All articles
Buyer GuideTroubleshootingFDMMaterialsEquipment

Why 3D Printed Parts Shatter on Impact

3D Prototyping Hub·
Why 3D Printed Parts Shatter on Impact

It flew all season. Then one landing came in a little flat, the arm hit the grass, and it did not bend or crack — it came apart into four pieces and a spray of chips. Nothing had changed. The same part, the same spool, the same slicer profile that had worked for months. 3D printed parts shatter on impact because impact asks for a property that nothing on the spool advertises, and the number people choose filament on is not that property.

This post contains affiliate links. If you purchase through these links, 3D Prototyping Hub may earn a small commission at no extra cost to you.

This post also contains Amazon affiliate links. As an Amazon Associate we earn from qualifying purchases.

Read the Break Before You Buy Anything

Pick the pieces up and look at the fracture face. It tells you which of five different problems you have, and they have different fixes.

What the break looks like What it is Where to start
Many pieces, glassy faces, chips and dust Brittle fracture — wrong polymer for a sudden load Material
One clean split exactly along a layer line Interlayer bond, not material strength Orientation and print settings
Whitened, stretched, bent region before the tear Ductile failure — the material did its job and lost Section and geometry
Crack starting at a sharp internal corner or hole Stress concentration Fillets, not thickness
Fine crack visible before it finally let go Accumulated damage from earlier hits Inspect and replace on a schedule

The useful tell is the surface. A glassy, faceted fracture that reflects light means the part broke without deforming — it stored the energy and released it as new cracks. A dull, whitened, slightly stretched region means the material absorbed energy by deforming first, which is what you want even when the part still failed. The first is a material problem. The second is a geometry problem, and buying different filament will not fix it.

Strength Is the Wrong Number, and the Spool Only Prints That One

Tensile strength is what a bar withstands when you pull it slowly. PLA does well on it — commonly 50-60 MPa, above ABS on the same test. That figure is what gets printed on the box, and it is why "PLA is stronger than ABS" is both technically true and practically useless here.

Impact asks a different question: how much energy can the part absorb before something breaks? That is roughly the area under the whole stress-strain curve, and it is dominated by how far the material stretches rather than how hard it pulls. PLA stretches around 2-5% and lets go. ABS, ASA and nylon stretch considerably further, absorbing energy the whole way.

Two more effects stack on top, and both work against you:

  • Strain rate. Polymers get stiffer and more brittle the faster you load them. A material that bends on a bench press can shatter under a hammer, so a slow bend test tells you very little about a drop.
  • Temperature. The same part is meaningfully more brittle at 0°C than at 25°C. A PLA bracket that survives summer handling can explode on the first cold morning, which is why winter is when people decide their printer got worse. It did not; the part did.

Does It Actually Need Drying?

Tell moisture from the faults that imitate it. Then dry it without wrecking it.

The file downloads on this page as soon as you submit. No waiting on an email.

Orientation Decides More Than Material

An extruded part is not one material. It is a stack of bonded layers, and the bond between them is weaker than the extrusion itself — typically well under half the strength in the layer direction, and worse on a cold, fast, draughty print.

That gives the crack a ready-made path. A drop that puts tension across the layer plane finds it instantly. The same part rotated so the layers run along the load survives a far harder hit, with no change of filament and no extra cost.

Three things to do with that, in order:

  1. Work out which way the hit arrives, then orient so the impact load runs along the layers rather than peeling them apart. For a drone arm that means laying it flat, not standing it on end for a prettier surface.
  2. Print hotter and slower for the parts that get hit. Raising nozzle temperature toward the top of the material's range and reducing part cooling on the walls genuinely improves layer bonding. It does not make PLA tough — it just stops you losing what the material already has. The full picture is in why 3D prints snap along layer lines.
  3. Add walls before you add infill. Energy is absorbed in the shell. Four to six perimeters does far more for a dropped part than raising infill, and it does it without the extra mass that makes the next drop harder.

Every Hole and Corner Is a Crack Starter

Impact energy concentrates at geometry, and a printed part is full of concentrators that a machined one would not have.

A sharp internal corner multiplies local stress several times over. So does a bolt hole near an edge, a sudden change in thickness, a support scar left proud, and the small notch where a bridge or an overhang left a step. The crack starts at whichever of these sits closest to where the energy arrived.

The fixes are cheap and they are all CAD, not filament:

  • Fillet every internal corner. Even a 2mm radius transforms the stress distribution at a corner that was square. This is the single highest-value change on the page.
  • Keep holes away from edges. Leave at least one hole diameter of material around a hole, and more where the part is thin.
  • Taper thickness changes rather than stepping them. A thick boss meeting a thin plate is a crack waiting for an excuse.
  • Remove the support scars on surfaces that see load. A nub left on a tension face is a notch you printed deliberately.

Fatigue cracks grow from the same places under repeated smaller loads, and the two problems travel together on anything that gets knocked regularly — that side of it is why 3D printed parts crack under vibration.

