Metal Injection Molding
Metal injection molding molds a powder-and-binder feedstock exactly like a plastic, then debinds and sinters it into a dense small metal part.
- Part
- Forming
- Revised
- 2026-08-11
At a glance
- Family
- Metal
- Typical tolerances
- About ±0.3% of the dimension as standard (roughly ±0.003 in on a 1 in feature), with ±0.1–0.2% achievable on controlled features or after sizing; machine anything tighter
- Surface finish
- Roughly 32 µin Ra (0.8 µm) as sintered, reflecting the mold finish
- Typical volumes
- 20,000–1,000,000+ parts per year; below about 5,000 machining is usually cheaper
- Lead time
- 8–14 weeks for tooling and first articles; 4–8 weeks per production run thereafter, set largely by furnace scheduling
- Materials
- Metal
What it is
Metal injection molding (MIM) blends fine metal powder with a thermoplastic binder into a feedstock that is molded on a conventional injection molding machine, then chemically and thermally strips the binder away and sinters the remaining powder skeleton into a dense metal part. The molded "green" part shrinks 15–20% linearly during sintering — uniformly, and by an amount the toolmaker builds into the cavity.
The result is a small, geometrically complex metal part at injection-molding economics: sintered density typically 96–99% of theoretical, so mechanical properties approach those of wrought material, unlike conventional press-and-sinter powder metallurgy.
MIM's sweet spot is parts under about 100 g with sections under 0.25 in (6 mm) — surgical instrument jaws, firearm components, orthodontic brackets, watch cases, connector shells, and phone hinges. Standard dimensional tolerance is roughly ±0.3% of the dimension. Below about 5,000–20,000 parts a year, machining usually wins.
How it works
- Feedstock. Metal powder finer than about 20 µm is compounded with a multi-component binder — typically wax or polymer blends — to roughly 55–65% powder by volume. The powder must be spherical and fine, which is why MIM material costs far more per pound than bar stock.
- Molding. The feedstock is injected into a steel tool on a standard injection molding machine. The green part looks like the finished part but oversized, and it is fragile — it is essentially powder held together by wax.
- Debinding. The primary binder is removed by solvent immersion, catalytic decomposition, or thermal means, leaving an open pore network and a secondary backbone binder holding the shape. This is the slowest step and its rate is governed by section thickness, which is exactly why MIM has a wall thickness ceiling.
- Sintering. The brown part is fired in a controlled atmosphere or vacuum furnace below the melting point — roughly 2,300–2,450°F (1,260–1,340°C) for stainless steels. The residual binder burns out and the powder particles fuse and densify, with the part shrinking 15–20% in every direction.
- Secondary. Optional sizing, machining of critical features, heat treatment (17-4PH is commonly solution treated and aged), and finishing. See the surface finish chart for achievable finishes.
The whole process depends on the shrinkage being isotropic and repeatable. Part-to-part variation comes from density gradients in the green part — the same molding defects that cause sink and warp in plastic — which is why MIM design rules track plastic injection molding rules closely. The injection molding design guidelines apply almost directly, and the metal melting points chart shows why sintering happens well below the melt.
Design guidelines
Wall thickness: uniform and thin
Nominal walls of 0.040–0.160 in (1–4 mm) are ideal, with a practical ceiling around 0.25 in (6 mm). Thick sections are not a strength problem, they are a debinding problem — the binder has to diffuse out of the center, and time scales steeply with thickness. Core out heavy sections just as you would in plastic.
Part size and mass
MIM is economic on small parts. Typical parts fall between about 0.1 g and 100 g; above that, feedstock cost and furnace capacity make it uncompetitive against machining or investment casting.
Draft
0.5–2° of draft is normal and makes ejection reliable. Zero draft is achievable on short walls with a good tool finish, since sintering shrinkage pulls the part away from the steel, but it should be a deliberate decision rather than an oversight.
Ribs, bosses, and corners
Follow injection molding practice: ribs at roughly 0.6× nominal wall to avoid sink, radiused internal corners, and cored bosses rather than solid ones. Sharp internal corners crack the green part during handling and debinding, before the metal ever exists.
Undercuts, threads, and holes
Side actions and slides work the same as in plastic tooling. Cross holes, external threads, and internal detail are all moldable. Small holes down to about 0.010 in (0.25 mm) can be molded, but blind holes deeper than roughly 5× diameter risk pin damage. Threads are moldable but often finished by machining when the class of fit matters.
