Metal Stamping & Progressive Die Design: The Complete Engineering Guide — the 6% Clearance That Shears 2 mm CRCA Cleanly and the 16% It Takes at 1,400 MPa, the 235 kN Behind 420 mm of Cut Edge, the 3 mm Scrap Bridge That Sets 63% Material Yield, and the ₹8 Lakh Tool That Pays for Itself at 18,200 Parts

Die clearance from 6% to 16%, 235 kN of cutting force, 62.9% strip yield, springback, 22MnB5 hot stamping and the ₹8 lakh tool crossover at 18,200 parts.

2026-10-22 · FabFlow Team

Metal Stamping & Progressive Die Design

A laser cutter makes a shape. A stamping die makes a rate.

That sentence contains the whole economics of the press shop. A fibre laser cutting 2 mm CRCA at 4 m/min produces a bracket every ninety seconds and costs ₹15–25 of machine time to do it. An eight-station progressive die in a 100-tonne press produces the same bracket every second, from the same coil, at a conversion cost near ₹1. There is no contest at volume — and no contest at low volume either, because the die costs ₹8 lakh before it makes a single good part.

The transition between those two regimes is arithmetic, not judgement. It is set by five numbers.

The first is 235 kN. That is the shearing force behind a 2 mm CRCA bracket with 420 mm of cut perimeter at a shear strength of 280 MPa. Add 10% stripping force and a 1.3 safety factor and the die asks 336 kN — 34 tonnes — of the press, which is why a 100-tonne machine is the sensible home for it and a 40-tonne machine is not.

The second is 6%. Per side. That is the clearance at which mild steel shears with a clean burnish band and a controlled burr. It rises to 10% for HSLA, 12–15% for DP590, 14–16% for DP780 and martensitic grades above 1,200 MPa — and above 1,400 MPa it starts falling again, because the punch tip, not the edge, becomes the failure mode.

The third is 62.9%. A 110 × 70 mm blank on a 76 mm strip at a 113 mm pitch uses 62.9% of the coil. The remaining 37.1% is not waste in the abstract; it is 3 mm of edge scrap, 3 mm of bridge and the carrier that holds the part while six punches hit it.

The fourth is 18,200. That is where a ₹8 lakh progressive die amortises against a ₹45/part laser-and-brake route on this bracket. Below it, the laser wins. Above it, the die wins by an increasing margin — ₹25.40/part at 50,000 against ₹53.36.

The fifth is 250 tonnes. A 600-tonne mechanical press has 600 tonnes available at half an inch above bottom dead centre. Three inches up the stroke, it has 250. A die designed against the nameplate rather than the tonnage curve is a die that stalls, deflects and spits burrs.

This guide runs in twelve sections: the cutting mechanics, the clearance ladder, strip layout and yield, the anatomy and timing of a progressive die, bending and springback, deep drawing, hot stamping of 22MnB5, press selection by energy rather than tonnage, tool steel and die life, the tolerances a die can actually hold, the full Indian cost model, and the DFM rules that decide whether the tool is buildable at all.

1. The Four Zones of a Sheared Edge

Every mechanical cut — blanking, piercing, shearing, trimming, slitting — produces the same four-zone edge. The proportions change with clearance, tool sharpness and material, but the sequence does not.

Zone · Depth (typical) · What happens · What it tells you

Rollover · 10–25% of t · Material bends over the punch radius before fracture starts · Too much = clearance too large or punch too dull

Burnish (shear) · 25–50% of t · Clean plastic shear, no fracture · The only zone that seals a cut edge

Fracture · 25–60% of t · Crack propagation, rough and angled · Too much = clearance too large

Burr · 0–10% of t · Final tear, work-hardened lip · The acceptance criterion, and the wear indicator

On mild steel the target is a burr below 10% of thickness — 0.20 mm on 2 mm stock. On AHSS the burr is a poorer indicator, because these grades fracture before they can grow a large burr; the useful criterion becomes burnish uniformity and the transition into the fracture zone. That is the distinction that separates a shop that understands AHSS from one that keeps sharpening a die that was never the problem.

The cutting operation also changes the material it cuts. Mechanical shearing work-hardens a zone that extends roughly half a metal thickness from the edge, dropping the local work-hardening exponent n and the edge ductility with it. A hole that will later be stretch-flanged therefore has a different allowable strain from the parent sheet — which is why the hole-expansion ratio, not the tensile elongation, governs edge-critical designs.

The force identity

Blanking force is the cut perimeter multiplied by thickness and shear strength:

where L is the total cut length in mm, t the thickness in mm, \tau_s the shear strength in MPa and R_m the tensile strength. The factor 0.8 is the usual engineering approximation for \tau_s/R_m on cold-rolled steel; for austenitic stainless it is closer to 0.65–0.7 because the grade work-hardens through the cut.

For the bracket:

Stripping force — the force needed to pull the strip off the punch — is 8–12% of blanking force for a well-supported stripper. The press must supply both, plus a margin for tool wear and material variation:

Select the press at 1.5–2× that figure. A 63-tonne press is marginal, a 100-tonne press is comfortable, and the 100-tonne machine also has the bed area and the shut height to carry an eight-station die. The tonnage calculation is not there to size the press to the part; it is there to stop you sizing it to the part.

Shear angles: buying force with stroke

A flat punch hits the whole contour at once, so the force peaks at full perimeter. Put a shear angle on the punch face — a roof-top, a bevel, a chamfer — and the cut starts at one edge and propagates across, trading stroke for force.

