⚡ Key takeaways
- The shape of the chip is the shape of the cost. A basket of swarf beside a lathe is money that was bought and then cut away.
- Subtractive versus forming: turning starts from bar at least as large as the biggest diameter; cold heading starts from wire close to the effective section and moves the metal instead.
- Grain flow is the hidden quality difference. Cold-headed bolts test about 7–10 % higher in tensile strength than turned bolts of the same steel, and far better in fatigue at the fillet.
- Work hardening is both a dividend and a red line. Yield and tensile strength rise while elongation and reduction of area fall; over-deform and the part cracks.
- The material account is the sharpest blade. One documented case: 8.6 kg of wire versus 26.3 kg of bar for 25,000 solid pins, with scrap falling from 65 % to under 1 %.
- Cycle time differs by an order of magnitude — around 300 pieces per minute against about 4 — but tooling and changeover front-load the cost.
- Tolerance is the real boundary. Cold heading holds ±0.025 mm; anything at ±0.0125 mm or tighter has to be machined, and ±0.0025 mm needs ground bar.
- The practical answer is hybrid: cold form the body near-net, then machine only the few critical fits. A 6061 lifting sleeve went from CNY 50 to CNY 20 per piece and from 20 minutes to under 4.
1. Introduction: the shape of the chip is the shape of the cost
2. Subtractive versus forming
3. Effect one: grain flow
4. Effect two: work hardening
5. Effect three: the material account
6. Effect four: cycle time and tooling on a see-saw
7. Effect five: tolerance and process boundaries
8. Case study: a 6061 battery pack lifting sleeve, 60 % cheaper
9. Which parts to move from CNC to cold heading
10. Implementation roadmap
11. Conclusion: change the way the chips are made
12. Sources and standards
1. Introduction: the shape of the chip is the shape of the cost
Walk into a machine shop and look at the basket of curled aluminium or steel swarf beside the lathe. That basket is discarded cost. Its weight equals metal you paid to buy and then paid to cut away.
The trade often cites the 15/85 rule from the iceberg model of total cost of ownership: the direct purchase price of a fastener is typically only about 15 % of what the joint actually costs, with the other 85 % spread across design selection, assembly efficiency, quality rework, logistics and after-sales. Inside that visible 15 %, material is usually the largest block — and it is exactly the block cold heading rewrites best.
Many design engineers treat CNC as the default, on the grounds that it is slower but safer. On a single prototype that is fine. On a line producing several hundred thousand parts a year, it is the most expensive sentence in the project. This article avoids slogans and breaks the difference into five measurable dimensions: grain flow, work hardening, the material account, cycle time and tooling, and the tolerance boundary. Each comes with numbers, so the question becomes when to switch, how much it saves, and when not to switch at all.
2. Subtractive versus forming
CNC turning is subtractive: take a bar and cut away everything that is not the part. Cold heading is formative: in a sequence of dies and punches, a cut length of wire is squeezed beyond its elastic limit into the required shape. The material is redistributed, never removed.
That difference has a consequence that is easily missed: the bar diameter for turning must be at least the largest diameter of the part, whereas the wire diameter for cold heading only has to be close to some intermediate section. A hollow sleeve 30 mm across with a 3 mm wall has to be cut from solid bar over φ30 by turning; by cold heading it can grow from wire close to the finished minimum effective diameter.
Lay the capabilities side by side and the difference is plain:
| Manufacturing factor | Cold heading | Machining (CNC) |
|---|---|---|
| Part quality | ✓ | ✓ |
| Yield and tensile strength | ✓ Clearly better | — Grain severed, strength reduced |
| Close tolerances | ✓ | ✓ |
| Retained characteristics (press fit, locking) | ✓ | ✓ |
| Scrap generated | — Setup stock only | ✗ On every part |
| Complex geometry | — Limited by the forming envelope | ✓ Maximum freedom |
| Long or thick shanks | ✓ | — |
| Lowest tooling cost | — Tooling is expensive | ✓ Very low tooling |
| Short changeover | — 6–12 hours | ✓ 2–4 hours |
| Fast cycle | ✓ Order of magnitude | — |
| Lowest total cost for solid pins | ✓ | — |
| Lowest cost for hole processing | ✓ | — |
The decisive line in that table: given enough volume, cold heading wins on strength, total cost and hole cost at the same time; the genuine advantage of machining is the first part, where tooling and setup are nearly free. Cold heading carries a substantial tooling and setup front-load, but once it runs the unit cost collapses. Understand that and every later decision becomes simpler.