Support during sintering
The part is soft at sintering temperature and sags under its own weight. Long unsupported spans, thin overhanging flanges, and tall thin walls distort. Flatness and straightness on a MIM part are set by the sintering setter, not the mold, so give the supplier flat surfaces to rest on.
Tolerances
Standard tolerance is about ±0.3% of a dimension, with ±0.1–0.2% achievable on well-controlled features or with post-sinter sizing. On a 1.000 in dimension, that is roughly ±0.003 in standard. Anything tighter — a bearing bore or a sealing face — should be machined after sintering.
| Feature | Recommended | Limit | Why |
|---|---|---|---|
| Nominal wall | 0.040–0.160 in (1–4 mm) | 0.25 in (6 mm) | Debinding time rises steeply with section |
| Part mass | 0.1–60 g | ~100 g | Feedstock cost and furnace capacity |
| Draft | 0.5–2° | 0° on short walls | Sintering shrinkage aids release |
| Rib thickness | 0.6× nominal wall | 0.8× wall | Thick ribs sink and create density gradients |
| Molded hole diameter | ≥ 0.020 in (0.5 mm) | 0.010 in (0.25 mm) | Fine core pins survive only short depths |
| Blind hole depth | ≤ 3× diameter | 5× diameter | Long core pins deflect in the feedstock |
| Linear tolerance | ±0.3% of dimension | ±0.1–0.2% with control | Shrinkage repeatability governs everything |
| Flatness | Provide a flat setter face | — | Parts sag on their supports during sintering |
Cost drivers
Three cost centers: the tool, the feedstock, and the furnace. The mold is a conventional injection molding tool sized up for shrinkage, so its cost is comparable to a plastic tool of the same complexity — modest against a die-casting die. Feedstock is expensive because MIM requires fine spherical powder, so material cost per part is high relative to bar stock. Furnace time is a batch operation, and long debinding and sintering cycles mean throughput, not cycle time, sets capacity.
Volume breakpoints: below roughly 5,000 parts a year, machining or 3D printing beats MIM because the tool cannot amortize. From 20,000 to a few hundred thousand is the classic MIM window. Above a million, the process is fully in its element, with multi-cavity tooling driving molding cost toward negligible and furnace throughput becoming the constraint.
- Consolidate an assembly into one part. MIM's biggest wins come from replacing three machined components and a welding step with one molded part.
- Thin the walls. Thinner sections debind faster, use less expensive powder, and cost less per part in every dimension.
- Avoid post-machining. Every machined feature adds a fixturing operation on a small part that is awkward to hold. Tolerance to ±0.3% wherever function allows.
- Use multi-cavity tooling. Molding is the cheap step; more cavities spread it further.
- Pick a common alloy. 316L and 17-4PH have the deepest supply base and the best-characterized shrinkage; exotic alloys carry both a material and a process-development premium.
Questions
6 questionsHow much does a MIM part shrink during sintering?
15–20% linearly, in every direction. The mold cavity is cut oversize by exactly that factor, which is why shrinkage repeatability — not molding precision — is what sets MIM's dimensional tolerance of roughly ±0.3%.
What is the maximum wall thickness for metal injection molding?
Around 0.25 in (6 mm), with 0.040–0.160 in (1–4 mm) ideal. The limit comes from debinding: the binder has to diffuse out of the section, and the time required rises steeply with thickness. Core out heavy sections exactly as you would in plastic injection molding.
How dense are MIM parts?
Typically 96–99% of theoretical density, which is why mechanical properties approach wrought material. This is the main distinction from conventional press-and-sinter powder metallurgy, which usually leaves 10–15% porosity and correspondingly lower strength and fatigue performance.
What materials can be metal injection molded?
316L and 17-4PH stainless are the highest-volume grades, along with low-alloy steels such as 4605, soft magnetic iron-nickel alloys, tool steels, titanium, and tungsten heavy alloys. Alloy availability follows powder availability, and fine spherical powder is what drives the material cost.
What volume justifies MIM over machining?
Roughly 5,000–20,000 parts a year as a crossover, depending on how much machining the part would otherwise need. MIM's advantage grows with geometric complexity — a part needing five machining setups crosses over far sooner than a simple turned one.
MIM or investment casting?
MIM wins below about 100 g, where its molded detail, 0.040 in walls, and 32 µin Ra surfaces beat what a ceramic shell can deliver. Investment casting wins above that mass and for larger sections, since MIM's debinding limit caps section thickness near 0.25 in regardless of part size.