The published numbers are large. A roof-top punch reduced punching force by up to 80% against a flat punch on 1 mm, 1,400 MPa martensitic steel. A bevel of 3–6° on DP500/780 and DP550/980 reduced peak piercing force by more than 50%, and at 17% clearance improved the hole expansion ratio by more than 60% — because the bevel also produces a cleaner burnish-to-fracture transition.

There is a cost. A shear angle imposes a lateral load on the punch, and a large chamfer risks plastic deformation of the punch tip. The design rule is to make the chamfer deep enough to initiate the cut before the full face contacts the sheet, and no deeper.

Snap-through: the force you did not design for

At breakthrough, the stored elastic energy in the press frame and die releases in milliseconds. The ram accelerates downward and the frame rebounds — reverse tonnage. In high-strength blanking it routinely exceeds 40% of the cutting force, and it is the load that cracks punches, loosens guide pins and hammers the press counterbalance.

On the 235 kN bracket that is 94 kN of reverse load — 9.6 tonnes slamming upward through the tool at 60 strokes a minute. Hydraulic cushioning, die springs, dampers and, increasingly, a servo press with a programmed deceleration at breakthrough are how that energy is managed. Ignore it and the die life you calculated in §9 never materialises.

Energy per stroke

Force is not work. The blanking energy is the area under the force-displacement curve, and for a cut that completes at roughly 30% penetration:

For the bracket: 0.15 \times 235{,}000 \times 0.002 = 70.6 J. At 60 strokes per minute that is 4.2 kJ/min — 71 W of average cutting power. Blanking is a force problem, not an energy problem.

Drawing is the opposite. A 193 kN draw force over a 60 mm stroke stores 5.8 kJ per stroke, and at 30 strokes per minute the press must average 2.9 kW while delivering 174 kJ/min through a flywheel that has only part of a revolution to recover. That is why a press can be rated for the tonnage of a draw and still stall on it.

2. Clearance: The 6-to-16% Ladder

Die clearance is the gap between punch and die, expressed as a percentage of sheet thickness. Almost every clearance argument in a press shop is really a units argument, because half the catalogues quote the gap per side and the other half quote the total.

Rule one: establish whether the number is per side or total, and double or halve accordingly. A catalogue that recommends "15% clearance for mild steel" is quoting total gap; used as a per-side value it produces a 30% total gap, a huge rollover and a burr you will spend a shift chasing.

The per-side ladder, assembled from the AHSS Guidelines and tooling practice:

Material · Tensile strength · Clearance per side, % of t

Mild steel / CRCA (DC01–DC04) · 270–400 MPa · 5–8

HSLA 340–550 · 410–550 MPa · 8–10

DP590 · 590–700 MPa · 12–15

DP780 · 780–900 MPa · 14–16

Martensitic 1,200–1,400 · 1,200–1,400 MPa · 14–16

Austenitic stainless 304 · 520–700 MPa · 8–12

Aluminium 5052-H32 · 215–275 MPa · 5–8

Copper / brass · 220–320 MPa · 4–6

Fine blanking (any grade) · — · 0.5

The physics is a leverage argument. The clearance is what allows the punch to bend the slug down and start a crack ahead of the edge; a stronger material needs a longer lever to initiate that crack. Below 6% the punch rubs, the die wears and the press needs perfect alignment. Above 16% the rollover grows until the edge is a tear rather than a cut, and on a martensitic grade the bending stress on the punch edge starts chipping it.

The practical form is:

with k_c taken from the table. For a 2 mm DP590 bracket at k_c = 0.13, the per-side gap is 0.26 mm, so the die opening is 2c larger than the punch — 0.52 mm on the dimension. That is the number the toolmaker grinds to, and it must be written on the die drawing in both forms.

Two refinements matter in production. First, clearance should increase slightly with thickness at the same strength, because the crack has further to travel and the bending moment scales with t^2. Second, the optimum clearance for hole expansion is not the same as the optimum for tool life: ISO 16630 specifies 12% for the standard hole-expansion test, while production studies on CP1200 found 15% outperformed both 10% and 20%. Clearance is a process choice, not a constant.

Fine blanking: the other end of the scale

Fine blanking uses roughly 0.5% clearance per side — a tenth of conventional practice — in a triple-action press that clamps the sheet with a V-ring around the cut line, applies counter-pressure through the ejector, and shears at 4–15 mm/s. The result is a cut face that is 85–100% burnish, with dimensional tolerance into the ±0.01–0.02 mm band and a burr measured in hundredths of a millimetre rather than tenths.

The mechanics explain the cost. The V-ring (or "stinger") embeds into the sheet outside the part contour and generates a high hydrostatic compressive stress at the shear zone, suppressing crack initiation. Holding pressures are 10–30 MPa, counter-pressure is similar, and the total force is typically two to three times a conventional blank — which is why fine blanking presses are built with a third ram and why the tooling is 2–3× the price. It replaces a machined edge, not a laser edge, and that is the comparison that justifies it.

3. Strip Layout: The Only Place Yield Is Decided

Before a die exists, a strip layout fixes three things permanently: how much coil becomes part, where the burr lands, and whether the strip can feed at 60 strokes a minute without buckling.

The economy factor is:

where A_b is the blank area, n the number of rows, W the strip width and p the feed pitch. A layout below about 60% is usually a layout that has not been nested; a layout above 85% is usually a part with straight sides and no windows.

