Die Casting vs Injection Molding: Metal or Plastic?

Die casting and injection molding are the same machine architecture wearing two uniforms. Both clamp a two-piece steel tool, inject a melt under high pressure, let it solidify, and eject a part seconds later. What changes is the melt: die casting forces molten metal alloy into a die, and injection molding forces molten thermoplastic into a mold. That one difference decides the strength, weight, heat behavior, tooling economics, and finishing work of everything the two processes produce, which is why experienced buyers settle metal versus plastic before they argue about any machine parameter.

The comparison below sticks to selection-relevant numbers: material properties, accuracy, tooling cost and life, per-part economics, surface finish, and the volume crossover where each route pays. It goes deeper than the manufacturing process comparison matrix, which tracks machining, cutting, printing, and forming but carries no row for either of these two volume workhorses.

Who each process is for

Die casting serves teams that need a metal part at volume: housings and enclosures that must shed heat or shield electronics, brackets and frames under sustained load, drivetrain and power hardware, fittings, and consumer-metal trim. The alloy list is short and non-ferrous: zinc in the Zamak family, aluminum such as A380 and ADC12, and magnesium AZ91D. Zinc and magnesium run hot-chamber, where the machine’s gooseneck sits in the melt; aluminum runs cold-chamber, ladled shot by shot, because molten aluminum would attack the gooseneck steel.

Injection molding serves teams that need a plastic part at volume: consumer and electronics housings, appliance and vehicle interior components, packaging and closures, medical disposables, and any part that wants electrical insulation, chemical immunity, color through the material, or living hinges, a flex-fatigue duty that polymers absorb and metals cannot. The catalog is broad: commodity resins such as PP and PE, engineering resins such as ABS, PC, PA, and POM, elastomers such as TPU, and filled grades of most of them.

The buyer profiles overlap almost completely on everything except material. Both processes want thousands of parts or more, both amortize expensive hardened-steel tooling across the run, both form complex three-dimensional geometry that would otherwise take many machining operations, and both see their tightest dimensions within one tool half and their loosest across the parting line. If your part could be built in either an alloy or a resin, the material section below decides most of the case. If it cannot, the choice has already been made.

The material question comes first

Neither process can reach into the other’s material catalog, so the first gate is functional: does the part need what metal gives, or what plastic gives? The property gaps between the two catalogs are wide enough to settle most decisions on their own, and they come with datasheet numbers attached.

Strength, stiffness, and weight

On representative published values, common die-cast alloys test at roughly 230 to 324 MPa tensile. AZ91D magnesium sits around 230 to 240, Zamak 3 zinc around 270 to 283 with about 208 MPa yield, and A380 aluminum around 324 with about 160 MPa yield and 3.5 percent elongation. Common unfilled molding resins test far lower on the ASTM D638 tensile method: PP at 25 to 40 MPa, ABS at 30 to 45, PC at 60 to 72, POM at 60 to 70, and PA66 nylon at 80 to 90. That is roughly a 3 to 8x strength gap across typical alloy-and-resin pairings in favor of metal, and stiffness separates further: die-cast alloys run about 45 to 96 GPa across the alloy families in elastic modulus against roughly 1 to 3 GPa for unfilled resins, more than an order of magnitude apart.

One plastic closes part of the gap. For example, a 30 percent glass-filled nylon reaches roughly 210 MPa, close enough that engineers now specify it for housings and brackets that were cast aluminum a generation ago. That substitution has limits, covered below, but it is real and it is why the two processes compete for work more often than their material tables suggest. The engineering plastics page carries the filled-grade detail.

Weight cuts the other way, decisively. Resins sit at 0.9 to 1.4 g/cm3 against 1.81 (AZ91D) to 6.6 (Zamak) for the casting alloys. The same geometry in A380 aluminum at 2.71 g/cm3 carries about 2.6x the mass of ABS and about 3x that of PP. A design that counts grams and sees modest loads is a plastic part before any cost analysis starts. Where loads still demand metal, magnesium and aluminum keep the weight penalty small, while zinc’s 6.6 g/cm3 argues against it for anything a person carries.

Heat, EMI, and the chemical environment

Temperature and electrical behavior finish the material case. Die-cast aluminum conducts roughly 96 W/mK and spreads heat into its own walls, so a cast enclosure doubles as a heat sink. Plastics conduct about 0.1 to 0.4 W/mK and insulate instead. On service temperature, common resins deflect at 80 to 150°C, with ABS around 95°C, PA66 around 80°C under load, and PC around 130 to 140°C, while aluminum castings serve continuously past 150 to 200°C, losing strength rapidly only above that band.

