Cold Forging and Cold Heading Parts from China: What Actually Matters

I have spent more afternoons than I care to count standing next to cold-heading machines in Zhejiang and Guangdong, watching a bolt or a pin pop out every fraction of a second. It is a hypnotic process, and that is part of the problem: it looks so simple that buyers assume the quote is the whole story. It never is.

What cold forming actually is, and why it is cheaper

Cold forging and cold heading take wire or bar stock at room temperature and squeeze it into shape under enormous pressure. Because the metal is not heated, you avoid the scale, the shrinkage allowance, and the energy bill that come with hot forging. Material waste drops to almost nothing compared with machining, and the grain flow follows the part's contour instead of being cut across, which gives you better fatigue strength for the same alloy.

That is the upside. The catch is that the shape has to be achievable. Every heading blow moves a finite volume of metal, and if your design asks the material to flow further than it wants to go, you get laps, cracks, or a tool that dies after ten thousand parts.

The deformation question your supplier should answer in writing

Ask a shop how many blows a part needs and where the material is coming from and going to. A good factory can sketch the progression in two minutes. A shop that just says "yes, no problem" without looking at the sketch is a shop that will learn on your tooling budget.

Rough rule of thumb people in the trade use: carbon steel can typically be upset by around 2 to 2.5 times its diameter in a single blow, and about 4 to 5 times total across multiple stations before the material work-hardens to the point of cracking. Stainless and high-strength alloys are more stubborn. If your part geometry sits past those numbers, the shop should be telling you about annealing steps between blows, not quietly quoting a single-stage die.

Tooling: who owns it, and how long it lives

Cold-forming dies are not the simple blocks many buyers imagine. They are multi-station punches, and the life is measured in hits. For a straightforward mild-steel fastener you might hear 300,000 to over a million parts per die set depending on material and lubrication. Push into 10.9 or 12.9 grade steel and that number collapses.

Two things you need nailed down before the first payment:

Who owns the dies, and where do they physically sit. If the answer is vague, assume you do not own them.
What the die life assumption is in the unit price. If the quote assumes a million hits and the die realistically does 200,000, your price will drift upward on reorders and nobody will have to explain why.

Material: the cheapest place to get cheated

Wire rod grade is where the money hides. A shop quoting "carbon steel" could mean anything from a clean drawing-quality grade to whatever their trader had cheap that week. For anything structural, ask for the grade designation and a mill certificate, and for critical parts send a sample to a third party for chemical analysis and hardness. I have seen fasteners sold as 10.9 that pulled hardness numbers you would expect from a much softer grade.

If the part will be plated, remember that hardness and hydrogen embrittlement go together. High-strength parts that get electroplated need a proper bake per ASTM B850 or the same spec your buyer requires, or you will ship a batch that looks perfect and snaps in the field after a week.

Deburring, edge condition, and thread quality

Cold-formed parts come out with a burr. Whether the shop removes it, and how, is a real cost line the buyer rarely discusses up front. Tumble deburring is fine for many parts and disastrous for ones with fine threads or critical sealing surfaces. Vibratory finishing leaves a particular edge radius; abrasive flow or a secondary cut can leave another. If your assembly has an O-ring, a sharp burr will slice it and you will spend three weeks blaming the seal supplier.

Ask for parts with specified edge breaks and check a sample under a loupe. It takes five minutes and saves entire production runs.

Inspection that is worth the money

For cold-formed parts, the tests I actually pay attention to are:

Dimensional check on critical diameters and lengths, with the gauge and method named in the report.
Thread gauging with go and no-go for threaded parts, not just a visual glance.
Hardness in the core and, for hardened parts, at the surface, because case depth is where heat treatment regimes differ between shops.
For safety-relevant parts, magnetic particle inspection or a proper forging test for laps and seams.

A supplier who offers all this before you ask is worth a premium. A supplier who offers none and calls it "standard" is telling you who does the quality control. It is you.

First-article inspection is non-negotiable

Before volume, require a first-article inspection report covering the drawing dimensions, material cert, and any functional test. Hold the approved sample. When a reorder comes in two years later and the die has been rebuilt, compare against that sample, not against a fuzzy memory of what it looked like.

What tends to go wrong, and how often

In my experience the failures cluster in three places: material substitution, under-engineered tooling that fails mid-run, and skipped deburring or heat treatment. None of them show up in the quote. All of them show up in your returns. Budget for a real first article, name the material, count the die hits, and look at the edge under magnification. That is most of the battle.