Die Casting vs Injection Molding: Which Process Fits Your Part?

Die casting and injection molding are not competitors so much as two
different answers to the same question: how do you make this shape out of metal or
plastic, thousands of times, to a cost you can live with? Getting the answer wrong costs
money twice — once in tooling you cannot use, and again when the parts that arrive do
not behave the way the design assumed. This guide compares the two processes on the
criteria that actually decide it: part weight and stiffness, wall thickness, tolerance,
surface finish, tooling cost, lead time and volume.

Short answer

Choose die casting when the part is predominantly metal, weighs more
than roughly 200 grams, needs real stiffness without becoming heavy, and has a geometry
that is simple to fill and release — typical examples are automotive brackets, gearbox
housings, heat sinks and structural castings. Choose injection molding
when the part is a thin-walled polymer component, needs a cosmetic or Class A surface,
requires tight tolerances, needs flame-retardant or medical-grade certification, or the
production volume is below a few thousand pieces.

The single most useful rule is this: compare parts, not processes.
Take the drawing and put the two candidate parts side by side, including wall thickness
and material. A die cast aluminium bracket that is 400 g will usually replace a glass-
filled nylon bracket that is 600 g; in that case die casting wins on function and cost
simultaneously. But if the candidate part is a 300 g ABS housing with a 1.5 mm wall and
a visible surface, injection molding is the only sensible answer, because no die casting
process produces that combination.

What actually differs between the processes

Both processes fill a cavity with molten material and let it solidify, but the
material, the pressure and the tooling are different enough that the two industries
share very little equipment, suppliers or process know-how.

That last row is worth pausing on, because it is counter-intuitive. A die casting die
is more expensive to build than a simple injection tool, but it is not built from a
softer, cheaper steel — it has to be H13, hardened, because each shot dumps a large
volume of molten metal into a cavity that has to survive the thermal shock indefinitely.
Injection tooling can often start in pre-hardened P20 and stay there if the part is
simple and the volume is modest.

Weight and stiffness: the dividing line

The single biggest technical difference between the two processes is material
density. Aluminium is roughly 2.7 times denser than glass-filled polyamide and roughly
2.9 times denser than ABS. That sounds like an argument against metal, but it works in
the opposite direction once you account for stiffness, because stiffness scales with
section depth in a way that is brutally nonlinear.

A short beam’s deflection falls with the cube of its thickness. So a wall that needs
to be 3 mm thick in glass-filled nylon to hold a 500 g load can often be 1.8 mm thick in
die cast aluminium and hold the same load, because aluminium’s modulus is roughly three
times higher. The part gets lighter and stiffer at the
same time. This is the reason automotive structural parts moved to die casting and never
came back.

The reverse case is just as common. If the part is a thin cover, a bezel, a housing
with clips or a panel with a visible face, metal is the wrong material on weight and
wall control alone, and injection molding wins regardless of the unit price quoted for
the metal.

Wall thickness and geometric freedom

Injection molding can hold a 0.8 mm wall across a 300 mm part because the cavity is
steel, the gate is small, and the process is fast enough to freeze the flow before the
part can distort. Die casting cannot do that, because molten metal does not behave like
a viscous polymer under a fast gate. In a conventional die casting, the metal has to be
allowed to fill and the part has to stay hot long enough to avoid cold shuts, so
minimum walls are set by the metal’s fluidity and the machine’s gating design.

This single constraint propagates into almost every design decision downstream.
Thin webs, long unsupported walls, tall thin ribs and snap-fit arms that rely on
polymer elasticity are all natural in injection molding and awkward in die casting.
Conversely, die casting handles things polymer molding cannot: heavy bosses, integral
thread forms, and thin-walled housings with integrated metal inserts.

Two further differences matter in real projects. First, integrated inserts.
A metal insert can be placed in the die and become part of the casting, which is how
electrical contacts and threaded metal features get into an aluminium part without a
secondary assembly step. Polymer molding cannot do this without overmolding, which is a
different and more expensive process. Second, internal passages: die
casting can form cored passages and hollow sections in metal with a reasonable price,
because the core is metal and the casting is large enough to amortise it.

