If your print snaps cleanly along a layer boundary — a smooth, flat break that follows the horizontal lines — the bond between layers never fully formed. That is the definition of a layer adhesion problem: the material welded fine within each layer (the X-Y plane) but failed to fuse to the layer beneath it in Z. Fix the bond and the part behaves as one solid piece; leave it weak and the object stays fragile along its weakest axis, splitting, delaminating or crumbling under modest load.
Almost every case of poor layer adhesion traces back to one thing: the interface between two layers didn’t get enough heat energy, for enough time, over enough contact area to let the polymer chains diffuse across the boundary. Everything below — temperature, cooling, flow, speed, moisture — is just a lever on those three variables. Here’s how each one works and exactly what to change.
The mechanism: interlayer bonding is thermal welding
When the nozzle lays down a new bead, that molten plastic has to partially re-melt the surface of the layer below so that polymer chains from both surfaces can intermingle and entangle across the interface. Three conditions govern how strong that weld ends up:
- Temperature — both the incoming bead and the layer below must be hot enough for the chains to move freely.
- Contact time — the interface needs long enough in the hot state for meaningful diffusion to occur.
- Contact area — good physical squish across a wide area maximises the chain entanglement.
This is why Z is always the weak axis. Along X-Y the part is one continuous extrusion path; in Z it’s dozens or hundreds of stacked layers, each boundary a potential fault line. And because the layer below has already started cooling by the time the next one arrives, the bonding window is always closing on you.
How much weaker is the Z axis?
The honest answer is: it depends heavily on material and settings, and the published figures genuinely conflict. Treat any single number with suspicion. The common ranges look like this:
| Metric | Typical figure | Notes |
|---|---|---|
| Z strength vs X-Y | ~50–70% of X-Y | Material- and settings-dependent |
| Z generally weaker by | ~20–50% | Most commonly cited band |
| FDM vs injection moulding (X-Y) | 70–90% of moulded strength | Well-tuned parts |
| FDM vs injection moulding (Z) | 40–75% strength, 10–30% ductility | Ductility hit is the bigger problem |
| Aggressive/outlier estimate | X-Y up to 4–5× stronger than Z | Poorly-tuned or brittle materials |
A representative lab datapoint: one FDM test measured Z-axis yield strength at roughly 55% of the X-axis. The practical takeaway isn’t the exact percentage — it’s that you should orient functional parts so the primary load runs parallel to the layers, and never rely on a tensile load pulling straight across the layer lines.
Cause 1 — Nozzle temperature too low (the usual culprit)
Insufficient nozzle temperature is the single most frequent cause of weak layers. If the extruded plastic isn’t hot enough, it freezes before it can soften and fuse into the layer below. The fix is deliberately boring: raise the nozzle temperature in 5°C increments until layer adhesion improves and your snap test starts breaking rough rather than clean.
There’s a subtle trap here involving flow rate. If pushing more flow makes adhesion worse, the filament may be moving through the hot end too fast to reach the set temperature — the melt is thermally starved. This is measurable: raising the nozzle to 260°C on one material lifted the critical flow rate from 10 mm³/s to 16 mm³/s. So if you print fast and hot-end capacity is the limit, either slow down or run hotter, don’t just crank the flow.
Cause 2 — Too much part cooling
Rapid cooling slams the bonding window shut. The right fan strategy is entirely material-dependent, driven by each polymer’s glass transition temperature (Tg): PLA ~60°C, PETG ~80°C, ABS/ASA ~100–105°C. The higher the Tg, the less it tolerates forced cooling.
| Material | Tg (approx.) | Part cooling starting point |
|---|---|---|
| PLA | ~60°C | 50–70% (not 100% from layer one) |
| PETG | ~80°C | 30–40%, adjust for stringing vs delamination |
| ABS/ASA | ~100–105°C | Fan off; enclosure; brief directional air only for bridges |
| Nylon | — | Fan off for reliable adhesion |
Symptoms of over-cooling read differently per material: PETG and ABS split along layers; thin PLA features go matte and chalky; sharp corners crack as the next pass lands on already-cold plastic. There’s a real trade-off with PLA — strong cooling gives you crisp overhangs and clean bridges but costs Z-strength. If you need strength over surface finish, print PLA about 5°C hotter than usual and drop the fan to widen the bonding window. For high-Tg materials the answer is almost always an enclosure and a still, warm environment rather than any fan tweak — see our notes on PETG behaviour for how touchy that material is.
Cause 3 — Under-extrusion, flow and layer geometry
Less material means less contact area, so a bit more flow generally means a stronger bond. Increase the extrusion multiplier in ~5% steps. But don’t overshoot: both under- and over-fill hurt strength. Once the fill ratio is dialled in it isn’t especially sensitive — one calibrated test settled at a fill ratio of 0.965, which is exactly the kind of narrow, correct window you’re aiming for.
Two mechanical faults masquerade as settings problems:
- Partial clogs and nozzle debris. A dirty or partially blocked nozzle throttles flow inconsistently and produces weak spots. Clean the nozzle regularly.
- Filament diameter variation. Tolerance wider than ±0.05 mm causes uneven extrusion and weak boundaries. Quality brands hold ±0.02–0.03 mm; cheap filament at ±0.1 mm+ can genuinely print weaker. (Manufacturer figures — verify per brand.)