3. Effect one: grain flow
Metal is not uniform putty. It is a mass of grains, and rolling and drawing align them into a fibrous direction along the axis — the grain flow. When a part is loaded, the load travels along those flow lines.
The problem with CNC is that it severs the material and severs the grain flow with it. At the head-to-shank transition, at fillets and at steps, the flow lines are cut clean through, creating a mechanical origin for fracture.
The value of cold heading is that the metal flows in the die and the flow lines follow the contour — round the head, through the neck, down the shank, unbroken. The classic mid-century text on the subject puts it plainly: cold-headed products are stronger than the same steel turned on an automatic lathe, because in turning the metal fibres under the head are cut, whereas in cold heading the fibres remain continuous and are directed towards the centre of the head.
Measured comparison on bolts of the same steel, turned versus cold headed:
| Bolt size | Steel | Turned tensile (kgf/mm²) | Cold-headed tensile (kgf/mm²) | Gain |
|---|---|---|---|---|
| M8 | 40 steel | 62 | 68 | about +9.7 % |
| M10 | 35 steel | 56 | 60 | about +7.1 % |
| M12 | 40X steel | 60 | 65 | about +8.3 % |
| M18 | 35 steel | 52 | 57 | about +9.6 % |
The absolute numbers are modest but the meaning is not: the same grade of material gives a cold-headed part a step up in strength. Break testing along the shank confirms it — after heat treatment, a cold-headed blank carries 10–20 % more load before breaking than a turned blank.
More important than static strength is fatigue. Fatigue cracks almost always start where stress concentrates: a small nick, a tool mark left by a turning pass, a transition fillet. Every extra machining operation is another chance to leave a notch. Cold heading forms the shape and the outside diameter in a die, which reduces machining marks and keeps the flow continuous through corners, so the fatigue gain at the critical loaded fillet is particularly marked.
This is not theoretical. On a 6061 aluminium EV battery pack lifting sleeve, machining straight from bar severed the internal grain flow and risked weak points under repeated stress or heavy lifting. Cold forming left the flow continuous around the flange, body and hexagon. Tensile testing showed the cold-formed version withstanding stress fracture better than the pure CNC version. On a lifting point that carries an entire pack, that is not a refinement — it is safety margin.
4. Effect two: work hardening
Cold heading forces plastic deformation beyond yield at room temperature, which is itself a cold-working process. Deformation and hardening go together: yield point, elastic limit and ultimate strength rise, while elongation and reduction of area fall.
The dividend is obvious. Some parts can meet the drawing with less heat treatment or none at all, and surface hardness and wear resistance improve at the same time. That is why a cold-headed part often outlasts a turned part that looks identical.
The red line is just as clear. Ductility falls with deformation, and past a point the part cracks. The higher the deformation q, the more ductility is consumed, the harder further deformation becomes, and the sooner rupture arrives. Where that line sits varies enormously with material:
| Material / condition | Deformation q | Result |
|---|---|---|
| 30XГС steel | Below 87 % | No cracking |
| 60Г steel | 50 % | Already cracked |
| Carbon steel under 0.45 % C, head D₀/H = 2.0 | 60–70 % | High deformation achievable |
| High-carbon tool steel Y10 / Y12 (1.0–1.2 % C) | 20–30 % | Severely restricted, often not enough in practice |
| Steel up to 0.5 % C, three blows, no intermediate anneal | 65–85 % | Manageable |
Three engineering answers exist: progressive deformation across several stations, each blow taking only part of the total; intermediate annealing to reset ductility before continuing; and the right incoming condition, meaning annealed or normalised wire. The number of blows is not arbitrary. The industry rule reads the ratio of blank free height to wire diameter, h/dm: up to 2.5 in one blow, up to 4.5 in two, up to 6.5 in three, and beyond 8 the work splits across two machines with a heat treatment in between.
Cold heading places hard requirements on incoming material. The general technical requirement for cold-heading annealed wire sets both hardness and mechanical floors:
| Steel | Brinell indentation ≥ (mm) | Hardness ≤ HB | Tensile σb ≤ (kgf/mm²) | Elongation δ ≥ (%) | Reduction ψ ≥ (%) |
|---|---|---|---|---|---|
| 10 | 5.2 | 131 | 45 | 26 | 60 |
| 15 | 5.1 | 137 | 48 | 23 | 55 |
| 35 | 4.8 | 156 | 55 | 18 | 50 |
| 45 | 4.7 | 163 | 60 | 15 | 50 |
| 40X | 4.5 | 179 | 60 | 14 | 50 |
| 15XФ | 5.0 | 143 | 45 | 20 | 60 |
Summarised in one line: tensile strength of cold-heading wire should generally not exceed 70 kgf/mm², elongation should be at least 12 %, and reduction of area at least 50 %. If the wire cannot produce that entry ticket, heat treat it to adjust the structure rather than forcing the press.
