Will my snap-fit arm break, and what force does it take to assemble? Enter the arm length, thickness and width, the undercut it has to clear and the resin. The calculator gives the strain at the root against the resin's allowable strain, the force that bends the arm, the push to assemble the joint and the pull to separate it, and the shortest arm or deepest undercut that would still be safe.
Engineering estimate, not a quote. Final snap-fit geometry comes from the resin grade's design data, the mold design review and an assembly test on molded samples at the coldest service temperature.
A cantilever snap-fit is a short plastic beam with a hook on its free end. During assembly the mating part rides up the hook's lead-in ramp and pushes the arm sideways by exactly the height of the undercut. When the hook clears the edge, the arm springs back and the retention face holds the parts together. Two questions decide whether it works: does the arm survive being bent that far, and how hard is it to push together and to pull apart?
Step 1: the strain at the root. Treat the arm as a cantilever beam fixed at the root and loaded at the hook. Bending is largest at the root, so that is where the strain peaks and where arms crack. Beam theory gives the deflection at the hook as y = c ε L² ÷ h, which turned around gives the root strain for a known undercut: ε = h y ÷ (c L²). For a straight arm of constant section c = 2/3 (published as 0.67), so ε = 1.5 h y ÷ L². Tapered arms spread the bending along their length and get larger values of c: 1.09 when the thickness tapers to half at the hook, 0.86 when the width tapers to a quarter. Note what the equation says: strain rises in proportion to thickness and undercut, but with the square of how short the arm is.
Step 2: compare with the allowable strain. Resin suppliers publish a permissible short-term strain for snap-fits that are assembled once or a few times. For joints that are opened and closed often, the common guidance is to design to about 60% of that value. The calculator reports the strain as a percentage of the allowable, the shortest arm that would be safe for your undercut and the deepest undercut that would be safe for your arm.
Step 3: the deflection force. The force at the hook that holds the arm at that deflection follows from the root bending moment: the root stress is Es ε and the moment is P L, so P = b h² Es ε ÷ (6 L), with b and h taken at the root for all three arm shapes. Es is the secant modulus, the stress divided by the strain at your design strain, which for plastics is lower than the initial modulus.
Step 4: assembly and separation forces. The mating part does not push sideways; it pushes along the insertion direction, and the ramp converts that push into the sideways force P. With friction μ on the ramp, the push needed is W = P (μ + tan α) ÷ (1 − μ tan α). The same relation with the retention angle α′ gives the pull needed to separate the joint. When μ tan α′ reaches 1 the denominator reaches zero: no pull along the axis can lift the hook, and the joint is self-locking.
An ABS battery-door latch uses one straight cantilever arm 15 mm long from the root to the hook, 1.5 mm thick and 6 mm wide, with a 1.0 mm undercut. It is assembled once at the line and opened only for service. ABS is taken at a 2.5% allowable strain and a 2,300 MPa modulus, with μ = 0.5, a 30° lead-in ramp and a 45° retention face.
The latch is comfortably safe: the root sees 1.0% strain, 40% of what ABS can take once and two-thirds of the repeated-use limit, so it also survives service openings. It assembles with about 5 N of push per arm, and pulling it apart takes about twice that, which suits a door that should stay shut but come open with a deliberate pull. The limits show where the margin is: the same undercut would break an arm shorter than about 9.5 mm, which is exactly what happens when an arm is shortened late in a design to save space. At 8 mm the root would see 3.5%. Tapering the arm to half thickness at the hook would cut the strain and the push by about 39%. Press Load the worked example above to reproduce these numbers.
The strain equation has only three geometric terms, and each behaves differently. Length is the strongest lever because it is squared: going from 12 mm to 15 mm cuts the strain by 36%. Thickness is linear in strain but squared in force, so thinning an arm from 1.5 to 1.2 mm cuts strain by 20% and the deflection force by about 49% (0.8³ ≈ 0.51 at the same undercut, since force goes with h² × ε). Undercut is linear in strain and force, and it is also the retention: a smaller hook is easier on the arm but easier to pop out. Width b does not change the strain at all; it scales the force and the holding strength, which makes it the right knob when the strain is fine but the joint feels weak.
