Thermoforming ToolingCavity tools for trays and blisters — from simple vacuum-only negative tools to plug-assisted deep draws, prototyped in days and machined for series.

The polarity question is really a question about which surface of your part carries the specification. A negative (female) tool forms the film into a cavity, so the tool machines the pack’s outer surface: outer length, width, stack shoulder and — critically — the sealing flange all come out at tool dimension, shot after shot. That is why trays that must seal against a lidding film, or nest in a machine pocket downstream, are almost always negative-formed.

A positive (male) tool drapes the film over a core, giving the inner surface tool accuracy instead. Choose it when the part must fit snugly over or around a product — a formed insert that locates a device, or a cover whose inside diameter matters. The trade-offs: material is thickest at the top of the form and thinnest at the base flange (the opposite of negative forming), and the part shrinks onto the core as it cools, so draft and de-moulding need more care.

Wall distribution follows polarity too. In negative forming the last film to touch the tool is in the bottom corners — the thin zone. In positive forming the first contact is the top of the core, which chills and stays thick while the skirt stretches. Neither is better; they simply put the material in different places, and the right choice is the one that puts thickness where your part is loaded.

A plug is not just a pusher — it is a pre-forming die that decides where the film’s thickness ends up. The film that touches the plug cools slightly and stops stretching; the film between plug and cavity keeps thinning. A plug that fills ~70–85% of the cavity volume, with generous radii and slightly steeper walls than the cavity, parks material in the corners where vacuum-only forming starves them.

Material matters because contact cooling is the mechanism. Syntactic foam (glass microspheres in epoxy) is the industry default: low thermal mass, low friction, no marking. Machined POM or PTFE-coated aluminium work for simpler shapes. In our printed tool sets the plug is often printed too — same CAD, same iteration speed as the cavity.

Timing is the third lever. The plug should enter just before vacuum, pre-stretching the sheet to ~80–90% of depth; vacuum then snaps the film the last distance onto the cavity wall. Too early and the film chills and webs; too late and the plug does nothing. On machines with adjustable plug delay this is a one-parameter optimisation — and it is stored in the tool’s parameter recipe, so it loads automatically with the tool.

Air has to leave the cavity exactly as fast as the film arrives — the vent map is the tool’s respiratory system. Holes of 0.3–0.6 mm are the working range for typical packaging gauges: small enough not to leave a visible nipple on the product face, large enough not to choke flow. Behind each hole the bore opens to 2–3 mm so the restriction is only the last half-millimetre.

Placement beats quantity. Air collects where the film lands last: the corners, the perimeter of the base, the bottom of any lettering or logo engraving. A cavity with a dozen well-placed vents outperforms one with fifty drilled on a grid. Slot vents — 0.2 mm gaps at insert part-lines — vent long edges invisibly and are a free benefit of building tools from cavity inserts.

Diagnosis from the part is straightforward: a bridged or rounded corner means trapped air (add vents there); a shiny patch on a matte part means the film never touched the tool (same cause); a nipple means a vent is too large or its bore too shallow. In production, vents slowly load with plate-out and dust — falling part definition over weeks is usually a cleaning schedule, not a process drift.

Draft exists because the part shrinks while it sits in the cavity — onto walls that lean in, off walls that lean out. 2–3° releases smooth negative cavities; textured surfaces need 1–2° extra because the texture keys into the film. Zero-draft walls are possible with ejection support, but they cost cycle time and part stress.

Radii are wall-thickness insurance. The film thins most where it turns fastest, so a bottom corner radius below ~2× gauge concentrates both thinning and impact load in the same spot — the classic cracked-corner failure. Our R4-standard on precut sheet corners has the same logic: no sharp corner, no crack initiator.

Undercuts lock the part into the tool by definition, so the default is to design them out. Small functional undercuts — a snap rim on a lid, a retention bead for a device — are formable in tough films like PETG, which flexes over the lip on ejection: keep them under ~1.5 mm engagement, radius them fully, and place them where the ejection lift pushes, not where it pulls. Anything deeper needs split cavities or collapsing inserts, which moves the tool into another cost class.

The economics push toward more cavities per shot; the physics pushes back. Every cavity must see the same film temperature and the same vacuum curve, or the tool produces two grades of part in one stroke. Pitch comes first: cavity spacing must match the downstream sealing and cutting stations (and any denesting automation), which is why our forming, sealing and cutting modules share datum and pitch by design.