EMI follows the same split. A die-cast aluminum enclosure blocks radio-frequency interference by existing; it is a conductive box. A plastic enclosure is transparent to RF and needs conductive coatings, internal shield cans, or metal inserts to pass emissions testing, and that add-on content erodes the plastic cost advantage in electronics work.

Chemistry favors plastic. Resins are immune to galvanic corrosion and most chemicals without any coating, hold color throughout the material rather than on it, and damp vibration. Aluminum die castings are naturally corrosion-resistant and usually powder coated or chromated where appearance matters, while zinc takes plating so readily it owns the decorative-fitting niche. Plastics carry two caveats metals do not: nylon absorbs water, up to about 9 percent by weight for PA6 and around 1 percent for PA12, which swells dimensions, and every polymer creeps under sustained load. The aluminum page and the material properties database hold the per-grade numbers behind these bands.

Accuracy: overlapping bands, different failure modes

Many comparison tables hand the accuracy contest to one process or the other. The verified numbers do not support a winner. Both processes hold roughly ±0.05 to 0.3mm as-produced, and both are governed by the same geometry rule: dimensions formed within one tool half hold tightest, and dimensions that cross the parting line or move with a core run looser because they inherit tool-closure variation. Published hub ranges sit side by side, about ±0.05 to 0.25mm as-cast for die casting and about ±0.1 to 0.3mm typical for molding, with resin choice moving the molded number because shrink of 0.5 to 2.5 percent has to be compensated in the cavity steel.

Where tighter precision is required, the two routes diverge in method rather than in ceiling. A precision mold holds ±0.02 to 0.05mm on tight features by cutting the cavity accordingly. A die casting gets there by leaving machining stock on those same features and machining threads, bores, and mating faces after casting. Both land in machining territory for their tightest work, and the CNC tolerances reference table covers what that next step holds.

Repeatability is a shared strength. Once the tool is tuned, every part comes from the same cavity, so part-to-part variation is low in both processes, and the realistic accuracy question is not which process is tighter but which defect profile your inspection plan must catch. Die castings carry gas and shrink porosity, which is why standard high-pressure castings are not heat treated or welded. Molded parts carry sink marks, warp, and weld lines instead, plus moisture-driven dimensional drift in hygroscopic resins. Neither defect family is fatal; both belong in the part’s quality plan.

Tooling: two hardened steels, two different lives

Both processes are tooling-amortized, and for a like-for-like part the tooling-cost relationship is consistent: a single-cavity die casting die usually costs more than the equivalent plastic injection mold. The die is typically H13-class hot-work steel built to survive molten aluminum at 620 to 680°C cycling against its face, versus mold steel holding resin melt at 180 to 320°C. The relationship can invert, though: a complex multi-cavity production mold with hot runners and Class 101 construction can out-cost a die, because cavitation, not process family, is the biggest tooling-cost driver on both sides. This page deliberately stays at that relative level; tooling prices belong in a quotation, not a comparison.

Tool life, by contrast, publishes in shots and the numbers are instructive. Zinc dies last roughly 500,000 to more than 1 million shots because zinc pours near 385°C and treats the steel gently. Aluminum dies last about 50,000 to 150,000 shots as the hot metal erodes and solders to the die face, and magnesium dies land roughly 100,000 to 200,000 shots. On the molding side, hardened Class 101 production molds also run past 1 million cycles, with aluminum prototype molds around 100,000. Read as a pair: process family does not decide tool life, melt temperature does.

Lead time to first article is weeks on both routes, set mainly by tool-build complexity and cavitation rather than by process. The practical difference lands at launch: a die casting program brings more of the program’s timeline with it, because the trim die, machining fixtures, and impregnation or leak-test stations are usually part of the production package, while a molding program tunes the tool and then ships near-net parts. Schedule both as tooling projects, not as material decisions.

Per-part economics and where the crossover sits

Both processes buy their low per-part cost the same way: pay for steel once, then run thousands of shots through it. The per-part differences stack up from four places.

First, material mass. Raw material prices per kilogram are broadly comparable between commodity resins and casting alloys, so the per-part material cost gap comes from density, not price: the metal part simply contains two to three times more material by mass in the same geometry. Second, melt energy: pouring aluminum at 620 to 680°C costs more energy than melting resin at 180 to 320°C. Third, cycle time, where the common wisdom needs correcting. Die casting is not uniformly faster. Zinc hot-chamber work runs seconds per shot, around 15 cycles a minute, faster than typical molding. Aluminum cold-chamber work runs 15 to 90 seconds, comparable to or slower than molding’s 15 to 60 seconds, and molding’s cycle is dominated by cooling that scales with roughly the square of wall thickness. Fourth, secondary operations: every casting is trimmed and deburred, and critical features are frequently machined, while molded parts more often ship as-molded.