Stainless is the special examination case for this rule. It work-hardens severely, and before heading the surface usually needs dedicated treatment to reduce friction and prevent galling and tearing. The classic answer is a thin copper layer on the stainless wire before heading, applied through a ten-step chain: organic solvent degrease → electrochemical cleaning at 82–87 °C and 6 V, 30 s cathodic then 30 s anodic → water rinse → acid dip for about 30 s → water rinse → electro-acid pickling at 20 °C in 50 % HCl or 10 % H₂SO₄ for about 2 min → water rinse → copper cyanide plating → water rinse → hot water rinse, and only then into the press. That copper layer is the lubricating separation film, and it is one of the preconditions for stable production of stainless cold-headed parts such as SUS630, 316L and 302HQ.
5. Effect three: the material account
Material utilisation is the sharpest blade cold heading has. Turning must start at the largest diameter, and cutting itself generates chips. Stack the two and the scrap rate becomes startling.
A measured case from the fastener industry: 25,000 solid pins, φ3 mm × 30 mm, cold headed versus machined:
| Item | Cold heading | Machining |
|---|---|---|
| Raw material form | Wire | Standard bar |
| Raw material required | 8.6 kg | 26.3 kg |
| Changeover and setup | 6–12 hours | 2–4 hours |
| Total production time | 1.4 hours | 104 hours |
| Scrap rate | Under 1 % — setup stock only | 65 % |
| Outside diameter tolerance | ±0.025 mm | ±0.0125 mm (tighter) |
For the same batch, cold heading used about a third of the material and roughly 74 times less production time, with scrap falling from 65 % to under 1 %. A second comparison, from a multi-station former builder, shows the weight of the material account just as clearly:
| Cut-off method | Weight per 1,000 pieces | Comment |
|---|---|---|
| Cold heading blank | 326 lb | Wire diameter close to the effective section |
| Turning blank | 3,000 lb | Bar diameter equals the largest part diameter |
| Saving | 2,674 lb | About 89 % less metal purchased |
The logic behind that ratio is simple: cold heading moves material, machining removes it. The more expensive the raw material — stainless, superalloy, copper alloy, titanium, 7xxx aluminium — the larger the material account within total cost and the more dramatic the cold heading return. The same CNY 30 saving is a cost reduction on low-carbon steel and a viability decision on SUS630.
In the 6061 lifting sleeve case this was expressed in very plain numbers: the original route cut away about 50 % of the metal from solid bar; the cold-formed route reached material utilisation close to 98 %.
6. Effect four: cycle time and tooling on a see-saw
The cycle time gap is an order of magnitude. Cold heading forms roughly 300 solid pins per minute against about 4 by machining. A multi-station former produces several parts progressively per cycle, giving 50–200 pieces per minute, and a six-station process can go from wire to finished blank in under half a second. Turning, single or multi-spindle, typically averages under 50 pieces per minute.
Against that, cold heading carries a front-load that has to be acknowledged: expensive tooling and slow changeover. On the solid pin example, cold heading needs 6–12 hours of setup and about USD 5,000 of tooling, while machining needs 2–4 hours and about USD 200. That is the see-saw — machining buys very low tooling cost at a very high unit cost, and cold heading does the reverse.
The fulcrum is volume. The classical rule is blunt: because equipment productivity is high, tooling expensive and setup time long, cold heading is only appropriate for large batches, at minimum around 5,000 pieces. Below that, tooling amortisation pushes the unit cost above the alternative.
