How to Choose the Right Manufacturing Route for Small Metal Parts

Start with the part, not the process

The instinct to begin with a preferred process is understandable — a supplier that knows MIM will see MIM opportunities — but it produces unreliable comparisons. The correct starting point is the part itself: what it does, what it must survive, and which of its dimensions genuinely matter.

That reframing changes the questions. Instead of "can MIM make this?", the useful questions become "which features carry the function?", "which dimensions would cause a field failure if they drifted?", and "what does the part cost to make today, all in?". Answering those first turns route selection into an engineering exercise rather than a price comparison between incompatible quotes.

Factor 1: geometry complexity

Geometry is the strongest filter because it is binary. A process either can produce the shape or it cannot.

• Flat, thin parts point to stamping, provided the thickness is set by sheet stock and the features can be formed from a flat blank.
• Simple rotational parts — shafts, pins, spacers, bushings — point to turning or, at high volume, to PM.
• Parts with cross-holes, undercuts and side features cannot be pressed by PM in one piece. The realistic candidates are MIM, investment casting or machining.
• Large parts with internal cavities point to casting, where sand or ceramic cores create the void.
• Load-bearing parts where grain flow matters point to forging, with machining to finish the critical faces.

A useful test: count the number of distinct setups a machinist would need to produce the part from bar. One or two setups suggests machining is already efficient. Five or six suggests a forming route has something to remove.

Factor 2: production quantity

Volume does not choose the process. It decides whether a dedicated tool is affordable, and therefore whether any forming route is on the table at all.

These bands are a starting point, not a rule. A short program life of two years with 5,000 units a year is a different proposition from ten years at the same rate, even though the annual figure is identical.

Factor 3: material and performance target

Material narrows the field in two ways. First, not every alloy is available in every process: MIM and PM require powders that can be atomised to the right size and sintered to the required composition, so some high-alloy or free-machining grades are simply unavailable. Second, the process affects the properties that a given alloy delivers, because cooling rate, density and grain structure differ.

Two examples make the point. A part needing full density and corrosion resistance in 316L is a natural MIM candidate, because MIM reaches 95–99% of theoretical density. The same alloy pressed and sintered by PM typically lands at lower density, which is acceptable for a structural bracket but not for a pressure boundary. Conversely, a part needing maximum fatigue life in a load path is often better forged, because forging aligns the grain structure in a way no powder route replicates.

Factor 4: tolerance and finish

This is where route selection most often goes wrong, because tolerances are frequently applied to the entire drawing out of caution. That caution is expensive: if every dimension carries ±0.02 mm, only machining qualifies, and the comparison never gets the chance to consider a forming route.

A better structure separates three groups of dimensions:

• Critical features — fits, sealing surfaces, bearing journals, mating datums. These may need ±0.02 mm or better.
• Functionally significant dimensions — where deviation affects assembly but not performance, typically ±0.1 mm to ±0.3 mm.
• General dimensions — cosmetic or clearance features, which forming routes handle comfortably at as-formed tolerance.

A part with a handful of critical features and a majority of general ones is usually a forming part with a finishing operation, and it is often considerably cheaper than machining the whole component. That hybrid is the comparison worth making.

Factor 5: project stage

Prototype, pre-production and production pull in different directions.

• Prototype. No tooling. Machining or printing, usually from a near-net blank so the prototype's material behaviour is representative.
• Pre-production / pilot build. Often machining from the intended material, with process development running in parallel so the tooling decision is informed by real data.
• Production. Lowest landed unit cost, including finishing, inspection and logistics — and a tooling plan that can survive a design change.

Skipping straight from prototype to production tooling is a common and expensive mistake. If the design is still moving, tooling changes cost money and time; a short pre-production phase using machining usually costs less than a mould revision.

A filter you can run in ten minutes

Working down this list on a single drawing removes most routes before any supplier is contacted.

• Is the part small and difficult to machine? If yes, MIM and investment casting are worth evaluating.
• Are there cross-holes, undercuts or internal threads? If yes, PM is out unless those features are machined afterwards.
• Is the part essentially rotational? If yes, turning is the baseline, with PM as the volume alternative.
• Is the annual volume below a few thousand, or the program life short? If yes, plan for machining and revisit tooling only when volume is confirmed.
• Are more than a handful of dimensions tighter than ±0.05 mm? If yes, plan for machining, or a forming route with a defined finishing operation.
• Is the required alloy available as a powder? If no, MIM and PM are ruled out regardless of how well the geometry fits.

What to send first

A supplier cannot perform this analysis without the drawing. Sending a bare 3D model or a photograph means the answer will be generic. The useful package is:

• A 2D drawing with tolerances, datums and surface requirements.
• A 3D model, ideally STEP, so wall thickness and geometry can be assessed directly.
• The material, or the performance requirement if the grade is open.
• Annual volume and expected program life, plus any seasonal variation.
• A short note on what the part does and what happens if it fails.
• Any constraint already fixed — an existing assembly interface, a regulatory requirement, a finish specified by the end customer.

The honest answer is sometimes "do not change"

Route selection is not a mandate to convert. A part that is already made by machining, meets its requirements and costs an acceptable amount should usually stay there — particularly at low volume, or where the design is still evolving. A supplier that recommends conversion in every case is not doing the analysis.

The value of running the exercise properly is that when a conversion is recommended, it comes with a reason: this many machining operations removed, this much material no longer turned into chips, this tooling amortised over this volume. That is a case a buyer can defend, and it is the only kind worth acting on.

Related component families

These pages take the general framework down to specific component families and their DFM questions.