For the bracket — 110 × 70 mm envelope, 5,400 mm² of blank area after windows, 2 mm thick:

Edge scrap: 3.0 mm each side, so W = 70 + 6 = 76 mm
Bridge between parts: 3.0 mm, so p = 110 + 3 = 113 mm

The strip mass consumed per part is:

against a part mass of 84.8 g. At ₹62/kg coil, the material line is ₹8.36/part, of which ₹3.10 is the coil value of the scrap. That ₹3.10 is not a rounding error: at 100,000 parts it is ₹3.10 lakh of coil, nearly 40% of the tool cost — recoverable at scrap price, not coil price, but recoverable.

The rules that set the numbers

Scrap bridge. The bridge between adjacent openings carries the strip and must survive the punch load. The usual rule is 1.2t, with an absolute minimum of 2 mm. Thick or hard material needs more; a bridge that tears is a strip that jams.

Edge scrap. It keeps the strip rigid enough to feed. Below 0.2 mm thickness, 1.0–1.5 mm per side is enough because the strip is light; from 1.0 to 3.0 mm, use 2.0–4.0 mm because the feed force and the strip mass are higher.

Nesting. Straight-edged parts nest at 70–85%; curved and angled parts at 45–65%. Rotating a blank by 180° between rows, or running two rows in a mirror layout, can recover 10–15 points of yield — but it doubles the die width and may push the press beyond its bed.

Grain direction. Bending across the rolling direction is stronger; bending along it cracks. For parts that will be formed, orient the bend line at 45° or more to the coil axis. For parts that will only be blanked, grain direction is irrelevant and the tighter nest wins.

Burr side. In blanking the burr forms on the punch side; in piercing it forms on the die side. The layout must put the burr where the drawing says — usually away from a visible face, a sliding surface or a weld line. Retrofitting a burr-side requirement after the die is built means rebuilding the die.

Carrier type. A centre carrier leaves a central spine and pushes yield up; a single-side carrier is simpler and cheaper but limits station access; a double-side carrier is the most rigid and the least efficient. The carrier is not scrap — it is the fixture that positions every part for every station — so it must be designed before the strip is nested, not after.

A 1,000 mm coil slits into thirteen 76 mm strips with 12 mm left over — 98.8% slitting yield. That last 12 mm is unavoidable unless the part width is changed, which is why coil width is a design variable and not a purchasing afterthought.

4. Anatomy and Timing of a Progressive Die

A progressive die performs every operation on every stroke, at a different station, on a strip that advances one pitch each time. Nothing is finished until the last station, and nothing can be reworked, because the part is attached to a strip that is already moving.

flowchart LR
    A["Coil 1,000 mm"] --> B["Decoiler"]
    B --> C["Straightener + servo feeder"]
    C --> D["Progressive die"]
    D --> E["Finished part"]
    D --> F["Carrier and scrap"]
    F --> G["Scrap chopper"]
    D --> H["Miss-feed sensor"]
    H --> D

The station sequence for a typical bracket:

Station · Operation · What it establishes

1 · Notch and pierce pilot holes · Strip location for every later station

2 · Pierce internal holes · Hole positions relative to pilots

3 · Idle · Die strength between heavily loaded stations

4 · Form / bend up · Bend lines, springback compensation

5 · Restrike / calibrate · Final angle and flatness

6 · Pierce remaining holes · Holes that must not distort during forming

7 · Cut-off · Part separation, burr side

8 · Idle / scrap cut · Carrier release

Pilots are the spine of the design. A pilot enters a hole pierced in station 1 and locates the strip to within ±0.01 mm at every subsequent station; without it, the cumulative feed error of a 0.05 mm-per-stroke feeder becomes a 0.3 mm error by station 7. The pilot must be the first element to engage and the last to withdraw, and it must be strong enough not to bend when a misfeed occurs — which is why pilot diameter is usually 1.5–2× the hole it locates, with a tapered lead.

Idle stations are not waste. A die block carrying a piercing punch and a forming punch side by side needs a minimum ligament between them — 15 mm is a common rule for high-rate dies — or it will crack through the web. An idle station buys that ligament, and it also gives the tool room a place to add a feature later without rebuilding the die.

Miss-feed detection is not optional. A single misfeed at 60 strokes a minute destroys punches, scraps the strip and often cracks the die block before the operator hears anything. The sensor — a probe in a pilot hole, a strip-edge sensor or a feed-length monitor — pays for itself the first time it fires.

The timing diagram is the document that proves the design. Every punch, pilot, stripper and cam has an entry and exit angle; anything that engages before the pilot has located the strip, or withdraws after the strip starts to feed, is a crash waiting for the third shift. A progressive die is a machine with a twelve-part sequence that must be correct on the first stroke and the ten-millionth.

5. Bending, Springback and the Overbend

A bend is not a fold. The neutral axis does not sit at mid-thickness; it shifts toward the inside of the bend as the radius tightens, which is why the flat blank must be longer than the sum of the flanges.

The bend allowance is:

with \theta in radians, R the inside bend radius, t the thickness and K the neutral-axis factor — about 0.33 for R/t < 1, 0.40–0.45 for R/t = 1–3, and 0.50 for R/t > 3. For a 90° bend of 2 mm CRCA at R = 3 mm, K = 0.42:

The flange flat length is the flange dimension minus the setback, and the blank is the sum of the flats plus every bend allowance. Get K wrong by 0.05 and a four-bend part misses its overall length by 0.6 mm — outside the ±0.3 mm a typical drawing allows.