If the part has snap fits, thin living hinges or a wall thinner than 1.2 mm, die
casting is not the right process and it is worth saying so early, because changing the
process after the design is frozen means redoing the tooling.

Tolerance, surface finish and machining

Here the two processes separate cleanly, and the separation is usually decisive.

The practical consequence is that die casting is a semi-finished process.
The raw casting has a texture, a gate boss and tolerances in the range of plus or minus
0.1 mm. Where a feature has to be a sealing face, a bearing surface or a datum, it is
machined after casting, and the design has to leave material for that machining. The
most common die casting design error is treating the casting as a finished part and then
finding that there is nowhere to put a cutting tool.

Injection molding produces a finished part directly from the cavity. That is why it
remains the default for anything with a consumer-facing surface, and why die casting
parts that need a good finish are usually specified with a machining operation attached
rather than as a bare casting. Our CNC machining
capability exists partly for this reason — the same facility that casts the part
finishes it, so the tolerance stack across both operations is owned by one team.

Tooling cost, unit cost and the crossover

Both processes have the same economic shape: a large one-off tooling cost followed by
a falling unit cost. The difference is the size of the first number and the steepness
of the second.

Die casting tooling is materially more expensive than a comparable
injection tool. The die has to be substantially larger, made from H13, fully hardened,
and fitted with a proper gating and venting system. A simple aluminium die typically
costs noticeably more than a simple single-cavity injection tool of similar part size,
and the gap widens for multi-part tools. That upfront cost is what makes the volume
crossover the central question of any die casting project.

The unit cost curves cross over because the material saving is large.
For a 400 g aluminium casting, the metal itself is a meaningful part of the cost, but
for a 600 g glass-filled nylon part the resin is more expensive per kilo than the
aluminium and there is more of it. The die casting part typically arrives with a lower
cost per piece once the tooling is paid off, because you are shipping less material
volume for the same stiffness.

As a rough planning guide for a simple part with a single cavity and no machining:

Two corrections to that table, because they are the two most common ways die casting
projects are mis-specified. First, it assumes a single cavity. If the part is small
enough to run in a multi-cavity die, the tool cost per cavity drops sharply and the
crossover moves to a lower volume. Second, it assumes no secondary machining. If
critical features need CNC finishing, add that cost to the die casting route before
comparing — for a part with several sealing faces, machining can add a meaningful
per-piece cost and can tip the decision back to injection molding with a machined
insert.

For what tooling actually costs and where the money goes, the
injection molding cost breakdown covers the
structure of a tooling quotation in detail, and it applies closely enough to die casting
to be useful when you are comparing quotes.

Lead time and volume planning

Lead time behaves differently in the two processes, and in a way that catches projects
out. A die casting die is physically larger, has to be machined from larger blocks,
takes longer to harden, and needs a proper trial with gating and venting corrections.
A simple aluminium die typically takes materially longer than a simple injection tool of
comparable part size — and that is before the first-article iterations that every die
casting tool needs, because the fill behaviour of molten metal is harder to predict from
simulation than the flow of a polymer.

Injection tooling can be brought in faster, particularly in the fast lane: a bridge
tool in aluminium or a pre-hardened steel resin tool can be delivered and running
parts in a small fraction of the time. Our rapid tooling
page describes what is achievable on a compressed schedule, and the
mold lead time guide sets out the full
range by tool type.

For the trial itself, the same principle from the
mold flow analysis page applies: simulate
before cutting. A die casting simulation models fill, solidification and shrinkage, and
it will show you where the cold shut and the shrinkage porosity will appear before the
first shot. That is the difference between a first trial that produces saleable parts
and a first trial that produces scrap.

Full comparison table

Everything above, consolidated for quick reference and for sharing with a purchasing
team.

10 to 35 days depending on tool type

How to choose for your part

Run through these questions in order. The answer to each one eliminates options, and
usually by question four or five the process is decided.