Layer height: the genuinely contested lever
This one splits the room, because two real effects pull in opposite directions:
- Fewer interfaces argument: 0.2 mm layers have fewer boundaries than 0.1 mm, and fewer weak interfaces can mean better perceived Z-strength.
- Per-interface quality argument: several lab tensile studies find thinner layers give higher overall tensile strength, and report that layer height influences tensile strength far more than annealing time or temperature.
Both can be true — it comes down to whether you’re limited by the number of weak interfaces or by squish and contact quality at each interface. For most functional parts, 0.2 mm with good squish and a hot nozzle is a safe default; drop to 0.12–0.16 mm only if you’ve got flow and temperature already nailed. If you want the full settings breakdown, our guide to the settings that actually bond your layers goes deeper.
Cause 4 — Print speed
Speed works exactly like temperature in reverse. Move the head too fast and the fresh bead doesn’t spend enough time in contact while hot to diffuse into the layer below. Lowering print speed buys bonding time — often the cleanest fix when you’re already at the top of your temperature range and don’t want to scorch the material.
Cause 5 — Filament moisture
Hygroscopic filaments pull water out of the air. In the melt zone that water flashes to steam, forming micro-bubbles that disrupt extrusion and blow the interface apart — you’ll hear hissing or popping from the nozzle and see stringing, bubbling and rough surfaces. The strength hit is estimated at 20–50%, and worse: in PETG, ABS, PC and nylon the steam hydrolyses the polymer chains, which is permanent degradation no amount of drying reverses.
| Material | Drying temp | Time | Watch out for |
|---|---|---|---|
| PLA | 45°C | ~6 h | Softens above 55°C, deforms at 64°C — never a normal oven |
| PETG | 65°C | ~6 h | Stay below the material’s HDT |
| ABS | 70°C | ~4 h | — |
| TPU | 55°C | ~8 h | — |
| Nylon | 80°C | 12–16 h | Reabsorbs in ~30 min of humid open air |
These are starting points — defer to the manufacturer’s technical data sheet, and always keep the drying temperature below that specific filament’s heat deflection temperature. HDT varies wildly between “the same” material: Bambu PETG HF sits around 62°C while Prusament PETG Tungsten is about 94°C, so a one-size drying temp will slump one spool and barely touch another. Store filament in airtight containers with desiccant, and print straight from a dry box that holds below 15% RH for anything moisture-sensitive.
Cause 6 — Ambient environment
A cold or drafty room pulls heat out of the part before layers can bond, and also drives warping. Warm, still air helps layers stick. This is where diagnosis gets useful: weakness across every layer boundary points to a systemic cause — nozzle too cold, fan too strong, or a draught. Delamination on a single layer line instead points to a momentary event: a brief temperature dip, a partially-cleared clog, or a layer that sat too long during a long retraction move. Don’t confuse this with mechanical layer shifting, which is a positional fault, not a bonding one.
Annealing: a real but limited last resort
Heat-treating a finished part lets residual stress relax and additional crystallisation occur, improving Z-strength by roughly 10–20%. The catch is warping — thin, unsupported sections distort. A commonly cited procedure: preheat a conventional oven to ~65°C for PLA or ~80°C for PETG, sit the part on a flat heat-resistant surface, bake 1–3 hours by size, then cool slowly with the door ajar. Note that other sources anneal PLA hotter (80–100°C), so test on a sacrificial part first. Annealing is a bonus on top of good printing — it will not rescue a part that delaminated from a cold nozzle or wet filament.
The fix order that works
- Run the snap test — break a sample along the layers vs across them. Clean, smooth break along a boundary = poor adhesion.
- Dry the filament first if you hear hissing or see bubbles/stringing.
- Raise nozzle temperature in 5°C steps.
- Reduce fan for the material; enclose high-Tg filaments.
- Increase flow in ~5% steps; check for partial clogs.
- Lower print speed.
- Reorient the part so load runs along the layers.
If you’re not sure whether you’re looking at delamination, under-extrusion or something else entirely, upload a photo to our Diagnose tool — it reads the failure mode from the image and returns slicer-specific settings, including downloadable .ini patches for PrusaSlicer and OrcaSlicer.
FAQ
Why does my print break easily between layers but not across them?
Because the Z axis is always the weak direction in FDM — layers are stacked welds, typically only 50–70% as strong as the continuous X-Y plane. A clean break along a layer line means those welds never fully formed, usually from a nozzle that’s too cold, too much cooling, or wet filament.
Does printing hotter always improve layer adhesion?
Up to a point, yes — hotter plastic diffuses better across the interface. Raise it in 5°C steps. But if you’re already flowing fast, the melt can be thermally starved, so slowing down may help more than more heat. Go too far and you’ll scorch the material and get stringing.
Can wet filament permanently ruin layer strength?
Yes, for PETG, ABS, PC and nylon. Steam from absorbed moisture hydrolyses the polymer chains, which is irreversible — drying afterwards won’t restore full strength. PLA suffers extrusion problems from moisture but far less permanent chemical damage. Store hygroscopic filament sealed with desiccant.
Should I use thicker or thinner layers for the strongest parts?
It’s genuinely contested. Thicker layers (0.2 mm) mean fewer weak interfaces; thinner layers can give higher measured tensile strength per lab tests. For most functional prints, 0.2 mm with good squish and adequate temperature is the reliable default. Only go thinner once flow and temperature are already dialled in.
Related: First Layer Adhesion Problems on Bambu Lab Printers: The Real Causes and Exact Fixes