The Material Ladder, With What Each Rung Costs

Material Behaviour under a sudden hit What it costs you
PLA Shatters. High strength, almost no elongation Nothing — it is the cheapest and easiest to print
PETG Yields and tears rather than exploding Prints on any PLA machine; stays notch-sensitive, so fillets matter
ABS / ASA Genuinely tough; ASA adds UV stability Needs an enclosure and a draught-free room, or it cracks mid-print
Nylon (PA) Dents, bends, survives — the practical answer Absorbs moisture from room air in hours; must be printed dry
Polycarbonate Absorbs the most energy of the common filaments ~290-310°C and a hot chamber; still notch-sensitive in thick sections
TPU Effectively unbreakable at these energies Not rigid — right for the bumper, wrong for the structure
Carbon-filled grades Worse than the same polymer unfilled Buy for stiffness and heat, never for impact

Two rows deserve the extra sentence, because they are where people go wrong in opposite directions.

Carbon fibre is not the upgrade it looks like. Chopped fibre raises stiffness and dimensional stability, which is exactly right for a bracket that must not flex. But the fibres interrupt the energy-absorbing polymer chains, so a filled grade almost always has lower elongation and lower impact toughness than the unfilled version. For a part that gets hit, plain nylon beats carbon-filled nylon.

Nylon only works dry. Nylon filament is the toughest common material here and the one most often bought and then wasted: it pulls moisture out of ordinary room air within hours, and wet nylon prints with steam bubbles, poor layer bonds and strings, producing a part that fails at exactly the interface you bought the material to strengthen. Printing out of a heated dryer box is part of the purchase, not an accessory to it. The current range is compared in our nylon filament guide.

The honest middle of the ladder is PETG, which demands nothing of your printer, and ASA, which is a real step up in toughness and holds up outdoors, at the price of needing a still, warm chamber. If the part is structural and the answer keeps coming back as nylon or polycarbonate, that is a hardware purchase rather than a settings problem: QIDI's enclosed, high-temperature printer line lists chamber and hotend specifications per model, which is what to check the material against.

Design the Part to Lose Something Cheap

The strongest version of a dropped part is usually not the strong one. It is the one with a sacrificial piece in front of it.

A TPU bumper, foot, skid or standoff absorbs the energy over a longer distance before it reaches anything rigid, and doubling the time an impact takes roughly halves the peak force. It is the same principle as a crumple zone, it costs a few grams, and a part designed to be replaced after a bad landing is a better outcome than a part designed never to fail. Printing it is covered in our TPU guide; the broader design question of which parts should flex is in why 3D printed snap-fit clips break.

What Does Not Work

  • 100% infill. More mass means more kinetic energy on the next drop, plus internal stress from over-packing. Walls and fillets beat solid material.
  • Gluing it back together. The joint is weaker than the original and the crack usually restarts beside it. It gets you through a weekend, not a season.
  • Annealing PLA. It raises stiffness and heat resistance, often reduces impact toughness further, and shrinks the part unevenly. It is the wrong lever for this problem.
  • Thicker everywhere. Thickness without fillets just moves the same stress concentration somewhere else, and thick sections in notch-sensitive materials can fail at lower energies than thin ones.

Fix It in This Order

  1. Read the fracture. Glassy means material. Clean along a layer line means orientation. Whitened and stretched means geometry.
  2. Fillet the corners and move the holes. Free, and it fixes more dropped parts than any spool change.
  3. Reorient so the hit runs along the layers, and raise walls to four to six.
  4. Step up one rung of material — PLA to PETG costs nothing in hardware; PETG to ASA costs an enclosure.
  5. Add a sacrificial TPU part where the impact actually lands.
  6. Move to dry nylon or polycarbonate when the part is structural and gets hit as a working condition, and budget for the dryer and the chamber that come with them.

When the Part Should Not Be Printed at All

Three cases where FDM is the wrong process rather than the wrong material: the part is a safety item, it is crashed as a matter of routine, or you need a batch of them in a polymer your machine cannot run.

Sintered nylon from a service has no layer plane to split along, which removes the failure mode this whole article is about, and it is substantially tougher than the FDM version of the same polymer. Cast urethane covers low volumes in genuinely rubbery materials — the comparison is in urethane casting vs 3D printing, and the wider question of which parts survive real service is in functional and end-use 3D printed parts. You can browse providers by location and process and get two or three quotes before committing to another spool.

Most shattered parts are the first three steps on that list. The ones that still break after all three are telling you they were never a PLA job.

Hero photograph by Tom Claes on Unsplash.

Does It Actually Need Drying?

Tell moisture from the faults that imitate it. Then dry it without wrecking it.

The file downloads on this page as soon as you submit. No waiting on an email.

Recommended Resources

Disclosure: Some links below may be affiliate links. We only recommend services we have personally evaluated or that are used by providers in our directory. Clicking earns us a small commission at no cost to you.

Frequently Asked Questions

Ready to find a 3D printing service provider?

Browse 2,000+ verified providers across the United States and submit your quote request free.

Browse Providers