Springback

On unloading, the elastic strain in the bend recovers and the angle opens. For a wide sheet in plane strain, the angular recovery is approximately:

where E_c is the effective unloading modulus. For mild steel at \sigma_y = 250 MPa, R_i/t = 4, E_c = 210 GPa:

For DP780 at \sigma_y = 550 MPa with the chord modulus nearer 160 GPa:

The radius also grows:

which for DP780 at R_i = 8 mm, t = 2 mm gives R_f = 8.34 mm — a 4.3% radius change that a hole-to-edge tolerance will not forgive.

The rule of thumb from the AHSS Guidelines matches this ladder: allow a wall-opening angle of about 3° for mild steel, 6° for DP350/600, and 10° for DP850/1000 or TRIP450/800. Note that these are section angles, not bend angles — a channel opens more than a single bend because sidewall curl and the two flanges add to the recovery. The die compensates by overbending, by coining the radius, or by a restrike station; the design decision is which of the three the part can afford.

The forming limits that decide feasibility

Check · Rule · Consequence if violated

Bend radius · R \ge 1.0t mild, 2.0t AHSS · Outside cracking

Flange height · h \ge 3t + R · Flange pulls out of the die

Hole-to-bend · \ge 2.5t + R · Hole distorts into an ellipse

Hole-to-edge · \ge 1.5t · Edge tears during piercing

Bend relief · depth \ge t + R · Tear at the bend terminus

Grain direction · bend \ge 45^\circ to rolling · Cracking on hard grades

A hole closer than 2.5t + R to a bend line is not a tolerance problem; it is a different process. The hole must be pierced after forming, or the part must be redesigned with a relief slot.

6. Deep Drawing: Where the Press Runs Out of Energy

Drawing is not bending. A flat blank is forced into a die cavity while its flange is restrained, and the metal flows inward under a blankholder. The limiting draw ratio is the ratio of blank diameter to cup diameter at which the wall tears:

Typical first-draw limits are 2.0 for mild steel, 1.8 for austenitic stainless and 1.6 for aluminium. A 150 mm blank into an 80 mm cup is a ratio of 1.88 — feasible in mild steel, marginal in stainless, and a two-stage draw in aluminium.

The punch force follows Siebel:

For d = 80 mm, t = 2 mm, \sigma_{uts} = 350 MPa and D/d = 1.8:

The blankholder force is separate and must be high enough to prevent wrinkling but low enough to let the flange flow:

with p = 1.5–3 MPa for mild steel and 4–6 MPa for AHSS. On a 150 mm blank with a 30 mm annular contact, A_{bh} \approx 11{,}310 mm², so F_{bh} \approx 34 kN at 3 MPa — about 17.5% of the draw force, and the number that is usually set by trial rather than calculation.

The energy is the part that catches people. A 193 kN draw over a 60 mm stroke is 5.8 kJ per stroke. At 30 strokes per minute the press needs 174 kJ/min, and the flywheel must supply most of it in the 0.3 s of the working stroke while the 22 kW motor replenishes it over the remaining 1.7 s. A press with the right tonnage and the wrong flywheel stalls on the third stroke. This is why press selection in §8 is an energy calculation with a tonnage check, not the reverse.

7. Hot Stamping: 600 MPa In, 1,500 MPa Out

Press-hardening steel is the one place in sheet metal where the forming operation is also the heat treatment. A 22MnB5 blank arrives ferritic-pearlitic at about 600 MPa. It leaves the die martensitic at 1,500 MPa with roughly 5% elongation — a strength level that cold forming cannot reach at any practical tonnage, because the flow stress at 900 °C is a fraction of what it is at 20 °C.

The process window is narrow and entirely thermal:

Step · Parameter · Why

Austenitisation · 900–950 °C, 3–10 min · Full austenite, homogeneous grain

Transfer · < 3 s · Must not drop below ~850 °C before forming

Forming · 850–400 °C · Low flow stress, above M_s

Quench · > 27 K/s · Diffusionless martensite

Martensite start / finish · M_s 425 °C, M_f 280 °C · The transformation window

Die contact pressure · ~40 MPa · Heat transfer and shape calibration

Cycle · ~20 s · Furnace dwell dominates

The cooling rate is set by contact, not by the blank's internal conduction:

With h \approx 3{,}000 W/m²K, a sheet at 700 °C against a die at 50 °C, \rho = 7{,}850 kg/m³, c_p = 500 J/kgK and t = 1.5 mm:

comfortably above the 27 K/s threshold. The thermal diffusivity of steel, about 4.5 mm²/s, means the through-thickness gradient equilibrates in a fraction of a second; the limiting resistance is the die interface, which is why cooling channel spacing, die steel conductivity and contact pressure — not furnace time — are the levers on cycle time.

The coating is a separate metallurgy. An Al-Si coating 23–32 µm thick protects the blank from oxidation and decarburisation, but its melting point is below the austenitisation temperature. Heating must be slow enough — around 12 K/s — for iron to diffuse into the coating and form a temperature-resistant Fe-Al-Si layer; heat too fast and the coating melts, sticks to the furnace rolls and destroys both the blank and the line. That is why hot stamping furnaces are 30–40 m long and why the process is capital-heavy even before the press.

India has moved into this process quickly. Hot-stamped content in domestically built passenger cars has reached about 60 kg per vehicle, against 10 kg six years ago, and now represents roughly 15% of the average body-in-white. Bharat NCAP crash requirements, not weight targets, are the driver — a B-pillar that must survive a side pole impact is a 1,500 MPa part or it is a thicker mild-steel part, and the hot-stamped version is lighter and cheaper at volume. Gestamp's fifth Indian plant, a USD 62.3 million investment in Gujarat, brought its footprint to 14 stamping lines including five hot-stamping lines.