The decision factors, gathered into one table:
| Decision factor | Favours cold heading | Favours CNC machining |
|---|---|---|
| Annual volume | Large, long-term stable — tooling amortises | Small batch, one-off, prototype |
| Material share of cost | High — stainless, alloy steel, aluminium, copper | Low — ordinary low-carbon steel |
| Head-to-shank ratio | Large — head far bigger than shank | — |
| Geometric complexity | Surfaces of revolution, multiple diameters, asymmetric but within the wire envelope | Deep cavities, cross holes, complex grooves |
| Fit tolerance | ±0.025 mm or looser | ±0.0125 mm or tighter |
| Delivery pattern | Long-term stable, tooling can be prepared ahead | Frequent design changes, rapid iteration |
| Strength and fatigue | Wants continuous grain flow and longer fatigue life | Insensitive to static strength |
7. Effect five: tolerance and process boundaries
Every process has a boundary, and stating it is more valuable than enthusiasm. Tolerance is the first wall cold heading hits:
| Method | Stock | Stock cost | OD tolerance, metric | OD tolerance, inch |
|---|---|---|---|---|
| Cold heading | Wire | $ | ±0.025 mm | ±0.001 in |
| Machining | Standard bar | $$ | ±0.0125 mm | ±0.0005 in |
| Machining | Ground bar | $$$ — about 3× standard bar | ±0.0025 mm | ±0.0001 in |
The conclusion is clear: cold heading holds ±0.025 mm, which is finer than a human hair and covers the great majority of fastener and pin work. Where a part is a critical alignment feature needing tighter outside diameter than that, machining takes over: ±0.0125 mm is reachable on standard bar, and ±0.0025 mm requires ground bar at roughly three times the cost, at which point cold heading is no longer the first choice. Length tolerance is comparable between the two, around ±0.25 mm depending on part length.
Geometric and structural boundaries also apply:
- Chamfer angle: the optimum forming angle for cold heading is 30° or less, whereas the optimum cutting angle is 45°. A 25–40° chamfer suits most applications and gives maximum fill.
- Length-to-diameter ratio: upsetting inside a full die generally limits shank length to 9.5–10 times the diameter, with the ejector entering the die bore by at least 1.5–2 times the diameter. Beyond that, an open split die is needed and tooling complexity rises.
- Head proportion: above h/dm of 8, the work must be split across two machines and two operations, usually with a heat treatment between.
- Size ceiling: traditional cold heading is centred on small and medium surfaces of revolution in medium to large batches, typically up to 22–24 mm diameter and 225 mm shank length. Beyond that, machine footprint, weight, power and management complexity all climb, and hot forging usually takes over. Modern multi-station equipment pushes the boundary to about 41 mm cut-off diameter and 355–432 mm under-head length.
- Material cleanliness: cold-heading wire must be free of scale, porosity, laps, seams, inclusions and burrs, because these open directly under high-pressure deformation. Scale also behaves like emery, working into slideways and punch carriers and accelerating machine wear.
So the correct posture is not "cold heading replaces CNC". It is: use cold heading as near-net forming to take the main shape and most dimensions, and reserve CNC for the few critical fits that genuinely need it.
8. Case study: a 6061 battery pack lifting sleeve, 60 % cheaper
This case ties every dimension above into one part: a flanged hollow sleeve used to lift and secure an EV battery pack during assembly, requiring a precision thread and a smooth top mating face, in thousands per line.
The original route was classic over-engineering: pure CNC from 6061 solid bar. Not impossible — just slow, wasteful and expensive, at about CNY 50 per piece.
Re-examining the geometry showed a flanged hollow sleeve, a shape cold heading was born for. The route became "cold-headed blank plus limited secondary machining": form the base shape by cold heading for speed, repeatability and near-98 % material utilisation, then run a few fast CNC operations to turn the top thread and finish the seat face.
| Metric | Pure CNC (6061 solid bar) | Cold-formed blank plus limited CNC |
|---|---|---|
| Unit cost | About CNY 50 | About CNY 20 — CNY 30 saved per piece |
| Cycle time per piece | About 20 minutes | Under 4 minutes |
| Daily output per line | About 60 pieces | Over 300 pieces |
| Output gain | Reference | About 5×, with no extra equipment or floor space |
| Material utilisation | About 50 % — half the metal cut away | Close to 98 % |
| Grain flow | Severed, potential weak points | Continuous around flange, body and hexagon |
| Stress fracture resistance | Reference | Higher in tensile testing |
The value of the case is that three things improved at once: cheaper, faster and stronger. It breaks the reflex that cost reduction must cost quality, because the cost advantage comes from material and cycle time while the performance advantage comes from grain flow. The two do not conflict.
The same logic recurs across industries. Documented examples from multi-station cold forming:
- Airbag mounting pin: originally specified as a die casting, which could not meet the mechanical requirements at prototype stage. Changed to 10B21 cold heading with secondary machining, giving a stronger and more reliable part.
- Brake system piston rod: originally turned, requiring the largest bar diameter. Changed to cold heading plus shaving, saving USD 0.60–0.80 of material per piece.
- Steering shaft with an involute spline: machined splines could not reach the strength requirement. Changed to 1035 alloy steel cold heading with rolled threads, with the spline formed in a second cold-forming operation to limit damage.
