Injection Molding in China: Molds, Materials, and the Volume Math

Injection Molding · High-Volume Plastics in China Injection Molding in China: Molds, Materials, and the Volume Math Injection molding turns a few cents of resin into a finished plastic part in seconds — once you pay for the steel mold. Here is how the process works, which thermoplastic to pick, what the tooling really costs, the tolerances molded parts honestly hold, and the DFM rules that decide whether your part molds clean or fights the tool. Request a quoteDFM rules What injection molding is Injection molding is a formative process that forces molten thermoplastic into a closed steel cavity under high pressure, cools it until it solidifies, and ejects a finished part. It is the workhorse of mass-produced plastic: housings, caps, gears, connectors, and enclosures are all molded, and the economics are the whole story. A mold is expensive and takes weeks to build, but once it exists, each part costs a few cents of resin and a few seconds of cycle time. Its neighbors frame the trade. CNC machining cuts a solid block with no tooling — right for prototypes, metal parts, and tolerances in the micron range. 3D printing builds a part layer by layer with no tooling at all, right for one-offs. Die casting is molding's metal cousin, for zinc and aluminum. Injection molding sits at the far end of the volume curve: the highest fixed cost in the tool, the lowest variable cost in the part. The trade is geometry — a molded part must release from a two-piece mold, so undercuts and deep internal features are costly or impossible. The design artifact that matters is the mold design — cavity layout, gate placement, ejection, and cooling. It, not the part drawing alone, sets the tool cost, the cycle time, and the tolerance the part can hold. A molder who walks the mold design before quoting understands the process. The molding process, step by step Every molded part passes through the same five stages on every cycle. Understanding them is how you read a mold quote, spot a fill or shrink problem, and know where cycle time and tolerance are won or lost. Mold. The two halves close — the A-side cavity and the B-side core — aligning the steel that will form the part. Clamp. The machine builds clamp tonnage to hold the halves shut against injection pressure. Clamp force sizes the press: more projected part area needs more tonnage, or the mold flashes. Inject. The screw pushes molten resin through the sprue, runner, and gate into the cavity. Injection pressure, speed, and melt temperature decide whether the cavity fills cleanly or short-shots and burns. Cool. The part solidifies against the cavity walls. Cooling dominates the cycle — typically 50–80% of it — which is why cooling channels are drilled into the mold steel. Shrinkage happens here, and it is the root of every tolerance the part will hold. Eject. Ejector pins push the finished part off the core once it is rigid enough to release without deforming. Ejection is where draft angle earns its keep: without it, the part drags, sticks, or shears a boss off. The full cycle runs in seconds to tens of seconds depending on wall thickness. Thin walls and aggressive cooling shorten it; thick walls are the enemy, because they hold heat and extend the cool. Materials: ABS, PC, nylon, PP, and POM Five thermoplastics cover most molded parts. The choice follows stiffness, toughness, heat, chemical resistance, and cost — and each shrinks differently, so the material is also a tolerance and DFM decision. Cross-check any grade against the ASM Handbook before committing a drawing. ResinTypeWhat it gives youWhat to watch ABSAmorphousCheap, tough, easy to mold, great cosmeticsPoor solvent/UV resistance; softens above ~80°C Polycarbonate (PC)AmorphousStiff, tough, clear, heat-resistantStress-cracks; needs drying; more expensive Nylon (PA6/PA66)Semi-crystallineStrong, wear-resistant, heat-resistantAbsorbs moisture — dimensions move; needs drying Polypropylene (PP)Semi-crystallineCheap, light, chemical-resistant, hinges wellSoft, low stiffness, high shrinkage, hard to bond or paint POM / acetalSemi-crystallineLow friction, stiff, precise — gears and bearingsSlippery to paint or bond; degrades if overheated Amorphous resins (ABS, PC) shrink less and hold tighter tolerance; semi-crystalline resins (nylon, PP, POM) shrink more and move more as they cool — plan looser callouts on them. Two practical notes. First, hygroscopic resins — nylon and PC above all — must be dried before molding, or moisture turns to steam in the barrel and splay, brittleness, and voids appear on the part. Second, glass-filled grades raise stiffness and cut shrinkage at the cost of abrasion on the mold and a rougher, more anisotropic surface. The full material-selection guide. Mold types: two-plate, hot runner, and multi-cavity The mold is a precision steel assembly, and its architecture — how resin gets from the machine nozzle to the