Brass for sheet metal parts: grade, thickness, formability

Brass for sheet metal parts is specified for one of three jobs: carrying current, surviving outdoors without a coating, or drawing deeper than steel will before it tears. Whichever of those the part needs sets the grade, and the copper content — not the colour of the sample — is the number that decides whether the drawing can be made at all.

Most brass drawings stop at "brass, 1 mm". That line lets the mill supply anything from 62% to 71% copper in any temper. The gap between those two ends shows up as a split flange at the die, a bend that springs 4° off square, or a part that tarnishes green in three months outdoors.

Which brass grade does a formed part need?

CuZn30 (UNS C26000) for anything drawn or bent hard, CuZn35–CuZn37 (C27000, C27200) for general fabrication, and CuZn40 (C28000) is not a cold-forming sheet grade at all.

The mechanism is phase structure. Below about 35% zinc the alloy is single-phase alpha, which deforms by slip and takes 55% elongation in the soft temper. At 40% zinc a body-centred beta phase appears; beta is hot-workable and cheaper per kilogram, but it does not bend cold without cracking at the outside of the radius. A designer who specifies a cover in CuZn40 because the extrusion grade was cheap has specified a part that will split on the first fold.

Work hardening is the second half of the choice. Brass has no yield plateau, so every forming pass raises tensile strength by roughly 80–120 MPa. A part that is blanked, folded and then flanged has already used two of the three forming operations the temper allows. That is why the temper code on the mill certificate — O60, H01, H02, H04 under ASTM B36 — matters as much as the grade.

Brass for sheet metal parts: grades, tempers and shop limits

The grades below are the ones that arrive on Indian purchase orders. The Ganesh column is the route we would quote, not a promise that every grade is stocked.

Two rows in that table decide most arguments. C46400 is the right answer for a salt-water bracket but is hard to buy as thin sheet in India, so the design often moves to 316L stainless or to a coated steel part instead. Leaded brass is common in turned fittings and rarely appropriate in sheet; at 0.5–3% lead it breaches the 0.1% RoHS threshold unless an exemption covers the application.

Thickness, weight and the bend allowance you have to recalculate

Brass at 8.5 g/cm³ is 8% denser than CRCA at 7.85 g/cm³, so a like-for-like gauge change adds mass to the assembly even though the part looks identical.

A 300 × 200 mm blank in 1 mm C26000 has a volume of 60,000 mm³, which is 60 cm³, which weighs 511 g. The same blank in 1 mm CRCA weighs 471 g. On a lift car door made of four such panels the difference is 160 g — small on its own, but it lands on a counterweight that was calculated at the start of the project.

The bend allowance is where brass surprises people, because the neutral axis sits further out than in steel. Using the standard development,

• BA = θ × π/180 × (R + K·t)
• BD = 2 × (R + t) × tan(θ/2) − BA

for a 2 mm C26000 O60 part with a 2 mm inside radius and a 90° bend, K is about 0.45. That gives BA = 1.571 × (2 + 0.9) = 4.56 mm. The setback is (2 + 2) × tan 45° = 4.00 mm, so the bend deduction is 8.00 − 4.56 = 3.44 mm. Two outside dimensions of 100 mm therefore need a flat blank of 196.6 mm, not 200 mm.

Use K = 0.50 for H02 and H04 tempers. A half-hard part folded with the soft-temper K-factor comes out 0.2 mm long on a 90° bend of this size, which is enough to miss a 0.5 mm hole-to-edge callout on a small bracket.

Mill thickness tolerance is the other input. On 1.0 mm brass sheet the mill tolerance is typically ±0.05 mm, and that ±5% feeds straight into the bend deduction, because K·t and R + t both depend on the actual thickness. A press brake with a Delem control stores the correction per material and thickness; a shop that runs brass on a steel programme without that correction will chase the same 0.3 mm all week.

What the temper does to springback

Springback is the elastic recovery of the bend after the punch lifts, and in brass it is the difference between a bracket that closes square and one that has to be tweaked by hand.

Soft C26000 recovers about 1–2° on a 90° air bend at a 1t radius. Half-hard C27200 recovers 3–4°, and full-hard sheet can recover 5–6°. The recovery is proportional to yield stress divided by modulus, so the harder temper always springs more even though the alloy is the same. Two compensations exist: overbend by the measured angle, which the controller can do per step, or coin the bend at the bottom of the stroke, which needs roughly four times the tonnage and leaves a visible witness mark on soft brass.

For a part with three bends in series, springback accumulates. A 3° recovery on each of three 90° bends leaves the last flange 9° out of position relative to the first, and no amount of fixture pressure at assembly will pull that back without stressing the panel. Set the first-off against the drawing, record the overbend, and keep that number with the programme rather than with the operator.

