending too far from the inner perimeter, and a modest change to the slicer's documented overlap control closes the same coupon without creating ridges, show-through, or heavy dimensions.
Can speed cause infill to separate from walls?
Yes. If the path reaches the wall but the physical joint is weak, high melt-rate demand, low effective temperature, or aggressive cooling can prevent sound fusion. Compare slow and fast regions before changing several process settings.
Can thin model walls cause the problem?
Yes. Narrow or awkward features can leave no clean room for whole perimeter and infill paths. Preview the exact feature and test a deliberate thickness or line-plan change instead of retuning the whole machine.