cavity — sets most of its cost and cycle time. Three choices cover the ground. Mold typeCostBest forTrade-off Two-plate, cold runnerLowestPrototypes, low volume, simple partsRunner waste; slower cycle; gate vestige Hot runnerHigher upfrontHigh volume, no gate vestige, no regrindMore complex; more to maintain; slower color changes Multi-cavityScales with cavitiesHigh volume of one partLonger build; cavities must stay matched These are not mutually exclusive — a production tool is usually a hot-runner, multi-cavity mold. The two-plate cold-runner tool is the entry point; the hot-runner multi-cavity tool is where unit cost bottoms out. Two-plate is the simplest: the part and a cold runner are ejected together, then the runner is trimmed or reground. The runner is waste — resin you paid for that does not ship — but the tool is cheap and easy to modify. Hot runner keeps the runner molten inside a heated manifold, so only the part ejects: no waste, faster cycles, and a cleaner gate, at the cost of a more expensive, more temperamental tool. Multi-cavity puts several identical cavities in one mold so every cycle drops several parts — the path to the lowest piece price — but the cavities must be cut to match, and every cavity is more tooling, more tryout, and more to keep in balance. Mold cost and lead time: the front-loaded bet For a sourcing engineer. The number that dominates any molding program is the mold. A simple two-plate prototype tool for a small, flat part can run a few thousand dollars; a hardened, hot-runner, multi-cavity production mold for a complex housing can run five figures and beyond, plus design, steel, and tryout. The cost is front-loaded and fixed — which is why injection molding is a volume decision, not a capability decision. Lead time tracks the cost. Mold design, steel, cavity cutting, fitting, and first-article trials commonly run four to eight weeks, and longer if the first shots reveal a fill or shrink problem that sends the tool back for rework. You cannot compress it easily, because the mold is a precision mechanical assembly, not a file you print. The amortization mathMold cost ÷ (machined per-part cost − molded per-part cost) = break-even quantity. Worked example: a CNC housing costs $14 at any quantity; the same part molded costs $1.10 after a $12,000 mold. The lines cross near 930 parts. That is the naive break-even — the practical number is higher, because you also carry mold lead time, first-article revisions, and the risk of a design change that strands your tooling. Two hidden costs push the real crossover above the arithmetic one. First, design risk: change the part after the mold is cut and you pay for a revision or a new mold. Second, cash and time: a mold ties up capital for weeks before a single part ships. Buyers who tool too early — on the formula alone, ignoring revision risk and time-to-market — are the most common over-spend in molding. Two tooling facts matter at quoting. First, ownership: confirm who owns the mold and where it is stored, because the tool is an asset you paid for. Second, steel grade: a soft pre-hardened steel is cheaper but wears; a hardened tool steel with the right surface treatment runs hundreds of thousands to millions of cycles. The cheap mold and the good mold look identical in a quote — the difference shows up a year into production. Tolerances: what molded parts honestly hold For a quality lead. Injection molding is a net-shape process, not a precision one. A competent molder holds roughly ±0.1–0.5 mm depending on part size and resin, and the range widens as the part grows. Shrinkage is the root cause: plastic shrinks as it cools, and every resin shrinks differently — amorphous materials like ABS and PC move less than semi-crystalline ones like PP and nylon. A good molder compensates with gate placement, packing pressure, and cooling control, but you are still buying tenths of a millimeter of capability, not microns. The loosest number is not the molder's fault — it is the material's. Shrinkage varies with wall thickness, gate location, and even the batch of resin, so a dimension that holds in one shot drifts in the next. That is why molding tolerances are specified as ranges, and why a drawing full of ±0.05 mm callouts is a CNC drawing wearing a molded disguise. When a feature genuinely needs machining-grade accuracy, you have three honest options: machine the feature after molding, insert a machined component into the mold, or post-mold the part — drill, ream, or face the surfaces that matter. The last is where in-house CNC earns its keep: the molding holds the shape and volume, the mill holds the tenths. DFM: draft, wall thickness, ribs, bosses, sink, undercuts For a manufacturing engineer. Most molding rework comes from six rules broken at the drawing stage. They cost nothing to respect and a full mold revision to ignore. Draft. Every wall parallel to the draw direction needs draft angle so the part releases — typically 0.5°–1° per side on textured or deep surfaces, less on shallow shiny ones. Zero draft means the part drags, sticks, or tears on ejection. Wall thickness. Keep walls uniform — roughly 1–3 mm for most resins — and avoid sudden steps. Thin walls short-shot; thick walls cool slowly and sink. The rule is uniformity, not absolute thickness. Ribs. A rib stiffens a wall without adding thickness — but keep it to roughly 50–60% of the wall it joins and add draft. A rib thicker than the wall leaves a sink mark on the opposite face. Bosses. A boss is a molded standoff for a screw — but it is a local thickening, so it sinks. Keep its wall thin, tie it to a side wall or rib for support, and plan a counterbore to hide the mark. Sink. These marks form wherever a thick section cools after a thin one and pulls the surface in. The cure is not cosmetics — it is wall uniformity, thinner ribs and bosses, and coring out thick sections. Undercuts. An undercut needs a side action, slide, or lifter — mechanism inside the mold that raises cost and complexity — and sometimes cannot be done without a design change. Every undercut you add is money you add to the tool. None of these rules add cost; they remove it. A part drawn with draft, uniform walls, and no undercuts drops out of the mold on cycle; the same part drawn without them spends a cycle being re-tooled. The complete DFM guide. Surface finishes Molding wins cosmetics out of the box. The cavity surface transfers directly to the part, so you can specify a glossy polish, a matte texture, or a fine grain and get it on every shot with no secondary operation. Textures like soft-touch or leather grain are molded in — a real advantage for consumer-facing housings — but a textured surface also needs more draft to release, so the finish and the draft angle are one decision. SPI finishes are the standard language: A for high gloss, B for semi-gloss, C for matte, D for textured — each with sub-grades, each produced by polishing or blasting the cavity to a spec. Texture and grain hide sink and flow lines, but they add draft and can trap the part on the core if under-specified. Secondary processes — painting, plating, metallization, pad printing, laser marking — are common on molded parts, and the finish choice gates them: PP and POM are hard to paint or bond without surface treatment, while ABS plates and paints easily. Specify the finish before the mold is polished, because the cavity finish is cut into steel and changing it later means re-polishing the tool. Injection molding vs CNC machining: the break-even Molding and CNC compete for the same plastic parts, and the decision comes down to three questions: what volume, what tolerance, what material. Molding wins at volume for thin-walled plastic parts: the mold is expensive and the design is locked, but per-part cost collapses to cents and repeatability is excellent. It loses at low volume and while the design is moving. CNC machining wins for low volume, metal parts, and tolerances in the ±0.005 mm range, with no tooling and an MOQ of one. It loses on per-part cost and on thin walls, which deflect and chatter under the cutter. The practical pattern is a sequence, not a choice. Machine prototypes and pilot runs through EVT → DVT → PVT while the design settles, because machining absorbs every change for free and ships in days. Then tool the mold once the geometry freezes and the forecast justifies it — and machine the critical faces on the molded parts when they land. A partner that runs CNC in-house and molding through vetted partners executes that whole handoff under one quality system. The full CNC-vs-molding breakdown. Why China for injection molding China's manufacturing belts run the densest clusters of mold shops and molders anywhere. In the Pearl River Delta — Dongguan and its surrounding towns — the mold-base suppliers, steel stockists, wire-EDM and grinding shops, and the molders themselves sit minutes apart, so a tool can be designed, cut, and first-articled inside one district without importing a single component. Dongguan in particular is the mold capital of the region: the mold bases, standard components, and the labor to build and run them are all local. That density changes the economics three ways. First, tooling is fast and cheap: a mold that costs weeks and a premium in a Western toolroom is a same-district order in the Delta, where the steel, machining, and tryout press are next door. Second, labor depth: mold designers and machine operators have run your exact part family thousands of times, turning a good mold into a great one. Third, shared finishing: the texture house, plater, and printer are specialists serving every molder, so a small shop delivers a finished part without owning any of those lines. The honest caveat applies to every process in China: the cluster holds world-class shops and shops that cut corners, often quoting similar prices. In molding the risk concentrates in