Every mold designer eventually gets the same request: "I know this feature is an undercut, but can you just... make it work?" Sometimes the answer is yes for free. Sometimes it costs $8,000 and two weeks. Knowing which is which before you finalize a prototype design is the whole game, so here is how undercuts actually get handled in plastic injection molding.
What counts as an undercut
Any geometry that prevents the part from ejecting straight out of the mold is an undercut. The usual suspects:
- Snap-fit hooks and clip fingers
- Side holes, side slots, and cable pass-throughs
- Internal threads
- Textured grip surfaces deeper than ~0.1 mm in rigid resins
- Enclosed features the core can't escape from
A subtle point people miss: a through-hole on a side wall is an undercut only if its axis is parallel to the mold-opening direction. Rotate the part in the mold and the same hole becomes a simple core-out. Tooling orientation is a design variable, not a fixed constraint — one of the cheapest fixes is simply asking which way the part sits in the tool.
Sliders: the workhorse
A slider (or "slide and core") is a metal block on an angled gib that retracts sideways as the mold opens, pulled by an angle pin or cam. Sliders handle side holes, side ribs, and external snap features in virtually any resin.
The costs are real but bounded: a prototype-tool slider typically adds $800–2,500 and about a week to an aluminum soft tool, more in steel production tools. Wear points are the gib and the angle pin, and sliders leak flash at the shut-off face after a few thousand cycles if the wear plates aren't maintained — one reason prototype shops prefer to avoid them unless the geometry genuinely needs one. Sliders also leave witness marks, so plan cosmetic surfaces away from slide shut-offs.
Lifters: the inside job
Internal undercuts — a hook on the inside of a box, an inward-facing clip — need lifters (also called angled lifters or "jazkers"). The lifter is an angled insert set in the ejector plate: as it ejects, it moves both outward and forward, walking the hook free of the plastic.
Lifters have harder limits than sliders. The travel follows the angle, so releasing a 0.8 mm undercut at a 15° angle needs roughly 3 mm of ejection stroke devoted just to that feature. In deep boxes, that stroke competes with part stability during ejection. Lifters also trace an arc, not a straight line — the feature must tolerate it, which usually means keeping lifter undercuts shallow: 0.3–0.8 mm is the comfortable range for ABS and PC, less for glass-filled resins where the material doesn't stretch. If your snap hook needs 1.5 mm of engagement, a lifter alone won't do it; you either split the hook's release across two stages or redesign.
When you don't need any mechanism at all
Semi-crystalline and flexible resins buy you free undercuts. PP, TPE, TPU, and LDPE parts routinely pop over 0.5–1.5 mm undercuts because the material stretches past the feature at ejection. Rules that hold up in practice:
- PP snap hooks with 0.6–1.0 mm engagement: routinely stripped off the core, no mechanism
- ABS: 0.2–0.4 mm strippable if the hook has a lead-in radius and the ejection is warm
- PC: forget stripping below about 0.1–0.2 mm; polycarbonate's notch sensitivity means a sharp hook corner plus forced deflection is a crack waiting to happen
- Glass-filled grades (PA6-GF30, PC-GF20): essentially zero strippable undercut — the fibers don't stretch
Temperature matters too. Ejecting a PP part hot, right at ejection temperature, doubles the stretch it can survive compared to letting it cool fully. Prototype molds that cool aggressively can actually be worse for strippable undercuts than production tools.
The escape hatches
Three more options that avoid moving mold hardware:
- Collapsible cores for internal threads and deep internal features: the core segments fold inward on ejection. Expensive and maintenance-heavy, but sometimes the only way to mold an internal thread. For most prototypes, machining or tapping the thread afterward beats a collapsible core.
- Telescoping / two-stage cores for deep thin internal ribs where a lifter can't fit.
- Lost features: accept a slot instead of a hole, or split the part into two moldings joined later. On a 100-part prototype run, a $0.40 secondary assembly step often beats a $1,500 slider.
A decision sequence that works
When you review a design with undercuts, walk it in this order:
- Can the part be reoriented in the mold so the feature isn't an undercut? Free.
- Can the undercut be reduced below the strippable limit for your resin? Free, plus a lead-in radius helps ejection in every case.
- Slider or lifter? Cost is known and contained; budget roughly a week and $1–3k on a prototype tool, and keep the feature off cosmetic faces.
- Secondary operation? For low volumes or one-off threads, machining after molding is often cheaper than tooling for it.
Prototype-specific advice
Soft tooling changes the calculus slightly. Aluminum prototype molds handle sliders fine — the mechanism doesn't care what the mold is made of — but each added mechanism is added lead time, and prototype programs usually live on schedule, not per-part cost. Our default on a first article: strip what you can with material choice and radii, slider only what functionally must be molded, and defer the rest to secondary ops until the design is proven. When the design graduates to production, review the deferred list again — features like side holes that were drilled on 20 prototypes are usually worth a slider in the production steel tool. You can see how this plays out across our plastic injection molding services and our soft and production tooling options, and if the part might later move to metal, comparing with die casting draft and pull-direction rules early avoids redesigns later.
The one-sentence version: undercuts are a scheduling and cost dial, not a yes/no wall — and the cheapest turn of that dial is always geometry.