In practice we start by finding the undercut the joint really needs, set the length as long as the part allows, choose the thickness for strain, then use width for force. Tapering comes next, before adding material.
| Method | Use it when | Limit |
|---|---|---|
| This calculator | Concept sizing of straight or tapered rectangular arms; checking a supplier's latch; choosing length, thickness and angles in a design review | Small-deflection beam theory; rigid root and rigid mating part; no stress concentration; representative material values |
| Resin supplier's snap-fit guide | Allowable strain, secant modulus and friction for the exact grade; equations for ring, annular and torsion snap-fits | Generic shapes; does not know your wall, root or mating part |
| Finite element analysis | Short stubby arms (L under about 5 h), arms rooted in thin flexible walls, L-shaped or U-shaped arms, large deflections | Needs non-linear material data to be better than the hand method |
| Assembly and pull tests on molded samples | Final confirmation, with real weld lines, moisture and temperature; setting acceptable assembly force for the line | Needs molded parts; a prototype-material test can mislead |
| Screws or other fasteners | Joints that are opened very often, carry high loads, or must clamp a gasket evenly | An extra part and an extra operation; see our boss calculator |
The deflection, strain and force equations and the mating-force relation are those published in resin suppliers' snap-fit design manuals for cantilever arms of rectangular section, including the Bayer (now Covestro) Snap-Fit Joints for Plastics design guide and BASF's snap-fit design guidance: y = 0.67 ε L² ÷ h for a uniform arm, 1.09 for an arm tapering to half thickness and 0.86 for an arm tapering to a quarter width; P = b h² Es ε ÷ (6 L); W = P (μ + tan α) ÷ (1 − μ tan α); and the guidance to design to about 60% of the permissible strain for frequently separated joints. The uniform-arm coefficient follows directly from cantilever beam theory (deflection P L³ ÷ (3 E I), root strain 6 P L ÷ (E b h²)). The allowable strains, moduli and friction in the resin menu are representative published values, labelled as such and fully editable; your grade's design guide overrides them. The 0.5 h root radius, the L ≥ 5 h validity screen and the design tips are page-set rules of thumb from part-design practice, not resin data.
For a straight arm of uniform rectangular section, the strain at the root when the hook is deflected by the undercut y is ε = 1.5 × h × y ÷ L², where h is the arm thickness at the root and L is the length from the root to the hook. This is the beam equation used in resin suppliers' snap-fit design manuals, written there as y = 0.67 × ε × L² ÷ h. For example, a 15 mm long, 1.5 mm thick arm with a 1.0 mm undercut sees 1.5 × 1.5 × 1.0 ÷ 225 = 0.0100, or 1.0% strain.
It depends on the resin and on whether the joint is assembled once or many times. Representative short-term values published in resin suppliers' snap-fit guides are about 2.5% for ABS and PC/ABS, about 4% for unfilled PC, about 4% for dry and about 6% for conditioned PA6, and about 1.5–2% for 30% glass-filled PA6. For joints that are taken apart often, design to about 60% of the one-time value. Your grade's design guide overrides these numbers.
First compute the deflection force P = b × h² × E × ε ÷ (6 × L), the force at the hook that bends the arm through the undercut. The push needed along the insertion direction is then W = P × (μ + tan α) ÷ (1 − μ × tan α), where α is the ramp angle and μ the friction coefficient. A 6 mm wide, 1.5 mm thick, 15 mm long ABS arm at 1% strain and a 2,300 MPa modulus has P = 3.45 N; with a 30° ramp and μ = 0.5 the assembly force is about 5.2 N per arm.
Because the strain is highest at the root and grows with the square of the shortness of the arm. Halving the length quadruples the strain for the same undercut, so a short stubby arm can be far past the resin's limit even with a small hook. A sharp inside corner at the root concentrates the strain further, and weld lines, cold temperatures and dry nylon all lower what the root can take. Lengthen the arm, thin or taper it, reduce the undercut, and give the root a radius of about half the arm thickness.
Usually yes. An arm whose thickness tapers from h at the root to h/2 at the hook can deflect about 1.09 ÷ 0.67, or roughly 1.6 times, as far as a uniform arm at the same root strain, because more of the arm bends instead of only the root. For the same undercut its root strain and its deflection force are both about 39% lower. Tapering the width to a quarter at the hook gives a smaller gain, about 22% lower strain.
It depends on whether the joint should come apart. The separation force follows the same relation as the assembly force, using the retention angle. When μ × tan α′ reaches 1, the joint is self-locking and cannot be pulled apart without breaking or releasing the arm by hand; with μ = 0.5 that happens at about 63°. A 90° face is a permanent latch. Shallower angles, often 30–45°, give a joint that can be separated by pulling.
Use the secant modulus at your design strain, read from the resin's stress-strain curve at the right temperature, rather than the initial tensile modulus on the datasheet. Plastics soften as strain rises, so the tensile modulus overstates the deflection force, often noticeably at 2–4% strain. The strain check does not depend on the modulus at all; only the forces do.
Not sure whether to lengthen the arm, taper it or switch to a screw? Ask our AI assistant. It answers from this website, and our tooling team follows up if you want.
AI assistant. Answers can be wrong; nothing it says is a quote or a promise.
This calculator gives an engineering estimate. Our tooling team reviews every latch on your CAD for strain, assembly force and how the hook will come out of the mold, and proposes arm, root and lifter changes while each one is still a CAD edit.
Book a 30-min call Send a drawing or sample