Minimum bend radius by material and thickness

A 3 mm austenitic stainless bracket drawn with a 2 mm inside radius will usually form on the first bend and crack on the third. The radius is 0.67 × thickness, inside the practical range for CRCA, and well below what 304 will take once it has work-hardened twice. Minimum bend radius by material is the table that prevents that sequence, and it is best read as a multiple of thickness rather than a fixed number of millimetres.

The rule that fits most press-brake work: inside radius of 1.0 × t for cold-reduced mild steel, 1.5 × t for 304 stainless and 5052 aluminium, and 2.5–3.0 × t for 6061-T6. Read the table below before choosing a radius, then check whether the punch tip can produce it, because in air bending the tool decides the radius the part ends up with.

Minimum bend radius by material is a multiple of thickness

Doubling the thickness doubles the radius the material needs, because the strain on the outside of the bend depends on the ratio of radius to thickness, not on either value alone. A drawing that calls R2 on 1.5 mm CRCA (1.33 × t) is comfortable. The same R2 on 3 mm CRCA (0.67 × t) is at the edge of what forms cleanly and will show orange-peel or a fine crack on a sheared edge.

Inside, outside or neutral: say which

Three radii describe one bend. The inside radius is the concave face. The outside radius is inside radius plus thickness. The neutral radius sits somewhere between, at the K-factor line, and it is the one the flat pattern is calculated from.

Tooling is quoted by inside radius. Bend tables are written by inside radius. A drawing that calls an outside radius, or leaves the reference open, is the first source of a disagreement at first article. State "inside radius" or "R inside" on the drawing, and put the value in millimetres after a material and thickness callout the shop can read.

Why the multiplier changes with material

The outer fibre of a bend has to stretch. Formability is governed by how much stretch the material can take before it necks, and by how much it work-hardens on the way. Cold-reduced mild steel at 28–38% elongation takes a tight bend without complaint. Austenitic stainless has higher total elongation but work-hardens steeply, so the second and third bends are forming already-hardened metal. Heat-treatable aluminium in the T6 condition has 8–12% elongation and very little reserve.

The galvanised row is not a material property. Zinc cracks before steel does at a tight radius, and a cracked coating on the outside of a bend becomes a corrosion site within a monsoon. Adding half a thickness to the radius, or specifying a bend line that avoids the worst of the coating damage, is cheaper than a coating repair after forming.

Copper and brass need their own rows because they fail differently. Annealed copper at 0.5 × t bends like putty and then marks at the slightest die contact, so the tool surface and the handling between operations decide the finish as much as the radius does. Half-hard brass at the same thickness needs 1.0–1.5 × t and is far more sensitive to bend line direction: a bend along the rolling direction can open a visible grain-line crack that the same bend across the grain does not. Both are cut on the 12 kW laser and deburred before forming, which removes the edge feature that would otherwise start the crack.

The outer-fibre strain behind the table

For a bend with inside radius R and thickness t, and a neutral axis at mid-thickness, the engineering strain at the outer surface is 1 / (1 + 2R/t). At R = 1 × t that is 33%. At R = 2 × t it is 20%. At R = 3 × t it is 14%.

Read that against elongation and the table explains itself. A 6061-T6 at 8–12% elongation has no reserve at the strain a 1 × t radius implies, which is why it needs 2.5–3.0 × t, where the calculated figure falls to 14–17% and the margin is still thin enough that the first few parts are made as a cracking trial. CRCA at 30% elongation survives the 33% figure because the estimate assumes a mid-thickness neutral axis and uniform strain, and a real bend concentrates strain over a short arc with a gradient through the thickness.

The formula is a ranking tool, not a pass-fail gate. Use it to compare two candidate radii on the same part, and use the table to pick the starting point.

Bend line direction and the cut edge

Two parts cut from the same sheet can behave differently because of where the bend line sits relative to the rolling direction. A bend line running along the rolling direction puts the highest strain on the grain structure that is least able to take it, and the crack appears at the ends of the bend, where the strain is highest and the material is least restrained.