8. Press Selection: Tonnage Is a Rating, Energy Is a Limit

The nameplate tonnage is measured at a specified point above bottom dead centre — commonly 0.187 in, 0.25 in or 0.5 in. Above that point the mechanical advantage of the crank or eccentric falls away, and so does the available force.

A 600-tonne mechanical press delivers 250 tonnes at 3 inches off bottom dead centre. A die that needs 400 tonnes through a 4-inch draw will not run in it, no matter what the nameplate says. The tonnage curve, not the tonnage number, is the specification.

Energy behaves the same way and is easier to ignore. A representative 600-tonne press has 25% of its rated energy available when it runs at 12 strokes per minute instead of 24, because the flywheel has less time to recover between strokes. The press does not slow down politely; it stalls mid-stroke.

For the 5.8 kJ draw at 30 SPM, P_{avg} = 2.9 kW — small against a 22 kW motor, but the flywheel must deliver it in the 0.3 s working stroke, which is a 19 kW instantaneous demand and a 40% flywheel speed drop if the inertia is undersized:

This is also why the press bed matters. Tonnage rating assumes the load is distributed over roughly two-thirds of the bed area; a small die placed in the middle of a large bed deflects the ram and the bedplate, opening the die clearance on one side and closing it on the other. Off-centre loading does the same thing in the other direction. The die is designed for a press, not for a tonnage.

Servo presses

A servo press replaces the flywheel and clutch with one or more servomotors driving the slide through a screw or a linkage. The consequences for stamping are not marketing points:

Constant energy at any speed. The slide can be slowed to 2 mm/s for the shear and accelerated through the return without losing available energy.
30–80% more energy than a comparable mechanical press, which is what allows AHSS draws that would stall a flywheel machine.
32–42% lower electrical consumption in field tests at 75–80% capacity, because the motor only draws what the stroke needs and can regenerate on the return.
Programmable dwell at bottom dead centre, which lets the material set, eliminates a restrike station and holds a C_p \ge 1.33 on a part that a mechanical press could not hold.
Snap-through control. Decelerating before breakthrough can nearly eliminate the reverse load that cracks punches.

The cost is capital and maintenance. For a job shop running mild steel at 60 SPM, a mechanical press is the right machine. For an AHSS structural part with a 4-inch draw and a ±0.05 mm tolerance, the servo press is not an upgrade; it is the only machine that will hold the process.

9. Tool Steel, Coatings and the Cost of Wear

A progressive die is a consumable with a long life, and its economics are set by how many strokes it delivers between grinds.

Tool material · Hardness · Life between grinds (indicative) · Where it belongs

D2 / SKD11 (1.2379) · 58–62 HRC · 200k–800k strokes · General blanking and piercing

Powder metallurgy (ASP23, Vanadis 4) · 60–64 HRC · 500k–1.5M strokes · Abrasive and high-strength strip

Carbide · 78–82 HRA · 1M–5M+ strokes · Long-run connectors, laminations

H13 (hot work) · 48–52 HRC · 500k–800k (hot) · Hot stamping dies

Coating is the cheapest life multiplier available, provided the substrate is hard enough to support it:

Coating · Hardness · Life gain · Best for

TiN (PVD) · 2,000–2,500 HV · 1.5–3× · General carbon steel

TiCN · 2,500–3,500 HV · 2–4× · Abrasive high-strength strip

AlCrN · 3,000–3,400 HV · 2.5–3.5× · Stainless, galling grades

DLC · 3,000–5,000 HV · 2–5× · Aluminium, low-friction work

None (bare D2) · 600–750 HV · 1× · Low volume

Two rules govern coating economics. First, coating a badly finished edge wastes the coating — the substrate must be polished to R_a < 0.2 µm and often nitrided before deposition. Second, regrinding removes the coating, so a coated die must be stripped, ground, re-polished and re-coated after every sharpening; budget 0.5–2 days of turnaround and re-coat after every second or third grind, not after every one.

The regrind itself removes 0.05–0.15 mm from the cutting face. Too little and the worn layer survives; too much and the die loses working height it will never get back. A D2 die can typically be reground 15–25 times before retirement, which is where the total life of 1–3 million strokes comes from.

The maintenance arithmetic is the part most quotes omit:

A ₹8,000 regrind plus four hours of lost production at ₹1,500/h is ₹14,000 per event. Spread over 300,000 strokes, that is ₹0.047/part — trivial. Spread over 30,000 strokes because the die is running abrasive stainless at 1.2 mm, it is ₹0.47/part — 14% of the conversion cost. Die life is a per-part cost, not a tool-room metric.

Galling and the stainless problem

Austenitic stainless work-hardens at the cut edge and adheres to the punch, a failure mode called galling. It shows up as a pickup on the punch face, a torn part edge and a rapidly rising punch load. The countermeasures are a harder, smoother tool (AlCrN or DLC), a heavier chlorinated or sulphured lubricant, a larger clearance and a slower shear. Aluminium has the same problem for a different reason — it is soft and gummy rather than work-hardening — and the same remedies apply, with CrN or DLC preferred over TiN.

10. The Tolerances a Die Can Actually Hold

A progressive die does not hold the tolerances a machining centre holds, and quoting it as if it does is how a tool gets built twice.

