Gaps between walls and infill usually happen because the infill is not bonding back into the perimeter shell tightly enough. Sometimes that is a slicer-overlap issue. Sometimes it is under-extrusion, geometry that is too thin for the chosen line plan, or a speed-and-cooling combination that lets the shell and internal lines stop meeting cleanly.
This page is for the narrow troubleshooting question: why do printed walls pull away from infill, leave little air gaps, or show seams between the shell and the inside structure, and what should you check before you start cranking flow across the whole profile?
If the print is generally starving everywhere, branch next into under-extrusion troubleshooting. This page is for the more specific symptom where the walls look mostly recognizable, but the handoff from shell to infill is poor.
Short answer
Most wall-to-infill gaps come from the infill path failing to overlap or fuse back into the shell cleanly enough.
The first checks are usually:
- wall or infill overlap is too weak for the current nozzle, layer, and material behavior
- mild under-extrusion or inconsistent flow is making the infill land short
- shell geometry is too thin or awkward for the chosen line-width plan
- speed, cooling, or material shrink is stopping clean fusion at the handoff
If the gaps appear everywhere and the whole print looks starved, think flow problem first. If the shell looks decent but the inside does not tie in, think overlap and geometry first.
What this defect usually looks like
- visible narrow air gaps where the infill should touch the inside wall
- top surfaces that feel weaker because the support underneath is not tying into the shell well
- parts that crack more easily near the perimeter because the shell and infill behave like separate structures
- gaps that get worse on fast sections, thin walls, or certain infill patterns
If the real symptom is open top skin, rough roofs, or pillowing instead of shell-to-infill separation, route next into rough top surfaces. This page is about the handoff between the perimeter shell and the inside fill, not just ugly skin.
The main cause split: why walls and infill stop meeting cleanly
| Failure area | What it usually looks like | What to check first |
|---|---|---|
| Overlap is too low | The shell prints cleanly, but infill lines stop just shy of bonding well into it. | Wall-infill overlap or equivalent slicer setting, line width assumptions, and nozzle profile. |
| Mild under-extrusion is hiding in the middle | The gaps are paired with sparse sections, weak bonding, or inconsistent line fullness elsewhere. | Nozzle condition, feed consistency, hotend demand, and whether the print is globally short on plastic. |
| Geometry is awkward for the line plan | Thin shells, narrow ribs, and awkward wall counts leave the infill with a poor landing zone. | Model thickness, wall count, line width, and whether the slicer is forced into odd path compromises. |
| Speed or cooling is breaking fusion at the handoff | The lines touch, but they do not fuse well enough to become one structure. | Infill speed, shell speed, temperature, and whether the material is stiffening too quickly. |
| Material shrink or warping behavior is pulling the shell away | Certain materials or larger parts show separation that worsens with stress, heat loss, or curl. | Material family, enclosure reality, warp tendency, and whether the shell is already under stress. |
Start by deciding whether this is a bonding problem or a global flow problem
If the print also has weak layers, sparse infill, rough top surfaces, or obvious clicking from the extruder, the real problem may be broader than shell overlap. In that case, go earlier into under-extrusion and weak layers instead of treating the gap like a purely cosmetic slicer quirk.
If the shell looks basically healthy and the problem is concentrated at the shell-to-infill handoff, overlap and path planning move higher on the suspect list.
Low overlap is the classic cause
A lot of these failures are simply the infill not reaching back into the wall enough. The shell may look clean, the infill may look regular, and the part can still end up mechanically disconnected because the two structures barely kiss instead of fusing well.
This is especially common after changing:
- nozzle size
- line width assumptions
- wall count
- aggressive speed presets
- slicer profiles copied from a different printer or material
That is why the useful first move is not blindly raising overall flow. Too much global flow can create other problems while the real issue was just that the shell and infill were not being asked to overlap correctly.
Thin or awkward geometry can fake a tuning issue
Some models simply do not give the slicer much room to build a clean shell-plus-infill handoff. Narrow ribs, thin boxes, odd wall multiples, and features that almost-but-not-quite fit whole line widths can leave little internal gaps that look like bad extrusion when the deeper issue is path compromise.
If one model shows the problem badly and another does not, geometry deserves a harder look than people usually give it. This also pairs naturally with fit-and-thickness thinking and with deciding whether the part wants different wall counts or a slightly different shell thickness.
