Why Do 3D Prints Fail?

Prints fail. Not often on a well-set-up machine with a sensible file, but often enough that it is built into how the work is planned and priced. The failures are not mysterious — there is a short list, and each item has a physical cause.

Worth understanding even if you never touch a printer, because most of these are decided by the file and the material, not the machine.

Warping

What it looks like: corners lift off the bed. The bottom of the part curls. On a bad one, the whole thing peels off mid-print and gets dragged around.

Why: plastic contracts as it cools. The first layer is stuck to the bed and cannot move; the layers above it are free to. That difference builds internal stress until it wins against the bed adhesion, and the corners — where stress concentrates — go first.

What controls it:

• Material, more than anything. PLA barely warps. PETG a little. ABS and ASA are the fight, shrinking 0.6-1.5% — see do 3D printed parts shrink.
• Geometry. Large flat areas and sharp corners are the worst case. Small or tall-and-chunky is easy.
• Ambient temperature. A draught across a hot ABS part is enough to lift a corner. Enclosures exist for this reason alone.
• Design. Rounded corners on the footprint warp far less than square ones. A free win.

First-layer adhesion

What it looks like: the part comes loose, or the first layer is spaghetti.

Why: the first layer is the foundation and it is deliberately squashed into the bed. If the nozzle is a fraction too high, the beads sit on top instead of keying in and nothing sticks. Too low and the nozzle ploughs.

What controls it: bed levelling, first-layer height, bed temperature, and a genuinely clean bed. Fingerprint grease is a surprisingly common cause. This is a machine discipline more than a design one, but a part with a tiny footprint gives adhesion nothing to work with — which is an orientation decision, per print orientation and strength.

Layer separation

What it looks like: a clean horizontal crack partway up. The part comes apart along a layer line, sometimes on the printer, sometimes weeks later.

Why: the layers never welded properly. Each new bead has to reheat the one below enough to fuse. Too cool, too fast, too much cooling fan, or wet filament and you get a joint rather than a weld.

What controls it: temperature, speed, fan, and dry filament. It is also the underlying reason FDM is anisotropic in the first place — the layer weld is always the weak plane, and a poor one is dramatically weaker. See how strong are 3D printed parts.

Stringing and blobbing

What it looks like: fine cobwebs between features, little zits on the surface.

Why: molten plastic oozes from the nozzle when it moves without printing. Retraction pulls the filament back to stop it; too little and it strings.

What controls it: retraction settings, temperature, and — again — dry filament. Wet PETG or nylon strings no matter what the settings say, because the water is boiling inside the nozzle.

Stringing is cosmetic, usually cleans off, and is worse on some materials than others. It is not a structural problem.

Clogs and under-extrusion

What it looks like: gaps, thin layers, a part that suddenly stops being printed.

Why: something blocked the nozzle, or the extruder stopped gripping. Abrasive filaments wear a soft nozzle out — one of the real costs of carbon fibre, per nylon and carbon fibre parts.

Knocked over, or supports failed

What it looks like: a tall part on the floor of the printer, or an overhang that turned into a droop.

Why: either the part was too tall for its footprint, or the supports were too sparse to hold what was landing on them. Both are prevented at the setup stage, and both are why supports and overhangs is worth reading.

Why this matters to you

Two reasons.

It explains pricing. A part with a large flat ABS footprint carries more risk than a small PETG one, and risk is real cost — a failed print is hours of machine time and a length of plastic, spent twice. Material, size, geometry and supports all move a quote, as set out in how pricing works.

It explains our questions. When we ask where a part lives, what it holds and which faces matter, it is because the answers change the material and the orientation, and those change the failure odds. A part chosen well rarely fails.

Where FDM is the wrong process

Some things fail no matter how the print goes:

• Very large flat parts in high-shrink material. The physics is against you. Split it, or change material.
• Very fine, thin, tall detail. Below what a 0.4 mm nozzle can resolve, no settings help. Resin does that job, and we do not offer it — we would rather say so than take the order.
• Parts needing precision beyond about ±0.1 mm. Machining. See tolerances and fit.
• Safety-critical parts. A print that looks perfect can still hide a weak layer weld. Not a risk worth taking with brakes, lifting gear or structure.

Upload a file and we will flag what worries us before it costs you anything · get an estimate