Feature · Typical capability · Notes

Pierced hole diameter · ±0.03–0.05 mm · Within one station, same punch

Hole position, same station · ±0.03 mm · Referenced to the same pilot

Hole position, across stations · ±0.08–0.12 mm · Accumulated pilot and feed error

Blank outline · ±0.05–0.10 mm · Depends on strip width control

Bent feature position · ±0.10–0.20 mm · Springback dominates

Bend angle · ±0.5–1.0° · ±2° on AHSS without a restrike

Flatness · 0.1–0.3 mm/100 mm · Oil canning on large panels

Burr height · ≤ 10% of t · 0.20 mm on 2 mm stock

Fine-blanked feature · ±0.01–0.02 mm · 0.5% clearance, triple action

The general tolerance fallback is ISO 2768-m, but a stamping drawing should never rely on it. Specify the critical features, the datum scheme and the burr side explicitly; leave the rest to the general note.

Oil canning deserves its own line. A large flat panel with residual compressive stress in the skin will snap between two stable curvatures under light finger pressure — it is flat when you measure it and not flat when the customer unwraps it. The cure is a stiffening bead, a controlled crown in the die, or a restrike; the cause is uneven draw or insufficient blankholder pressure. It is a design and process problem, not an inspection problem.

Process capability is the acceptance language:

At C_{pk} = 1.33 the expected defect rate is 63 ppm; at 1.0 it is 2,700 ppm; at 0.67 it is 44,400 ppm. A press shop that cannot show a C_{ppk} study on the critical dimension is telling you it has not run the part enough to know — which, for a new die, is exactly when the first-article inspection matters most.

11. The Indian Ledger: Tooling, Rate and the Crossover

The market context first. India's metal stamping sector was estimated at USD 5.2–5.4 billion in 2025, with automotive stamping alone forecast to grow from roughly USD 5.4 billion to USD 10 billion by 2035 at 6.3% CAGR. Press machines account for 39.5% of the metal-forming equipment market; servo presses are the fastest-growing category at 8.2% CAGR; carbon and low-alloy steel is 43.5% of the material formed. The Production Linked Incentive scheme has drawn USD 4.24 billion of cumulative investment and 48,974 direct jobs through September 2025. The clusters are Pune–Aurangabad–Mumbai, Chennai, and Gujarat.

A progressive die for a bracket of this size costs ₹3–15 lakh in India, with 8–12 weeks of lead time for a straightforward eight-station tool and longer if the part needs draw development. Add tryout: two to four prototype loops are normal, and each loop is calendar time, not cutting time.

Building the press hour

The rate is built the same way as any machine hour, and the arithmetic is worth doing once because it is the number the quote rests on.

Component · Annual basis · ₹/hour

Depreciation · ₹45 lakh over 10 years, 3,000 productive hours/yr · 150.00

Cost of capital · 8% of ₹45 lakh over 3,000 hours · 120.00

Floor space · 60 m² at ₹250/m²/month · 60.00

Power · 22 kW at 60% load, ₹8.50/kWh · 112.20

Maintenance · consumables, spares, service · 40.00

Die maintenance allocation · regrind and tool-room share · 60.00

Operator · ₹28,000/month loaded at 1.35, 176 hours · 214.77

Subtotal · · 756.97

Overhead at 20% · · 908.37

So ₹908/hour, or about ₹900/hour for a 100-tonne press on two shifts at 3,000 productive hours. At 45 strokes per minute and 75% OEE, that is 2,025 parts/hour — ₹0.45/part of machine time. At 90 SPM it is ₹0.22. This is the number that makes stamping look impossible to compete with, and it is real: the machine is not where the money goes.

The crossover

For the bracket, the laser-plus-brake route is about ₹45/part of conversion — laser cutting 420 mm of contour in 2 mm CRCA, four bends, deburr. The stamping route is ₹1.04/part of conversion including die maintenance and consumables, plus the tool:

At 1,000 parts the die adds ₹800/part and the part costs ₹809; at 10,000 it adds ₹80; at 50,000, ₹16; at 100,000, ₹8. The full cost curve, with ₹8.36 of material:

Quantity · Tool share · Material · Conversion · Cost/part

1,000 · ₹800.00 · ₹8.36 · ₹1.04 · ₹809.40

10,000 · ₹80.00 · ₹8.36 · ₹1.04 · ₹89.40

18,200 · ₹43.96 · ₹8.36 · ₹1.04 · ₹53.36

50,000 · ₹16.00 · ₹8.36 · ₹1.04 · ₹25.40

100,000 · ₹8.00 · ₹8.36 · ₹1.04 · ₹17.40

500,000 · ₹1.60 · ₹8.36 · ₹1.04 · ₹11.00

At the crossover the stamping cost equals the laser cost exactly, by construction. Below it the laser is cheaper; above it the gap widens by ₹44/part for every additional part. A ₹15 lakh tool moves the crossover to 34,100; a ₹5 lakh tool moves it to 11,400. A part with twice the cut length and a ₹90 laser cost moves it to 9,000.

The model also shows what quotes get wrong. The material is 33% of the 50,000-off price, 48% of the 100,000-off price and 76% of the 500,000-off price. A die that improves yield from 62.9% to 70% saves ₹0.85/part — at 100,000 parts, ₹85,000, roughly 10% of the tool cost. Yield is not a sustainability metric; it is a price line.