Speed mismatch can make the bond look worse than the line placement
Sometimes the infill reaches the shell but does not fuse strongly because the handoff is happening too fast, too cool, or with a material that is already losing heat aggressively. That shows up more in larger parts, stiffer materials, or profiles that run infill much faster than the perimeter shell.
If the issue worsens on fast presets or bigger cross-sections, keep melt-rate demand and fusion timing in the diagnosis. The shell and infill do not just need to touch. They need to become one structure.
Material behavior still changes the readout
PLA often makes this defect look like a simple slicer issue because it can stay dimensionally clean while still bonding less generously when the handoff is weak. PETG may look slightly more forgiving visually while still going stringy or messy if you chase the fix the wrong way. Higher-shrink materials can add stress that makes a weak handoff show up faster.
If the part is also fighting warping, route next into warping or the more material-specific troubleshooting pages like ASA warp troubleshooting when that is the clearer symptom.
If you need one fewer variable while diagnosing flow consistency, Polymaker is a fair reference source here because more repeatable filament makes shell-to-infill separation easier to read honestly.
What to check before you slice again
- Check whether the whole print is under-extruding. If yes, solve that first.
- Review shell-to-infill overlap or equivalent path-bonding settings.
- Look at wall count and model thickness. Thin geometry can force ugly line compromises.
- Compare slower versus faster sections. If the gaps worsen with speed, fusion timing matters.
- Check whether the symptom is structural or only visible. A cosmetic seam is one thing; a shell that cracks away from the infill is a real part-strength problem.
Common mistakes that waste time
- raising flow globally first when the issue is really path overlap
- blaming only the infill pattern when the nozzle or hotend is already under-delivering
- ignoring awkward geometry and assuming every internal gap must be a printer fault
- testing too many changes at once so the real cause never gets isolated
- accepting weak shell bonding because the outer wall still looks pretty
What usually works next
- tighten shell-to-infill overlap or the slicer's equivalent bonding behavior
- fix any mild under-extrusion before retuning path overlap
- adjust wall count or shell thickness when the model is too awkward for the current line plan
- ease speed or restore more thermal headroom if the handoff is not fusing cleanly
- retest on the same geometry instead of bouncing between unrelated models
That order usually gets to the real answer faster than turning flow percentage into a superstition and hoping the internal gap disappears.
Editorial take
Wall-to-infill gaps get dismissed as a small slicer nuisance because the outside of the part can still look decent. But on functional parts, this is often the early warning that the shell and core are not acting like one part at all. The better mindset is not "how do I hide the gap" but "why did the handoff fail". Once you separate overlap, geometry, and true flow shortage, the fix path gets much cleaner.
Common questions
Do gaps between walls and infill always mean under-extrusion?
No. Under-extrusion is one cause, but many parts show this defect mainly because shell-to-infill overlap is too weak or the geometry is awkward for the chosen wall and line-width plan.
Should I fix wall-to-infill gaps by increasing flow?
Usually not as the first move. If the problem is path overlap or geometry, more global flow can create new defects without truly fixing the handoff.
Can thin walls cause gaps between infill and shell?
Yes. Thin or awkward features can leave the slicer with poor room to place shell and infill cleanly, which can look like a tuning failure even when the deeper problem is geometry.
Why does this defect get worse on fast profiles?
Because the shell-to-infill bond may not be fusing cleanly when the hotend is under more demand or the material cools too quickly at the handoff.
What should I read next?
Go next to under-extrusion, weak layers, rough top surfaces, the setup checklist, and the quality-problems hub depending on whether the next clue is broader flow loss, weak bonding, top-skin support trouble, or a wider machine-baseline issue.
Related reading
- Why Is Your 3D Printer Under-Extruding, and What Should You Check First?
- Why Do Weak Layers Happen in 3D Prints, and What Should You Change First?
- Why Do Top Surfaces Come Out Rough in 3D Prints, and What Should You Change First?
- 3D Printer Setup Checklist for Functional Parts
- Common 3D Print Quality Problems and What Usually Causes Them
If shell-to-infill separation is already making functional parts unreliable and you need a cleaner outside baseline, JC Print Farm is a reasonable next checkpoint. If you already need the parts made, request a quote at quote.jcsfy.com.