The wrong process choice rarely fails in the first build. It fails at the first price review, when a part that made sense at 300 pieces is still on a 5-axis mill at 30,000, or when a MIM tool was cut for a part that sold 2,000 units and was redesigned twice.
We make MIM parts, so this guide has an obvious bias to declare. It is also written by people who turn down MIM work every month because the part belongs on a mill or in a casting foundry. The aim here is a selection method you can apply without us: five questions, asked in order, and the numbers behind each one.
The three processes side by side
| Factor | CNC machining | Investment casting | MIM |
|---|---|---|---|
| Practical part mass | Any | ~5 g to 50 kg and up | ~0.1 g to 100 g, best under 50 g |
| Minimum wall | ~0.5 mm, depends on stiffness | ~1.5 to 2 mm (about 1 mm on small parts) | ~0.3 to 0.5 mm locally; 1 to 6 mm nominal |
| As-made tolerance | ±0.01 to ±0.05 mm | About ±0.5% to ±1% (ISO 8062 CT4 to CT6) | ±0.3% to ±0.5% as-sintered |
| As-made surface | Ra 0.8 to 3.2 µm | Ra 3.2 to 6.3 µm | Ra 1.6 to 3.2 µm |
| Materials | Almost any machinable alloy, including aluminum | Wide: steels, stainless, Ni and Co alloys, aluminum, bronze | Steels, stainless, Ti, Cu, Fe-Ni, W alloys; not aluminum |
| Density and properties | Wrought, 100% dense | Cast; possible shrinkage porosity in thick sections | 96% to 98% relative density; properties per MPIF Standard 35 |
| Tooling | Fixtures only | Wax-injection die, often $2,000 to $15,000 | Hardened steel mold, roughly high four figures to $60,000 |
| Tooling lead time | Days | ~4 to 6 weeks | Prototype tool: 4 to 5 weeks, then samples 10 to 15 working days; production tool: lead time confirmed after sample approval |
| Piece price vs volume | Nearly flat | Falls moderately | Falls steeply after tooling payback |
| Typical volume range | 1 to ~5,000 | ~500 to 50,000 | ~5,000 to millions |
| Cost of a design change | Reprogram | Modify wax die | Steel change or new insert |
These are working ranges, not limits. A very simple turned part can stay on a lathe at 100,000 pieces. A complex MIM part can pay back at 3,000. The table tells you where to start looking.
Five questions, in this order
1. What is the material?
If the part must be aluminum, MIM is out; aluminum powder carries an oxide film that prevents normal sintering. The choice is between machining, die casting and investment casting. If the alloy is a steel, stainless, titanium, copper, Fe-Ni or tungsten alloy, all three processes stay on the table.
2. How heavy and how thick is it?
Above about 100 to 150 g, MIM feedstock cost and debinding time work against it, and sections thicker than 8 to 10 mm risk cracks and porosity. Investment casting handles mass and thickness easily. Below about 20 g, casting struggles with fine features and gating yield, and MIM or machining take over.
3. What is the lifetime volume?
Tooling has to pay back. As a rough guide, below 2,000 to 5,000 pieces over the life of the part, CNC is usually cheapest even for complex shapes. From a few thousand to tens of thousands, investment casting and MIM compete, and geometry decides. Above that, if the part is small, MIM usually wins on piece price. The exact crossover depends on how expensive the part is to machine. A part with three setups and small internal features breaks even early; a simple turned pin may never break even.
4. How complex is the geometry, and in which directions?
Complexity costs money on a mill. Every setup, small tool and internal feature adds time. Investment casting and MIM both carry complexity in the tooling, so a complex part costs about the same per piece as a simple one of the same mass. The difference is scale: MIM reproduces fine features such as 0.3 mm slots, knurls, small threads and logos that casting cannot hold.
If the part is prismatic and can be pressed along one axis, such as a gear, a bushing, a cam or a flange, also price conventional press-and-sinter powder metallurgy. It is cheaper than MIM for those shapes, at a lower density (about 6.6 to 7.2 g/cm³ for iron-based grades).
5. How many features need tighter than about ±0.3%?
This is where many MIM projects go wrong. As-sintered MIM holds ±0.3% to ±0.5% because every part shrinks 15% to 20% linearly in the furnace. If one or two features need ±0.01 mm, a molded part plus one reaming or machining step is normal and still cheap. If most features need machining tolerances, you are buying a machined part with a MIM tool attached. Investment casting has the same problem at a coarser level, which is why most precision castings are machined on several faces.
Hybrids are often the right answer
The best route is often two processes in sequence:
- CNC first, MIM later. Machine the first few hundred pieces while the design settles, then move to MIM once it is frozen and volume is confirmed. A prototype MIM tool can bridge the gap and give real sintered samples 10 to 15 working days after the tool is ready (the tool takes 4 to 5 weeks).
- MIM plus one machining step. Mold the shape, then ream the one bore or face the ±0.01 mm callout lives on.
- Investment casting plus machining for parts above the MIM size range that still need precision interfaces.
Five example parts
| Part | Recommended route | Why |
|---|---|---|
| 3 g stainless watch-case lug, 50,000 per year | MIM | Small, fine 3D features, cosmetic surface, high volume |
| 250 g stainless pump housing, 3,000 per year | Investment casting plus machining | Too heavy for MIM; casting carries the shape, machining the seals |
| 40 test fixtures, aluminum | CNC | Aluminum, tiny volume |
| 15 g iron gear with hub, 200,000 per year | Press-and-sinter PM | Prismatic, one pressing axis, lowest cost per part |
| 12 g 17-4PH lock cam with cross hole, 20,000 per year | MIM, with one reaming step | Complex, small, volume pays back the tool; one bore needs ±0.01 mm |
Where each process gives you trouble
- CNC: cost stays flat with volume; burrs on internal intersections; material waste on complex parts; long cycle times in hard alloys.
- Investment casting: shrinkage porosity in thick-to-thin transitions; surface Ra and tolerance usually need machining on functional faces; fine features below about 1 mm are hard to hold.
- MIM: high tooling cost and limited room for design change after the tool is cut; size and mass limits; ductility below wrought in some grades (MIM 17-4PH in H900 has single-digit elongation per MPIF Standard 35); yield depends heavily on wall uniformity and how the part sits in the furnace. We do not promise one yield figure for every part. The quote states the yield assumption after the DFM review, and parts with features usually need sizing or other secondary operations after sintering.
Before you choose
Fill in the five answers: material, mass and maximum section thickness, lifetime volume, the features that drive complexity, and the number of features tighter than ±0.3%. Then send the STEP file with those answers and your current price, if you have one. Within 7 working days you get a free DFM review that says which process we would choose for the part, MIM or not, and if MIM fits, a tiered quote: prototype tool (4 to 5 weeks) and sintered samples (10 to 15 working days after the tool is ready), a pilot lot, and production pricing at your volume.
Share a link to your STEP file and drawing through our quote request form, or email leo@lichengmim.com.