12. DFM Rules and the RFQ That Gets a Real Quote

The design rules that decide whether a part can be stamped at all:

Rule · Limit · Failure mode

Minimum pierced hole · 1.0t (mild), 1.2t (AHSS) · Punch breaks, slug pulls

Hole-to-edge · ≥ 1.5t · Edge tear, die chipping

Hole-to-bend · ≥ 2.5t + R · Hole distortion

Minimum bend radius · 1.0t mild, 2.0t AHSS · Outside cracking

Minimum flange · 3t + R · Flange pulls out

Bend relief · t + R · Tear at bend end

Straight edge for feeding · ≥ 1.5t land · Strip wanders, misfeed

Burr side · specified, not assumed · Rework or scrap at assembly

Grain direction · bend ≥ 45° · Cracking on hard grades

Corner radius · ≥ 0.5t · Stress concentration, die wear

An RFQ that produces comparable quotes contains: the drawing with datum scheme and burr side, the material grade and standard (IS 513 D/DC04, EN 10130, ASTM A1008), thickness and tolerance, annual volume and call-off, the tolerance band for each critical feature, the finish (bare, galvanised, painted), the press capacity available at the supplier, and whether the tool is customer-owned. The last item alone changes the price by the full tool amortisation.

The six questions a buyer should ask a stamper:

What press will this run in, and what is its tonnage at the working height? Not the nameplate — the curve.
What is the strip layout and the material yield? A quote without a yield number is a quote with a hidden material margin.
How many stations, and which ones are idle? Idle stations are die strength, not waste.
What is the die life between grinds, and who pays for regrinds? This separates a tool price from a tool cost.
What is the C_{pk} on the critical dimension, and on how many parts was it measured? A capability claim without a sample size is an opinion.
What happens on a misfeed? The answer tells you whether the tool has a sensor or a scrap bin.

Frequently Asked Questions

What is the difference between blanking, piercing and stamping?

Stamping is the family name for sheet-metal press forming. Blanking cuts the outer contour of a part from the strip — the cut piece is the part. Piercing cuts holes inside the part — the cut piece is scrap. Both are cutting operations inside a stamping die, and both use the same force and clearance arithmetic; they differ only in which side of the cut is wanted.

How do I calculate the tonnage for a stamping die?

Multiply the total cut length in mm by the thickness in mm by the shear strength in MPa, then divide by 1,000 for kN. For 420 mm of cut edge in 2 mm CRCA at 280 MPa, that is 235 kN. Add 10% for stripping and multiply by 1.3 for safety, then select a press at 1.5–2× the result. For forming, add the draw force from Siebel's equation and check the press energy, not just its tonnage.

Why does die clearance increase with material strength?

Clearance is the leverage that bends the slug and starts the crack ahead of the punch edge. A stronger sheet needs a longer lever to initiate that crack, so the gap grows from about 6% of thickness per side on mild steel to 14–16% at 1,400 MPa. Below the optimum the punch rubs and wears; above it the edge becomes a tear with excessive rollover.

Is a catalogue clearance per side or total?

It depends on the catalogue, and getting it wrong doubles or halves the gap. Most tooling suppliers quote the total punch-to-die gap; most technical guidelines quote the per-side value. Always confirm: a 15% total clearance is 7.5% per side, and a 15% per-side clearance is a 30% total gap that will produce a heavy burr and a rounded edge.

What burr height is acceptable on a stamped part?

The conventional target is below 10% of sheet thickness — 0.20 mm on 2 mm stock — with a slightly larger allowance on thick or soft material. On AHSS the burr is less useful as a wear indicator because the grade fractures before a large burr can form; burnish uniformity and the fracture-zone transition become the acceptance criteria instead.

When is stamping cheaper than laser cutting?

When the tooling amortises. For a ₹8 lakh progressive die against a ₹45/part laser-and-brake route, the crossover is about 18,200 parts. Below that the laser wins because it carries no tool; above it the die wins because its conversion cost is near ₹1/part and falls with stroke rate. The crossover moves with tool cost, part size and cut length — a ₹15 lakh tool pushes it past 34,000 parts.

Can a progressive die hold ±0.05 mm?

Within one station, yes — a pierced hole referenced to its own punch holds ±0.03–0.05 mm. Across stations, accumulated pilot and feed error typically opens that to ±0.08–0.12 mm, and formed features are worse because springback dominates. If the drawing needs ±0.02 mm across stations, the process is fine blanking or machining, not conventional stamping.

What is the difference between progressive and transfer stamping?

A progressive die carries the part on the strip through every station and cuts it free at the end. A transfer press moves a blank or a partly formed part between separate dies using mechanical transfer fingers, which allows larger parts, independent tooling per station and in-die operations such as tapping or welding. Progressive tooling is cheaper and faster for small parts; transfer is the route for body panels and large structural members.

Why does hot stamping produce 1,500 MPa parts?

Because the forming operation is also a quench. 22MnB5 is austenitised at 900–950 °C, formed while still austenitic, and quenched in a water-cooled die at more than 27 K/s, which suppresses diffusion and forces the austenite to transform to martensite. The as-delivered strength is about 600 MPa; the martensitic strength is about 1,500 MPa with roughly 5% elongation.

What material yield should a stamping quote assume?

Between 55% and 75% for most progressive parts; below 60% usually means the layout has not been nested, and above 85% usually means a simple straight-sided blank. The yield is fixed by the strip layout before the die is built, so it should be shown in the quote rather than buried in the material rate. On the worked bracket the yield is 62.9%, and the 37.1% scrap carries ₹3.10/part of coil value.