The volume crossover follows. Published break-even guidance puts injection molding ahead at lower volume, commonly from about 1,000 to 10,000 parts for simple work and 10,000 to 50,000 for complex parts, with die casting usually quoted from about 10,000 to 50,000 per year. For example, in one published engineering case, a 280-gram housing at 50,000 per year came out meaningfully cheaper per part molded in glass-filled ABS than cast in aluminum, and the gap traced to mass and secondary operations rather than to material price. Treat that as one worked case, not a rule: a zinc part on a hot-chamber cell at very high volume, or a multi-cavity mold that costs more than the die it replaces, moves the arithmetic. The manufacturing cost comparison by process and low-volume manufacturing pages carry the volume-side framework, and the injection molding vs 3D printing bridge covers the plastic process’s lower-volume crossover.

Surface finish and the work after the press

This is where the processes diverge hardest in daily practice, and the divergence belongs in any honest per-part cost comparison.

Injection molding produces its finish inside the tool. A polished cavity yields a glossy surface, an etched cavity yields texture, and pigment yields color through the material, so every part in the run is identically finished with zero per-part labor and no paint line to mask or maintain. Post-press work is light: degating is automatic or designed out with hot runners, inserts are installed by heat or ultrasonic welding where needed, and machining is occasional rather than routine.

Die casting always finishes outside the press. The trim die shears runners, gates, and flash under press force, then parts are tumbled or deburred. Threads, tight bores, and flat mating faces are machined where the design requires them, castings that must hold fluid are impregnated to seal porosity and leak-tested, and appearance parts move to powder coat, paint, chromate, or plating, which zinc takes especially well. On a conventional cold-runner shot, only about two-thirds of each shot becomes product; the gates, runners, and flash are trimmed off and re-melted, where a hot-runner mold can eliminate runner scrap entirely. The surface finishing page maps the coating options both process families feed into.

Net: molding ships closer to finished, and casting carries a standard secondary chain. Buyers comparing per-part cost across the material divide should price the secondaries, not just the shot.

Four questions that pick the process

Material needs. Does the part require metal properties, roughly 3 to 8x tensile strength across typical pairings, an order of magnitude more stiffness, heat past about 150°C, heat spreading, or inherent EMI shielding? Then die casting, and the alloy choice follows from weight and corrosion needs. Does it require insulation, chemical immunity, molded-in color, or living hinges? Then injection molding. If glass-filled nylon can carry the loads, molding can stand in for a cast housing until heat, EMI, or threaded metal-duty fasteners send it back.

Weight target. If every gram counts and loads are modest, plastic wins by physics: 0.9 to 1.4 g/cm3 against 1.8 to 6.6. If lightness must coexist with structural duty, magnesium or aluminum casting is the route.

Volume. Both processes need thousands of parts to pay for tooling. Molding breaks even from roughly 1,000 to 10,000 simple parts; die casting is usually quoted from 10,000 to 50,000 per year. Within high volume, remember the cycle correction: zinc hot-chamber cells are among the fastest near-net production anywhere, while aluminum cold-chamber work does not beat molding on speed and must win on material capability.

Finish and secondary budget. If the part wants cosmetic surface straight out of the press, molding delivers it in the cavity. If the part wants machined precision features anyway, die casting’s as-cast-plus-machining pattern absorbs that naturally. Size matters too: die casting is strongest on small-to-medium complex metal parts, with very large structural parts moving to sand or investment casting, while molding reaches very large panels on presses up to 9,000 tons, limited per resin by flow length rather than tonnage.

When the answer is neither

Below the crossover, both routes lose to processes without tooling. A metal part needed in the tens or hundreds is usually machined from CNC stock, buying precision and lead time at a higher per-part cost but no tool investment. A plastic part still under development belongs on a 3D printing bed while the design moves, which is the low-volume route the rapid prototyping page develops.

Two adjacent processes fill specific gaps. Metal injection molding serves small, very complex parts, roughly 0.1 to 100 grams, in stainless steel or titanium that die casting cannot pour; its molds live around 150,000 to 300,000 shots, and die casting is cited as the cheaper route, by up to about 30 percent in published comparisons, where the alloy allows it and die life runs into the millions. Squeeze casting exists specifically to remove high-pressure porosity, producing aluminum parts that can be heat treated and welded, which standard die castings cannot.

And sometimes the answer is the other process. The glass-filled-nylon substitution runs both directions: designs migrate to plastic for weight and cost and back to metal for heat, shielding, and threaded strength. The shared vocabulary the two processes trade in is collected in the manufacturing glossary, and the two hub pages on either side of this comparison carry the full process depth.