Vacuum balance is plumbing: cavities nearest the vacuum port evacuate first and form first, stealing film from their neighbours. A balanced manifold — equal path lengths, or a plenum under all cavities — keeps the forming front simultaneous. On the film side, edge cavities run cooler than centre cavities under a uniform heater; zone-mapped heating or a slightly wider film margin compensates.

A practical sequence for scaling: prove the geometry single-cavity on a printed tool, then build the multi-cavity aluminium tool with the proven cavity as an insert. Insert construction keeps all cavities literally identical, lets you replace a damaged one without re-machining the tool — and gives you the slot-vents at the part lines for free.

Forming takes tenths of a second; cooling takes the rest of the cycle. The part must drop below the film’s heat-distortion range before ejection or it warps as it relaxes. In an aluminium tool with water channels 10–15 mm behind the cavity surface, a 0.5 mm PETG part sets in a few seconds; in an uncooled printed tool the same part needs several times longer — which is exactly the cycle-time difference between prototype and series tooling.

Tool temperature is a quality parameter, not just a speed one. Too cold a tool chills the film on contact and prints stress into the part (visible as blush or warp off the tool); too warm and cycle time balloons. 40–60 °C is the usual window for PETG/APET negative tools — tempered, not merely cooled, so shot one and shot one-thousand see the same tool.

The classic startup failure is condensation: a tool colder than the room’s dew point sweats, and the first shots carry water marks. The cure is procedural — bring the tool to temperature before threading film, and keep coolant above dew point. It is also why tool temperature belongs in the parameter recipe the machine loads automatically with the tool.

The generic industry view of printed thermoform tooling — a fragile pattern good for a demo — is outdated. Printed in the right high-temperature polymer, with vent bores designed in rather than drilled after, a cavity holds its geometry for thousands of shots at packaging-typical gauges and temperatures. For small-series production, seasonal formats and frequent design revisions, the printed tool is not a compromise: it is the economically correct tool.

What the printed route buys you is iteration speed: a cavity revision is a re-print, not a re-machining — days and a fraction of the cost. The vent map, draft and radii transfer unchanged to aluminium later because both tools are cut from the same CAD with the same datum. Our combine-ready interface applies too: a printed forming module can run in the same frame that later carries the machined one.

Where aluminium still wins is heat: without water channels a printed tool cycles slower and drifts warmer over a long run, and its surface will not survive hundreds of thousands of shots. The honest rule: printed for validation, pilot and small series; machined aluminium when volume makes seconds-per-cycle the dominant cost. We quote both routes on every project so the crossover point is visible in numbers, not opinions.

Thermoformed parts shrink as they cool — ~0.4–0.7% for PETG and APET (estimate; grade-dependent) — so the cavity is cut oversize by the shrink factor. The subtlety: shrinkage is not perfectly uniform. Restrained geometry (deep ribs, tight corners) shrinks less than free walls, and the flange, clamped between guide pins, shrinks differently from the base. Series tools for tight-tolerance parts budget one correction loop after first-article measurement — which is precisely what the printed prototype pass eliminates: measure the printed part, correct the CAD once, machine the aluminium right.

On achievable numbers: tool-side dimensions (the surface the film touches) hold ±0.1–0.2 mm in a tempered aluminium tool; free-side dimensions depend on film gauge and draw and should carry ±0.3 mm or more. Wall thickness is a distribution, not a dimension — spec the minimum at the identified thin point (bottom corners in negative forming), never a uniform nominal.

Practical drawing discipline: mark which face is tool-side, put tolerances only on dimensions that matter downstream (seal flange, stack features, retention geometry), and state the design material — a tool corrected for PETG shrink will run APET, but critical dimensions shift by the shrink difference.

An aluminium forming tool has no wear mechanism in normal service — the film is softer than the tool. What degrades is everything around the cavity: vents load up with plate-out and dust (falling part definition), seals and O-rings in the vacuum path harden (slower forming), ejection pins gum up (marking). A monthly routine — blow vents through from the back, wipe the cavity with a non-scratch solvent, cycle the ejector by hand — prevents essentially all of it.

The real lifetime risks are mechanical and human: a dropped tool, a crash with a mis-indexed film, corrosion from wet storage. Store tools dry, on their transport base, with the cavity face protected — and let the engraved tool ID carry the history: revision, design material, shot count if the machine logs it (our auto-detection does), last service date.

Refurbishment beats replacement in most cases. A polished-out scratch, a re-drilled vent field or a single replaced cavity insert restores the tool at a fraction of new-tool cost — another argument for insert construction. When a format retires, the frame, base and plug system usually live on under a new set of inserts.