- Wheelchair lift mechanism shaft: originally a turned shaft welded to a washer. Now formed as one cold-headed part with continuous grain flow; even with subsequent drilling, tapping and broaching, the result is stronger and cheaper than the assembly.
- Commercial vehicle air-conditioning valve stem: originally two turned 302 stainless parts brazed together, below strength requirement. Changed to a single cold-headed 3/8 in × 3 in part, eliminating the brazing operation and reducing both material and part cost.
9. Which parts to move from CNC to cold heading
Compressed into a checklist you can take into a design review. Meeting any one criterion means the part deserves a cold heading feasibility review; meeting several means cold heading is almost certainly the better route.
| Criterion | Comment |
|---|---|
| Large head-to-shank ratio | The shape cold heading does best, and where its advantage is greatest |
| Part length more than 1.5× its largest diameter | Typical cold-formed proportion |
| Multiple diameters | One progressive forming replaces several turning steps |
| High production quantity | Tooling cost amortises |
| Raw material is a large share of part cost | Saved material converts straight into margin |
| Fine surface roughness required | Formed surfaces beat normal turned surfaces |
| Specific strength, hardness or toughness | Work hardening can replace or reduce heat treatment |
| Specific geometry, drive system or extrusion | Blind and through holes, drive recesses and asymmetric bodies form directly, removing a drilling step |
The reverse list — keep CNC — covers small custom batches, critical alignment features needing ground-bar tolerances, geometry that functionally demands complexity, and features outside the forming envelope such as cross holes, deep cavities and internal grooves.
10. Implementation roadmap
Five steps stand between "this should work" and "this runs in production":
| Step | Key actions | What you should have at the end |
|---|---|---|
| 1. Select material and wire diameter | Choose the grade against strength and corrosion needs; work back from the finished minimum effective section to wire diameter and tolerance | Wire specification, incoming material requirements |
| 2. Cold heading feasibility review | Work through the checklist in section 9; confirm head ratio, length-to-diameter, envelope and hole form are within capability | DFM conclusion, risk list |
| 3. Station and tooling design | Set the number of blows from h/dm; plan progressive deformation and any intermediate anneal; design carbide inserts | Process sheet, tooling drawings, press tonnage |
| 4. Trial run and first article | Verify dimensions, appearance, flow-line metallography and mechanical properties; confirm equivalence with the CNC version | First article report, flow-line micrographs, tensile and fatigue data |
| 5. Production and tool life management | Monitor die bore wear (0.05–0.10 mm small, 0.08–0.12 mm medium, 0.10–0.15 mm large), tool life and batch consistency | Stability data, cost-saving reconciliation |
A final piece of advice from the trade itself: work with joining and assembly specialists to establish the lowest-cost solution for the assembly. Cold heading is not a switch that can be thrown automatically; it needs the supplier involved at the design and material selection stage. Tooling cost is front-loaded and needs a volume commitment, and once the dies are cut, geometry and material are locked. That is why the design-for-manufacture review belongs before the drawing is frozen, not after the quotation arrives.
11. Conclusion: change the way the chips are made
If you are designing or buying a bracket, sleeve, pin, shaft or any surface of revolution that is currently "cut from bar", do not default to CNC without thinking.
Cold heading for the blank plus the minimum CNC for precision fits is a well-proven route that lowers BOM cost, shortens cycle time and raises strength and fatigue life at the same time. Sometimes the largest saving comes from changing the way the chips are made — from removing surplus metal to letting the metal flow where it belongs.
12. Sources and standards
- Metal Cold Heading Technology (B. M. Misuzhnikov and M. Ya. Grinberg, translated by the First Ministry of Machine Building General Machinery Research Institute): grain flow, work hardening and deformation limits, blow-count rules, technical requirements for cold-heading steel, copper preparation for stainless.
- Cold Heading versus Machining of Solid Pins, industry technical white paper: capability comparison, outside diameter tolerance versus stock cost, the 25,000-piece cost case.
- Cold Formed Specials, international cold-forming reference: process capability, industry cases for replacing turning, blank weight comparison.
- Cold Forming Basics: process principles and part geometry suitability criteria.
- How Cold Heading Cut EV Battery Pack Lifting Sleeve Cost by 60 % and Made It Stronger: cost, cycle time and strength comparison for the 6061 sleeve.
- Standards: GB/T 3098.1 (mechanical properties of bolts, screws and studs), ISO 898-1. Surface roughness references Ra 1.6 μm (about 63 μin) and Ra 0.15 μm (about 6 μin).
👉 Engineering services ｜ 👉 Request a technical quote