mold steel — a soft substitute that wears early — and in material substitution, where a cheaper resin replaces the one on the drawing. The fix: specify the steel and resin grade, request a measured first article, and hold the partner to documented quality. Vetting a supplier without flying out. Nex-G: molding through vetted partners, QA-managed Nex-G's own floor is EMS and CNC under one roof in Dongguan Hengli — a 6,800 m² facility staffed by more than 100 people and running since 2006. Injection molding is sourced through a vetted partner network and QA-managed by Nex-G, which matters: the molded part comes back to a shop that can machine to the same drawing, inspect to the same standard, and integrate it into the same build the electronics live in. The anchor is the quality system behind the sourcing. Nex-G holds ISO 9001, IATF 16949 (build-to-print — design excluded under clause 8.3), and ISO 14001, with URS certification to 2027. SPC runs to a Cpk target of ≥1.67 on critical characteristics, backed by real data. That discipline is applied to sourced molding: the mold design is reviewed before steel is cut, the tool is approved from a measured first article, and incoming parts are inspected against the drawing before they enter your build. In-house, 80+ CNC machines from Mazak, Brother, TSUGAMI, and Sodick hold ±0.005 mm on milled and turned features and ±0.002 mm on grinding and wire EDM — the capacity that picks up any molded feature needing a machined tolerance, and that machines the prototypes that de-risk the mold. The service runs no MOQ — a minimum of one part — carries programs from EVT → DVT → PVT, and quotes lead times of 3, 7, and 30 days by scope. One partner manages the molding, the machining, and the board, so the fit is checked at the bench, not at your dock. See the full capability. On the floor the checks are concrete rather than aspirational. Every BOM line is cross-checked against the drawing and the purchase spec before release, so an ABS callout cannot silently swap to a cheaper off-spec resin. Critical resins are bought through a cross-checked second source with a matching material certificate, and each incoming lot is verified against that certificate before it reaches a press. For aerospace programs we run FAI per AS9102 so the first article is dimensionally signed off before the run begins. Frequently asked questions When does injection molding beat CNC machining?Above the break-even volume, where the mold cost amortizes across enough parts that the low per-part resin cost beats flat CNC machine time. For a small part that crossover is often in the low thousands, but design risk and lead time push the practical number higher. How much does an injection mold cost?A simple two-plate prototype tool can run a few thousand dollars; a hardened, hot-runner, multi-cavity production mold for a complex housing can run five figures and beyond, plus design and tryout. The cost is front-loaded, which is why the mold should be priced from a mold design, not a guess. What tolerances can I realistically hold?Plan for ±0.1–0.5 mm depending on part size and resin. Amorphous resins like ABS and PC hold tighter than semi-crystalline ones like PP and nylon. For a machining-grade feature, machine it after molding or insert a machined component. How long does a mold take to build?Commonly four to eight weeks from design through steel, cutting, fitting, and first-article trials — longer if the first shots reveal a fill or shrink problem. You cannot compress it easily, so machine prototypes in parallel to keep development moving. Which thermoplastic should I choose?ABS for cheap, tough, cosmetic parts; PC for stiffness, clarity, and heat; nylon for strength and wear; PP for cheap, light, chemical-resistant parts; POM for gears and bearings. Confirm the grade against the ASM Handbook, and dry hygroscopic resins before molding. Hot runner or cold runner?Cold runner is cheaper upfront and easier to modify, but wastes a runner on every shot and cycles slower. Hot runner eliminates the waste and the gate vestige and cycles faster, at a higher tool cost and more maintenance — the standard for production volume. Do you make molds and molded parts in-house?Nex-G sources injection molding through a vetted partner network and QA-manages it in-house — mold-design review, tool approval from a measured first article, and incoming inspection under the same ISO 9001 / IATF 16949 / ISO 14001 system and SPC discipline that runs the CNC floor. CNC and EMS are in-house, so molding, machining, and electronics integrate under one quality gate. Sourcing molded plastic parts for your build?Send the drawing, resin, and annual volume — we will review the mold design, confirm the tool cost and the break-even, and quote the molding with the machining and electronics under one quality gate.Request a quote Related articlesCNC Machining vs Injection MoldingPlastics Machining GuideCNC Machining Cost in China