References

[1] AHSS Guidelines, "Cutting, Blanking, Shearing & Trimming," WorldAutoSteel — the four-zone sheared edge, the 6%–16% clearance range, the 10%-of-thickness burr target, ISO 16630's 12% test clearance, the half-thickness work-hardened zone, the 60% hole-expansion gain from a 3–6° bevel at 17% clearance, and the DP590 15% recommendation. [2] AHSS Guidelines, "Springback" and "Correcting Springback," WorldAutoSteel — the wall-opening ladder of 3° mild, 6° DP350/600 and 10° DP850/1000 or TRIP450/800, and the overbend, stake-bead and restrike methods. [3] AHSS Guidelines, "Tooling and Die Wear," WorldAutoSteel — D2/SKD11/1.2379 equivalence, coating families, the R_a < 0.2 µm substrate requirement, and the 10× coating life spread on 1.6 mm 1,180 MPa steel. [4] Amada, Die Clearance and Punching Force tooling catalogue — the total-versus-per-side clearance convention and the shear-strength bands for mild steel, stainless, aluminium and copper. [5] H. Karbasian and A. E. Tekkaya, "A review on hot stamping," Journal of Materials Processing Technology, 210(15), 2010, pp. 2103–2118 — 22MnB5's 600→1,500 MPa transformation, the 27 K/s critical cooling rate, M_s 425 °C, M_f 280 °C, and the 40 MPa die contact pressure. [6] R. Neugebauer et al., "Investigation of the thermo-mechanical properties of hot stamping steels," Journal of Materials Processing Technology, 2006 — austenitisation at 900–950 °C, the 23–32 µm Al-Si coating, the 3 s transfer, and forming between 850 and 400 °C. [7] M. Merklein and J. Lechler, "Investigation of the thermo-mechanical properties of hot stamping steels," CIRP Annals, 2006 — the 3.5-minute dwell for a 1.75 mm blank and the hardness-based austenitisation window. [8] "Performance Comparison of Zn-Based and Al–Si Based Coating on Boron Steel in Hot Stamping," Materials, 2021 — the 930 °C / 3–4 min schedule, 20 mm/s forming speed and the 5.4–6 s quench from 730–800 °C to 200 °C. [9] The Fabricator, "Servo-driven mechanical presses meet HSS head-on" and "Sizing up servo presses" — 30–80% more available energy, 32–42% lower electrical consumption, dwell at BDC, and reverse loads above 40% in HSS blanking. [10] AHSS Guidelines, "Servo Press Force and Energy Considerations When Working With AHSS" — the 600-tonne press curve (250 tonnes at 3 in off BDC) and the 75% energy loss when stroke rate halves. [11] Wikipedia, "Stamping press" — mechanical press full-tonnage rating points at 0.187 in, 0.25 in and 0.5 in above BDC, and the hydraulic-versus-mechanical trade-off. [12] The Fabricator, "Designing progressive dies" — carrier and scrap-bridge conventions and the one-to-two-thickness trimming allowance. [13] Die design standards (SAPCO), strip layout — the 1.25t minimum bridge, the 15 mm minimum ligament between die holes, and idle-station practice. [14] Strip-layout economy factor and scrap-bridge conventions, press-tool practice notes — E = A_b n/(Wp), the 60% minimum target, and the edge-scrap table by thickness. [15] Fine-blanking process literature (ASM Handbook, Vol. 14B; Feintool technical papers) — 0.5% per-side clearance, the V-ring, 10–30 MPa holding pressure, 4–15 mm/s shearing speed, and the 85–100% burnish face. [16] Tool-life data, D2/PM/carbide ranges and regrind practice — 0.05–0.15 mm removal per grind, 15–25 grinds, coating strip-and-recoat cycles, and the 200k–800k stroke interval for D2. [17] Mordor Intelligence, India Metal Forming Equipment Market, 2026 — press machines at 39.5% of the market, servo presses at 8.2% CAGR, hot-stamped content of 60 kg per Indian passenger car versus 10 kg six years earlier, 15% of body-in-white, Gestamp's fifth plant (USD 62.3 million, 14 lines, 5 hot-stamping), and the PLI's USD 4.24 billion of investment and 48,974 jobs through September 2025. [18] Future Market Insights and Claight, India automotive/metal stamping market estimates — USD 5.2–5.4 billion in 2025 and the 6.3% CAGR to 2035. [19] Indian progressive-die tooling quotations, 2026 — ₹3–15 lakh for an eight-station bracket tool, 8–12 weeks lead time, 30–400 tonne press range. [20] ISO 2768-1:1989, General tolerances; ISO 16630:2017, Hole expansion test; ISO 6892-1:2019, Tensile testing; IS 513, Cold reduced carbon steel sheet and strip; EN 10130, Cold rolled low carbon steel flat products; VDA 238-100, Bend angle test — the standards behind the tolerance, clearance and material tables.

Note on sources: the governing relations are F_b = L t \tau_s for cutting force, c = k_c t for clearance, E = A_b n/(Wp) for strip yield, L_b = \theta(R + Kt) for bend allowance, \Delta\theta/\theta = 3\sigma_y R_i/(E_c t) for springback, F_d = \pi d t \sigma_{uts}(D/d - 0.7) for draw force, \dot T = h(T_s-T_d)/(\rho c_p t) for die quenching, and N^ = C_{tool}/\Delta c for the tooling crossover. Every ₹ figure is derived from the parameter set stated in the text — a ₹45 lakh 100-tonne press, 3,000 productive hours, ₹8.50/kWh, a ₹28,000 loaded operator wage, ₹62/kg CRCA and a ₹8 lakh eight-station die — and will move with those inputs. The method transfers; the numbers must be re-derived for a specific shop, city and year.*