Cold Heading vs CNC Turning for Fasteners

Two ways to make the same fastener

A bolt has a head on one end of a shank. There are two completely different ways to produce that shape, and the difference is whether metal is moved or removed.

Cold heading takes a length of wire or rod, puts it in a die, and strikes it with a punch. The metal flows — the shank is upset into the die, the head is formed from material pushed out of the shank, and the part takes the shape of the tooling. Nothing is cut away. The process is the standard production route for high-volume fasteners and is the reason a box of screws costs what it does.

CNC turning takes a bar and cuts the shape out of it. The head is formed by turning down the shank to a smaller diameter, so the head is whatever is left of the original bar. Material is removed rather than moved. The process itself is covered on the CNC turning page.

Both routes produce a bolt that fits the same nut. The differences show up in how the part is made economically, what geometries are possible, and — most importantly, and least often discussed — in what happens to the grain of the steel.

	Cold heading	CNC turning
How the shape is made	Metal displaced into a die	Metal cut away
Material waste	Very little	A large fraction of the bar
Grain flow	Follows the part contour	Cut through at every step
Fatigue at the head fillet	Better — grain runs along the load path	Weaker — grain severed
Cycle time	Very fast	Slow by comparison
Tooling	Part-specific dies and punches	Fixtures and programs only
Tolerances	Moderate	Tight
Undercuts and sharp internal corners	Not possible	Routine
Material requirement	Must be ductile enough to form	Any machinable grade
Diameter range	Governed by the wire and the machine	Wide, including large
Threads	Normally rolled after heading	Cut or rolled
Best quantity	High	Low to medium
Typical parts	Screws, bolts, rivets, pins, studs, nuts	Non-standard shapes, large or hard parts

Two rows in that table deserve to be pulled out and examined: the grain flow, and the material utilisation. The first is an engineering difference that affects whether the part survives service. The second is an economic difference that can outweigh the machining cost on expensive alloys.

Grain flow: the difference that decides fatigue life

Steel is not isotropic. Rolled or drawn bar has a grain structure — the result of the rolling and drawing operations — that runs lengthwise along the bar. Mechanical properties follow that structure: the material is tougher and more fatigue resistant along the grain than across it.

Now consider what each process does to that structure at the most highly stressed point of a fastener.

In a cold-headed bolt, the metal flows. The grain that ran along the wire is deformed by the heading operation, and it bends around the head and follows the fillet where the head meets the shank. The grain ends up running along the load path — the path the force takes from the head into the shank.

In a turned bolt, the head is the original bar and the shank is cut down from it. The grain still runs lengthwise, but at the head-to-shank transition the material that was outside the shank has been cut away, and the fillet is a cut surface. The grain is not bent around the corner; it terminates at the cut surface.

The head-to-shank fillet is where bolts fail. It is the point of stress concentration, the place where fatigue cracks initiate, and the location every fastener standard pays attention to. A grain structure that runs along the load path at that point is materially better than one that is cut across it.

This is not a subtle metallurgical curiosity — it is why cold-headed fasteners are specified for fatigue-critical joints, and it is the strongest engineering argument for heading that exists. A turned bolt and a headed bolt of the same material, the same dimensions and the same thread can have different fatigue performance, and the drawing does not show why.

A related point applies to the fillet itself. A formed fillet has a smooth, continuous radius because the die made it that way. A cut fillet can be just as accurate, and a turned part can have an excellent radius — but it is a cut radius, not a flowed one.

Threads: the same principle, one operation later

Threads are usually rolled rather than cut, and for exactly the same reason.

Thread rolling takes a blank at roughly the pitch diameter and rolls it between threading dies. The material flows into the thread form, so the grain follows the thread profile — bending around each thread root and crest rather than being severed at the root. The rolling operation also work-hardens the flanks and leaves a smooth, compressed surface with no torn material.

Thread cutting removes material. The grain is cut across at every thread root, which is the point of maximum stress concentration in a threaded joint, and the cut leaves a sharper root profile and a rougher surface.

The consequences in practice:

• Rolled threads are stronger in fatigue, for the same grain-flow reason as heading.
• Rolled threads consume less material, because the blank starts at about the pitch diameter rather than at the full major diameter.
• Rolled threads are produced faster and with no swarf.
• Rolled threads work-harden the flanks, which improves wear resistance in repeated assembly.

On a headed fastener the thread is normally rolled after heading, so the whole part benefits from the same principle applied twice. On a turned part the thread may be cut, or rolled if the geometry and the blank allow.

The one case where rolling needs care is a thin wall or a thread running to a shoulder, where the forming pressure can distort the part or the thread cannot be rolled to the end. There, cutting is the safer route. This is the kind of thing that belongs in DFM feedback rather than in a specification. Thread standards and callouts are covered on the thread standards page.

Material utilisation, which can outweigh the machining cost

The second structural difference is how much of the bought material becomes a part.

Cold heading moves metal, so almost all of it becomes the fastener. The waste is a small trim and the remnant at the end of a coil — a few percent.

