Living hinges and flexures

A living hinge is a strip of plastic so thin that it bends instead of pivoting: the lid of a Tic Tac box, a case that snaps shut, a clip that joins two halves without a single screw. The idea is beautiful because it replaces an entire assembly—pin, knuckles, spring—with a single part that comes off the bed in one piece. And it's treacherous for the very same reason: that thin strip is exactly where an FDM part is weakest, and one slip in material or orientation breaks it on the third fold. The difference between a hinge that survives a thousand cycles and one that snaps on the first fold barely shows in the drawing. It's in which plastic you pick and how you stack the layers.

Why a rigid plastic cracks instead of flexing

When you fold the strip, its outer face stretches and the inner face compresses. The tighter the fold and the thicker the strip, the greater that stretch at the surface. A flexible plastic absorbs the strain by yielding and recovering; a rigid plastic can't, so it opens a microcrack on the face in tension. Every cycle deepens it a little. That's fatigue, and it's cumulative: it doesn't fail from a single overload; it fails from repetition.

And here's what nobody likes to hear: PLA makes terrible hinges. It's rigid and brittle, its strain at break is far too low, and no matter how well you orient it, it cracks within a few dozen cycles. There's no print trick that saves it, because the limit isn't the weld between layers: it's the material itself, which can't tolerate the repeated strain a fold asks of it.

The materials that do flex repeatedly share one virtue: they take a lot of elastic strain before they're damaged.

• PP (polypropylene) — the gold standard. It's the reason the living hinge exists as a concept at all: injection-molded PP hinges survive millions of cycles. Printed PP is fussy about bed adhesion, but nothing else on this list comes close to it in fatigue life.
• PE (polyethylene) — the same polyolefin family, soft and tough, with that same ability to fold without cracking. It's just as awkward to print as PP because of its shrinkage and poor adhesion, but as a hinge, it lasts.
• TPU — folds indefinitely because it's elastomeric. It's worth understanding what that implies: it's soft and recovers slowly, and it isn't rigid. It's perfect for straps, gaskets, and bellows, but it won't snap a lid shut crisply the way a PP hinge does, and for a clip that has to latch with force it's a poor choice: its low modulus provides almost no retention force.

The thin web: 0.4–0.8 mm, perimeter not infill

The section that bends—the web of the hinge—wants to be thin enough to fold without resistance and thick enough not to tear. Aim for 0.4–0.8 mm at the thinnest point. Below one bead width you can't even fit a full perimeter, so the wall comes out brittle and inconsistent; above 0.8 mm the strain on the outer surface rises so steeply when you fold that the material itself cracks. That lower figure, 0.4 mm, is no accident: it's exactly one bead wide from a standard nozzle, the minimum at which the web prints as a solid wall and not as a poorly stitched patch of infill.

What's decisive isn't just the thickness, but what that thickness is made of. The web has to be 1 or 2 continuous perimeters, no infill. A perimeter is one bead that runs the band from side to side without interruption; it works and bends as a single fiber. Infill, by contrast, is a mesh of crossing threads with gaps: in a section a few tenths thick it never forms reliably, and where it does appear it leaves voids where a crack starts with no effort at all. Size the web as a whole multiple of the bead width so the slicer fills it with full perimeters, and leave it no internal void.

Spread the bend over a band, not a line

With the material and the thickness settled, the shape does the rest of the work. A bend concentrated on a sharp line is a guaranteed crack initiator: all the stretch piles up in one millimeter and that's where the first fissure starts. Give the hinge a 3–5 mm band to fold across and that same strain spreads along the width, so no single point reaches the limit.

The same reasoning applies to the transitions. Where the thin web meets the thick panel there's an abrupt change in stiffness, and abrupt changes concentrate stress. If that junction is a sharp step, it tears there; if you taper it gradually—a gentle concave throat that thins toward the center instead of a square notch—the strain flows along the transition rather than slamming into a shoulder. Fillet every corner of the flexing zone. On a part designed for fatigue, a generous radius isn't styling: it's service life.

There's a failure mode that creeps into wide hinges and that the outer face alone doesn't explain. A long hinge doesn't open uniformly: it almost always starts to fold from one end, so the root isn't loaded across its whole width at once, but from one edge. That edge concentrates the tearing and the crack runs sideways along the root, like a sheet of paper tearing from a corner. That's why it pays to finish the ends of the flexing band with a fillet in plan, not a sharp corner, and to keep the opening from always starting at the same edge. A uniform web and a root filleted in both planes—the plane of the fold and the plane of its width—leave that tear nowhere to start.

The orientation mistake that ruins the hinge

It's the failure that ruins more living hinges than any other, and it's counterintuitive. The temptation is to print the part flat on the bed and let the bend axis fall wherever it lands. Left to chance, that axis most likely crosses the layer lines: when you fold, you're pulling right on the plane that separates one layer from the next, in the thinnest and most stressed section of the whole model. The weld between layers sits well below the strength of the solid bead (Layer adhesion and anisotropy develops this), and here you've handed that weld all the work. The hinge delaminates on the first or second fold, and it does so dead straight, along a layer line.

The rule that prevents it is geometric: print flat, with the bend axis aligned to a layer and the web beads parallel to the fold. That way, when the hinge closes, you're stretching continuous plastic within the plane of one layer—the strong direction—instead of peeling one layer off the next. The fold loads the fiber along the bead, which is exactly what the bead is strong for, and not across the seam between beads, which is what it's weak for. The same hinge, printed on edge with the layers perpendicular to the fold, delaminates on the first cycle no matter how right the material is.

If the geometry of the assembly leaves no orientation with the beads parallel to the fold, don't force the thin membrane: accept that the hinge will be a low-cycle one and split it into two parts with a real pin hinge.

A flexure obeys the same physics

A snap fit, a spring clip, a latching tab: they're all flexures, and they obey exactly the same principle as the living hinge, only run in reverse. The hinge aims to bend thousands of times; the flexure usually bends few times—sometimes just once, during assembly—but in exchange it's asked for a larger deflection in one go. In both cases the rule is the same: keep the strain in the part below the material's limit. Stay in the elastic zone and the part bends and returns intact as many times as you need; cross it and it fails—immediately if the material is brittle, by fatigue if you repeat it.

From there come the same three factors. The material has to tolerate the strain you ask of it: a clip that flexes a lot at every latch wants a tough plastic that also springs back firmly—PP, copolymer, or nylon—rather than a soft TPU (which bends without latching) or a rigid PLA (which snaps). The geometry spreads that strain instead of concentrating it: a tab that thins toward its base and grows out of a fillet survives; one that springs from a sharp corner snaps right there. And the orientation decides which way the weakness points: a tab printed standing up flexes against the layer planes and breaks on the first try, while laid down, with the beads running along the flex, it bends through solid plastic. Where the flexure really gives—the fillet at the root, the bead running through the zone of peak stress—is the same place a poorly oriented living hinge would fail.

That's why it pays to treat hinges and flexures as a single family: the cycle count and the magnitude of the bend change, but the physics that decides whether they survive is identical. Get the flexible material right, set the web at one or two beads of continuous perimeter, and run the beads along the fold, and you'll have a part that replaces an entire assembly. For the opposite case—when what you want isn't to bend a thousand times but to snap in and stay put—carry on to Print-friendly snap fits. And to understand fully why the seam between layers is the weak link this orientation sidesteps, read Layer adhesion and anisotropy.