Metal Injection Molded: When Small Metal Parts Need the Plastic Playbook
Most buyers meet metal injection molding by accident. They've got a stainless part that's tiny, oddly shaped, and needed in the tens of thousands, and the CNC quotes come back at a price that kills the project. That's usually the moment someone says "ever looked at MIM?" Metal injection molded parts start life as a plastic-like feedstock, get shaped in a mold, then turn into solid metal during sintering. You keep the cheap, fast tooling of injection molding, but you end up with something that machines and rusts like 316 or 17-4PH.
We've run MIM for medical tweezers, lock parts, and a stack of automotive sensor housings. The appeal was never the tech demo. It's that you reach geometries Swiss turning simply can't, at a unit cost that drops hard once the tool is paid for.
How the process actually runs
Skip the brochure version. The real flow is four messy steps. First, fine metal powder gets mixed with a polymer binder into a feedstock that flows like warm plastic. We press that into a steel cavity on a standard injection machine, nothing exotic there. The "green" part comes out at full size but it's mostly binder.
Next comes debinding, where most of that plastic is washed or burned out. Then sintering in a furnace around 1300 to 1400°C pulls the particles together. The metal moves, shrinks, and welds into a dense part. Plan on 15 to 18% linear shrink, which is the number that bites newcomers. You design the tool to grow, not the part.
Where it beats the obvious alternatives
Buyers almost always compare MIM against CNC and die casting first. They solve different problems, and the lines aren't where the sales decks draw them.
| Process | Best at | Weak spot | Typical part size |
|---|---|---|---|
| MIM | Complex small parts, 50k+ units | High tooling cost, big shrink | 0.1 to 200 g |
| CNC machining | Low volume, tight tolerances | Slow and pricey at scale | Any size |
| Die casting | Larger zinc or aluminum parts | Poor for tiny, thin features | 100 g to several kg |
For a 3-gram hinge with internal threads, CNC means five setups and scrap. Die casting can't hold the wall. MIM does it in one shot and the per-part number falls under a dollar past the first 100k. When volume is low, though, we'll point you to CNC Machining and save you the tool money.
What tolerances you can really hold
Sintered parts move, so nobody holds ±0.01 mm straight out of the furnace across a whole batch. You design to about ±0.3% of a dimension, then hit the tight spots with a light CNC pass. That hybrid is where the value hides.
| Property | Typical MIM result | Note |
|---|---|---|
| Density | 96 to 99% of wrought | Low porosity, can be sintered closed |
| Linear tolerance | ±0.3% as-sintered | Tight features need machining |
| Surface finish | Ra 1 to 4 µm | Better than die cast, worse than turned |
| Min wall | 0.3 to 0.5 mm | Depends on the material |
We once had a client whose drawing called for ±0.02 mm on a 40 mm sensor bracket. As-sintered we landed at ±0.12. Two milled locating faces fixed it, and the part still cost a third of Swiss turning. The lesson stuck: design the loose stuff for sinter, and reserve machining for the three faces that matter.
Materials worth considering
Low-alloy steel, 17-4PH, 316L, and 420 all run well. Tungsten-heavy alloys and some soft magnetics too. What you can't easily do is pure copper or anything that oxidizes badly in the sinter atmosphere. Those need special furnaces most shops don't keep on the floor.
For medical and food contact, 316L is the default. Automotive leans 17-4PH for strength. We've also done an Fe-Ni part for a watch brand that needed a specific thermal expansion. The binder system matters more than people think. A bad one leaves carbon behind and your "stainless" rusts in the salt spray test.
Design rules we hand every new client
Uniform walls. That's the big one. MIM shrink stays predictable only when the section is even, so a 2 mm boss next to a 0.4 mm rib will warp. Keep wall thickness in a 0.4 to 6 mm band and the tool behaves. Sharp inside corners are fine, unlike casting, MIM likes them. Don't expect a mirror surface without extra polishing, though.
Undercuts are possible with side actions, yet every extra slide adds cost and a failure point. If a feature can move to a light CNC op after sinter, we usually say do that. Good Precision Mold Making is mostly knowing what to leave out of the tool.
A real shop-floor story
A US importer came to us with a firearm-sight screw smaller than a grain of rice. Their old supplier machined it from bar stock at $1.40 each, and rejects ran 9% because the slot cracked under torque. We rebuilt it as a MIM part in 17-4PH, with the slot molded in. First article held torque to spec, and the reject rate dropped under 1%. Tooling was $11k, but at 400k units a year the part landed at $0.31. They paid the tool off in six weeks.
The catch? Their print had a 0.2 mm fillet we couldn't hold as-sintered. We opened it to 0.4 and nobody noticed. Buyers fixate on the drawing. What ships is what matters.
The cost picture, honestly
Tooling for MIM runs $8k to $30k depending on cavities and slides. That's the gate. Below roughly 20 to 30k units a year, you're often better off machining or using soft tooling. Above that, the curve bends hard in your favor. Material is cheap, cycle times are short, and labor per part is tiny.
Don't forget the hidden line: secondary ops. Sintering leaves a soft skin and maybe a parting-line feather. Plan for tumbling, a trim pass, or light CNC. When you're weighing it against Die Casting, the post-work is where the two processes actually meet. Our Mold Making Services quote always lists those steps separately, so the number you sign isn't a surprise later.
Bottom line
MIM earns its keep on small, complex, high-volume metal parts where machining bleeds money. Get the walls even, design the loose dims for sinter, and machine only what must be tight. If your volume is there, the tool pays for itself fast. If it isn't, we'll tell you to machine it instead.
About SHINY Mold
SHINY Mold has run precision molding and machining in Dongguan since 2003, across a 22,000 m² plant with 120+ engineers and 100+ injection machines, ISO certified. We take parts from DFM analysis through to mass production, backed by CNC, wire EDM, mirror EDM and CMM on every tool. If a part fits MIM, plastic, or CNC, we build it under one roof. See current work at shiny-mold.com.