Turning removes metal. On a small fastener the shank is a fraction of the head diameter, so the majority of the bar can end up as swarf. The bought weight is the head diameter; the shipped weight is the finished part.

On ordinary carbon steel this matters but does not dominate. On an expensive alloy — stainless, titanium, a high-strength grade — the arithmetic changes: you are buying material at that alloy’s price and turning most of it into chips that are worth a fraction of what you paid. The scrap value recovers some of it, but not most of it.

That is the reason heading is often the only economic route for an expensive-alloy fastener at volume, independent of the cycle time. The material saving alone can decide it.

It is also worth noting where the advantage does not apply: where a part is small and the head is not much larger than the shank, the removed volume is small and the material argument weakens.

Tooling, volume and where the crossover sits

The two processes have opposite cost structures, and the same logic applies as anywhere else in manufacturing.

Cold heading puts the cost into tooling. Dies and punches are specific to the part — the head form, the shank diameter, the shoulder, any knurl or flange. The tooling is a real cost that has to be justified by the volume, and a design change after the tooling is made means remaking at least part of it. Once running, the process is extremely fast, and the unit cost is low.

Turning puts almost nothing into tooling. Fixtures and programs are inexpensive relative to heading tooling, setup is quick, and a design change is a program change. Every part then carries its full cycle time, so the unit cost stays high and does not fall much with quantity.

The crossover therefore depends on:

• Volume — the primary factor, and the one that decides most cases
• Material — an expensive alloy shifts the crossover towards heading, because the material saving is larger
• Size — a large fastener removes more material and takes longer to turn, which also favours heading where the wire and machine can take it
• Geometry — a shape that forming can produce favours heading; a shape it cannot removes heading from consideration entirely
• Tolerance — a part needing tighter tolerances than heading holds has to be machined, at least on the critical features

There is also a third route that is common in practice and rarely mentioned in comparisons: head first, then machine. A part is cold headed to near-net shape and the critical features — a precise shoulder, a ground diameter, a tight concentricity — are turned afterwards. That gets the grain flow and the material efficiency of heading with the accuracy of turning, applied only where it is needed, and it is the route behind a great many production fasteners.

What neither process can do

Cold heading cannot:

• Produce undercuts, sharp internal corners or re-entrant shapes the metal cannot flow into
• Form a part whose shank cannot be pushed out of the die — some form of draft or release is required
• Reliably form high-carbon or very hard alloys without cracking
• Hold the tightest tolerances on diameter, concentricity and straightness
• Justify itself on a small quantity, because of the tooling investment

CNC turning cannot:

• Beat heading on unit cost at volume
• Avoid turning most of the bar into chips
• Reproduce the grain flow through the head fillet
• Compete on cycle time for a simple high-volume part

The two processes are complementary rather than rival. The practical question is not which is better but which is right for this part at this quantity.

Where each route lives

Cold heading shows up as:

• Screws and bolts produced in volume — see custom screws and custom bolts
• Rivets and pins — see custom rivets
• Studs, threaded rods and headed fasteners
• Nuts formed from wire — see custom nuts
• Any fastener where the head-to-shank fillet is fatigue-critical
• Parts where the material is expensive enough that waste matters

CNC turning shows up as:

• Fasteners produced in small quantities or as prototypes
• Large-diameter fasteners beyond the practical wire range
• Non-standard geometries with undercuts or sharp internal corners
• Hard or high-alloy grades that will not form
• Parts with tight tolerances on diameter and concentricity
• Headed blanks finished to tolerance on critical features

When neither is right

The part is a body of revolution needing the highest accuracy throughout. Turning is the route, and the discussion is which turning process — the comparison of conventional and Swiss-type turning covers where that line sits.

The part is a plate or a bracket rather than a fastener. Neither process applies, and the answer is milling.

The part has to be made in a material that will not form and cannot be turned economically. That is a genuinely difficult case, and the honest answer is usually to reconsider the material rather than the process — the grades we run as standard are on the materials page.

The part is cast or moulded. Where a fastener-like component is produced in very high volume and the material suits, casting or moulding may beat both.

The callout that belongs on the drawing

Instead of	Write
Fastener	State the quantity — it decides the process more than the geometry does
Bolt, high volume	State whether the joint is fatigue-critical — it argues for heading
Bolt, tight concentricity	Say so — heading holds looser tolerances; machining may be needed
Thread	Say whether rolled or cut is required, or leave it to DFM
Bolt, expensive alloy	Say the grade — material waste may decide the route
Bolt, undercut or internal corner	Heading cannot form it; the part has to be machined

The two lines worth adding to any fastener enquiry are the quantity and the grade. Together they decide whether heading tooling can be justified and whether the material is one that can be formed at all.

Both routes are quoted against the drawing, and where the part sits near the crossover the useful answer is a price for each rather than one. Send the drawing with the quantity, the grade and any fatigue or concentricity requirement, and the quote comes back within 24 hours with DFM feedback — including a note where the process and the tolerance or geometry are working against each other. Orders start at 50 pieces per style and material, samples from a single piece in 5–10 days, and every batch ships with a material certificate (MTR) tracing the grade to its mill heat plus pre-shipment photos.