How thick does brass sheet go before the route changes?

Brass sheet is normally stocked from 0.3 mm to 3 mm, and the mass per square metre follows directly: 8.5 kg at 1 mm, 17 kg at 2 mm and 25.5 kg at 3 mm. A 1 m² panel at 3 mm already weighs as much as a small steel frame, which changes how it is handled on the brake and how it is packed.

Above about 3 mm the material is usually supplied as plate rather than sheet, the flatness tolerance widens, and laser cutting becomes the only sensible profiling route because the shearing forces on a guillotine distort the edge. At 6 mm and above, brass is a specialist purchase: the mill lead time is measured in weeks, the price moves with the copper content, and a design that uses 3 mm with a formed rib is usually the better engineering answer than one that uses 6 mm flat.

How brass behaves under the laser and the brake

Brass cuts on a fibre laser, but not like steel. At 1,070 nm the polished alloy reflects 60–70% of the beam, against about 95% for copper, so absorptivity is three to five times better than copper and still far worse than the 60–70% absorbed by oxidised mild steel.

The second effect is zinc. Zinc boils at 907 °C, below the 900–940 °C melting range of the alloy itself, so the cut front is a zinc plume as much as a melt. That gives a clean, nearly dross-free edge on 1–3 mm sheet with nitrogen assist, and it also means the machine needs extraction at the nozzle rather than at the bed. Brass fume is a health control, not a nuisance.

On the brake, brass needs a larger V opening than steel for the same thickness — around 8t rather than 6t — because the alloy spreads under load and marks easily against a sharp die shoulder. Soft brass also picks up tool marks on the outside face, so a polished or brushed finish has to be protected with a urethane film or formed before the finish is applied.

Joining is the real limit. Brass is brazed with a copper-zinc or silver-bearing filler at 600–800 °C, or soldered, or riveted and clinched. It is not fusion welded in production: the zinc boils out of the weld pool and leaves porosity, and the fume causes metal fume fever in anyone standing close. Spot welding fails for a different reason — at 28% IACS the sheet conducts the current away faster than a nugget can form.

What Ganesh runs for brass parts

Brass arrives as sheet, is nested and cut on the 12 kW fibre laser with its 40 ft × 16 ft bed, formed on the 200 T, 5-axis CNC press brake with a 4,450 mm bed, and where the volume justifies it, drawn in press tools built in our own tool room. Our custom part range runs 0.8–40 mm in thickness, and the laser covers brass, copper, stainless, aluminium and CRCA on the same floor.

Volume changes the route. Below a few thousand parts a year, laser cut and brake formed is the sensible route and the tooling cost is zero. Above that, a blanking or draw tool starts to pay, and the tool is quoted separately from the piece price so the amortisation is visible rather than buried. The press side of that decision is covered in copper and brass pressed components, and the cutting constraints in laser cutting copper and brass.

Ganesh has supplied brass and copper parts alongside steel fabrications for customers including Mahindra and Godrej & Boyce since 1992, across five units in Mumbai (Vasai) and Nagpur, under ISO 9001:2015 and ISO 14001:2015. Brass work is a small share of that output; the capability record behind it is on the copper, brass and aluminium fabrication page.

What to put on the brass drawing

• UNS number and temper, for example C26000 O60, not "brass sheet".
• Thickness with the tolerance, because ±0.05 mm changes the bend deduction.
• Inside bend radius per bend, and whether it is a forming limit or a clearance.
• Grain direction relative to the bend line if the part is drawn or bent past 90°.
• Flat pattern or developed length, with the K-factor you used, so the first article can be checked against it.
• Finish: bare, powder coated, plated, or lacquered, with the surfaces that must stay conductive.
• Joining method and filler, since brass does not arc weld.
• Quantity per year and per release, because that decides laser versus tooling.

The DXF and RFQ checklist covers the file side of the same list, and the materials hub carries the equivalent guidance for stainless, aluminium and CRCA. If the grade is still open, send the drawing through request a quote and we will price the forming route with the grade question answered.

Where brass is the wrong call

Brass is a poor structural material. Its modulus is about 110 GPa against 200 GPa for steel, so a brass bracket of the same thickness deflects nearly twice as far under the same load, and matching steel stiffness needs about 26% more thickness and 36% more mass.

It is also the wrong choice wherever the part has to be welded into an assembly, wherever the service environment carries ammonia or warm chloride, and wherever the drawing is really asking for conductivity — at 28% IACS, brass carries about a quarter of the current of copper for the same section, and a copper busbar is the honest answer. The full decision rule, with the crossover numbers, is in when brass is the wrong material.

If the part is a flat cover with no electrical or corrosion duty, CRCA with powder coating is cheaper, stiffer and easier to weld. Sending a brass drawing because the previous supplier used brass is not a reason to keep it.