Snap-fits that don't break: strain limits, materials and print orientation

A cantilever snap-fit is a small arm with a hook on the end. It bends out of the way as the lid goes on, then springs back behind a ledge. It works only if the bend stays inside what the plastic can stretch without yielding. This lesson turns that limit into one formula you can work out by hand, then covers orientation, taper and the root fillet.

Numbers assume FDM with a 0.4 mm nozzle and 0.2 mm layers. If you need fits and clearances first, read Print to the right size.

What actually breaks a clip

When the arm bends, the outer surface at its root stretches the most. That stretch, as a fraction of length, is the strain (ε). Every plastic has a strain at which it stops springing back (the yield point). Go past it and the root whitens, the clip stays bent, or it cracks. So we design for strain, not for force.

For a straight arm of constant thickness, beam theory gives the deflection at the tip for a given strain at the root:

y = 0.67 × ε × L² ÷ h

• y = the deflection the arm must make = the height of the hook (undercut), mm
• ε = allowed strain as a fraction (2% = 0.02)
• L = arm length from root to hook, mm
• h = arm thickness at the root, mm

Read it this way: length helps a lot (squared), thickness hurts (a thicker arm reaches the strain limit sooner). This is the formula in Table 1 of the Bayer (now Covestro) snap-fit design guide for a rectangular arm.

Strain limits for printed plastic

The Bayer guide gives the rule: for one short snap, amorphous plastics may go to about 70% of their yield strain (partly crystalline ones almost to yield). For clips that are opened and closed often, use about 60% of that value.

Moulding guides list values for moulded resins, not for filament. For filament, take the elongation at yield from the data sheet, measured on printed test bars. Here is the calculation with Prusament’s published sheets, using the stricter 70% rule for both:

Material	Elongation at yield (printed flat)	Single snap (×0.7)	Frequent use (×0.6)
PLA	2.9%	≈ 2.0%	≈ 1.2%
PETG	5.1%	≈ 3.6%	≈ 2.1%

These are lab bars printed under controlled settings. Your part has seams, different walls and a different spool, so treat the numbers as an upper starting point and confirm with a test clip. Other filaments: look up “elongation at yield” on their data sheet and apply the same two factors.

PETG takes nearly twice the strain of PLA. For a battery cover you open every week, that difference decides whether the clip lasts.

Worked example: a battery cover clip

A battery cover is opened often. The hook needs an undercut of y = 1.0 mm. Arm length L = 12 mm. Material PETG, so design strain ε = 2% (0.02).

1. Thickest arm allowed: rearrange to h = 0.67 × ε × L² ÷ y = 0.67 × 0.02 × 144 ÷ 1.0 ≈ 1.93 mm.
2. Pick a thickness that prints as whole lines. Four perimeters with a 0.4 mm nozzle are about 1.8 mm wide (Prusa’s table), so h = 1.8 mm.
3. Check: allowed deflection = 0.67 × 0.02 × 144 ÷ 1.8 ≈ 1.07 mm, which is more than the 1.0 mm needed. It passes.

Now the same arm in PLA, frequent use, ε = 1.2%:

1. Allowed deflection = 0.67 × 0.012 × 144 ÷ 1.8 ≈ 0.64 mm. Less than 1.0 mm, so the root will be over-strained.
2. Fix it with length: try L = 16 mm. 0.67 × 0.012 × 256 ÷ 1.8 ≈ 1.14 mm. That passes.

Lengthening by 4 mm did what a material change did. If there is no room, make the hook shallower or move to PETG.

Shape: taper and root fillet

Taper the arm. The Bayer guide shows that an arm thinning linearly from h at the root to h/2 at the hook allows more than 60% extra deflection compared with a constant arm, because strain is spread evenly along the length instead of piling up at the root. In the example above, a 1.8 mm root tapering to 0.9 mm at the hook is a sensible shape. At the tip, 0.9 mm is still two perimeters.

Round the root. A sharp inside corner concentrates stress right where strain is highest. Hubs recommends a fillet of at least 0.5 × the root thickness (0.9 mm for our 1.8 mm arm). The Bayer guide notes that the benefit flattens out around R/h = 0.6 and never goes below about 0.38 mm (0.015 in).

Width. Hubs suggests a clip at least 5 mm wide. A wider arm is stiffer to press but no more strained, because width does not appear in the strain formula.

The hook. A shallow entry slope (the Bayer example uses 30°) makes assembly easy. The steeper the back face, the harder the clip is to pull off; a 90° back face only releases if you push the arm by hand.

Print orientation: the arm must lie flat

FDM parts are strong along the lines and weak between layers. Prusament PLA shows it clearly: about 51 MPa tensile strength printed flat, but only about 17 MPa between layers. An arm standing upright in Z bends across those layer lines, and that is where it cracks.

• Lay the part so the arm lies in the XY plane and the lines run from root to hook.
• If the case has to print standing up, turn the clip so it flexes sideways in XY, or print the clip as a separate part and fix it with a screw or glue.
• Hubs says the same: avoid cantilevers built vertically in Z.

More on layer direction and strength is in Part orientation.

Clearance around the hook

Hubs uses 0.5 mm as a nominal FDM clearance for snap-fits. The site’s fit tests usually land at 0.1–0.2 mm per side for a snug fit and 0.3–0.5 mm for loose (details). For a clip:

• Leave 0.3–0.5 mm between the arm and the wall it moves towards, so it can flex without rubbing.
• The ledge the hook catches on should engage the full undercut. If clearance eats half of it, the lid rattles.
• Chamfer the opening edge of the mating part to guide the hook in.

Print one clip and its ledge as a small test piece before you print the whole case. The test costs minutes; a cracked case costs hours.

Common mistakes

• Making the arm thicker “to be strong”. A thicker arm reaches its strain limit sooner and breaks faster. Make it longer instead.
• Arm printed vertically. It snaps along a layer line on the first press.
• Sharp inside corner at the root. The crack starts exactly there.
• Using moulded-plastic strain values for filament. Printed parts are weaker; use printed-specimen data.
• Designing for single use, then opening it daily. Repeated use needs the 60% factor.
• Heat. PLA softens from about 55 °C (Prusament data sheet HDT). A clip in a hot car will creep and lose its grip.

Sources

• Snap-Fit Joints for Plastics, A Design Guide — Bayer MaterialScience (PDF copy hosted at MIT)
• How to design snap-fit joints for 3D printing — Hubs
• Prusament PLA technical data sheet — Prusa Polymers (PDF)
• Prusament PETG technical data sheet — Prusa Polymers (PDF)
• Modeling with 3D printing in mind — Prusa Knowledge Base