1. Is the part metal or polymer in the final assembly? If metal, and
   the weight is above roughly 200 g, die casting. If polymer, and the walls are thin,
   injection molding.
2. Does the part carry load or heat? Aluminium’s stiffness and thermal
   conductivity are not substitutable with a polymer. A heat sink or a structural bracket
   belongs in die casting regardless of what the plastic version would cost.
3. Does the part have a visible surface or a tight tolerance? If yes,
   injection molding, or die casting with a specified finishing and machining operation
   baked into the design.
4. What is the volume? Under about 1,000 parts per year, use rapid
   tooling or CNC machining for either material. Between 1,000 and 10,000, run the
   comparison on the actual numbers including machining.
5. What can the part tolerate geometrically? Wall thickness, snap fits,
   thin webs and living hinges all point to injection molding. Heavy bosses, integral
   metal inserts and cored passages point to die casting.
6. Is one part family going across both processes? If so, the
   interface between the die cast body and the moulded cover has to be designed as a
   system, with the datum strategy and the tolerance stack agreed before either tool is
   cut.

Our die casting page covers the alloys, the machine
range and the tooling we build in-house, and the
multi-shot molding page covers the cases where
two materials have to be combined in a single tool cycle. If the part sits between the
two processes and you are not sure which way it should go, send the drawing and the
annual volume — that is enough information for an engineering answer rather than a
quotation guess.

Three real process decisions

The comparison above is general. These three are the situations our engineering team
actually argues about, and each one is a case where the answer was not obvious from the
drawing alone.

Scenario 1 — The 480 g gearbox housing

An automotive customer needed a gearbox housing weighing roughly 500 g, with a
bearing bore and two sealing faces, produced at 60,000 pieces a year. The incumbent
solution was a glass-filled nylon housing at 780 g, which met the load requirement but
added 300 g to every vehicle it went into.

The die cast alternative was an ADC12 aluminium housing at 480 g. The stiffness was
higher, the weight was 38 percent lower, and the material cost per piece fell despite
aluminium being the more expensive metal. What changed the plan was not the casting but
the machining: the bore and the two sealing faces were moved to CNC-machined datums,
with the as-cast surfaces deliberately left untouched to avoid cutting into a
porosity-prone skin. The design decision that made it work was
specifying which faces were datums and accepting the as-cast texture everywhere else,
rather than trying to machine the whole housing.

Scenario 2 — The 45 g electronic enclosure cover

A consumer electronics customer wanted a cover with a textured Class A surface,
tight clips, a 1.0 mm wall and a flame-retardant requirement. The volume was 120,000 a
year, which sounds like die casting territory.

It was not. The 1.0 mm wall was below what conventional die casting fills reliably,
the clips needed polymer elasticity to work, and the flame-retardant certification
attached to the resin grade rather than the alloy. An aluminium version would have
needed 2.5 mm walls to achieve the same stiffness, would have weighed more than the
nylon version, and would have needed a secondary finishing operation to reach the
required surface. This was injection molded in PC/ABS with a texture applied to the
cavity steel. The useful principle is that a cosmetic, thin-walled,
certified part should be moulded in polymer even when metal is available.

Scenario 3 — The 900 g heat sink that had to be machined

An industrial customer needed a heat sink for a power supply: aluminium, 900 g,
extensive fins, and a mounting face with a flatness requirement of 0.05 mm. The fins were
the argument for die casting, and they were a good one — thin fins in a machined block
would have been uneconomical to produce.

The flatness requirement was the argument against it, because an as-cast 900 g
aluminium part will not hold 0.05 mm flatness across a mounting face. The resolution
was to cast the finned body and machine only the mounting face and the bore, keeping the
part as a rigid assembly so the casting did not distort during the machining operation.
Cost was higher than an entirely machined block and far lower than a moulded plastic
alternative, which would have needed a metal insert and a two-part assembly. This is the
pattern most heat sink and bracket projects end up in, and it is worth designing the
datum scheme before the die is cut rather than after the first
casting arrives.

Process and tooling terms quick reference