Bend across the rolling direction where the layout allows. Where it does not, add 0.5 × t to the radius rather than argue with the mill.

The cut edge matters as much on thin sheet. A 12 kW fibre laser cuts cleanly but leaves a narrow oxide and heat-affected layer on the cut face, and that layer is harder and less ductile than the parent metal. On bends below 1.0 × t, a laser-cut edge can be the crack initiation site even when the material itself is fine. A light deburr, or a radius that keeps the outer fibre away from the edge, removes the risk.

Springback changes what you inspect

At the same radius-to-thickness ratio, 304 stainless springs back roughly twice as far as CRCA, and 6061-T6 more again. A 90° bend programmed to 90° finishes at 88–89° in CRCA and 86–88° in stainless.

That difference is why a bend table is per grade and per thickness, not per machine. The Delem control stores a table per material, and the operator proves it on the first part of a run. A part that measures correctly in CRCA and is then ordered in 304 has to be re-proved, and the flat pattern re-checked, because springback changes both the finished angle and the developed length.

The K-factor and bend allowance guide explains how the neutral axis position changes the flat length; the bend allowance calculator turns the same values into a blank size in a few seconds.

Measuring the radius after forming

Radius gauges are fast and good to about ±0.2 mm, which is enough to confirm that a 3 mm radius is 3 mm and not 5 mm. They are not enough to prove a minimum-radius callout, because the gauge reads the best-fitting arc over a short contact length and a slightly elliptical bend can pass two adjacent gauges.

For a critical bend, measure the profile on a shadowgraph or against a traced template, and record the thickness at the same point. Thickness tolerance matters here: a 3 mm nominal sheet that arrives at 2.85 mm has a radius-to-thickness ratio about 5% higher than the drawing assumes, so a tool that was comfortable on nominal stock is marginal on the thin end of the tolerance.

Measure before coating. A powder film of 60–80 µm sits on both faces and adds to an outside-radius reading, which is why a coated part often appears to have a radius 0.1 mm larger than the formed part underneath.

Minimum flange and the die the radius implies

A radius that is fine on its own can still be unformable because the flange is too short to sit on the die shoulders. The practical minimum flange is half the V-opening plus the thickness, and the V-opening follows from the radius you want.

This is where the radius decision meets the tonnage decision. A tighter radius needs a narrower die, and a narrower die needs more force per metre for the same thickness. On a long part, that force may exceed what the bed can carry, and the answer is a larger radius rather than a bigger press. The minimum flange length rules set out the geometry, and the press tonnage calculator checks whether the chosen die fits the machine.

How Ganesh holds the radius on a repeat part

Radius work runs on the 200 T, 5-axis CNC press brake with a 4,450 mm bed, Delem control and capacity to 10 mm. Tooling starts at a 0.6 mm punch tip, which covers most 0.8–1.5 mm enclosure work at 0.5–1.0 × t. For a repeat part, the tool set, the die opening, the bend table and the ram penetration per bend are recorded against the part number, so a radius proved at first article is the radius produced at part 2,000.

Blank preparation matters as much as the brake. The 12 kW fibre laser with its 40 ft × 16 ft bed cuts a 4,000 mm blank in one pass, so the bend line runs parallel to a cut edge that is straight, and the flange length measured from that edge is consistent along the part. After forming, parts go through seven-tank pretreatment and the 18 ft × 14 ft powder booth; the inside radius is measured before coating, because a 60–80 µm film changes a micrometer reading on the outside face.

Ganesh has formed sheet metal since 1992 across five units in Mumbai (Vasai) and Nagpur, holds ISO 9001:2015 and ISO 14001:2015, and works to ±0.5 mm typical on formed sheet metal, tighter on critical features by agreement. The full route sits under sheet metal bending.

What to put on the drawing

• Material grade, temper or condition, and thickness with its tolerance.
• Inside radius in millimetres, marked "R inside", per bend if they differ.
• Bend line direction relative to the rolling direction, where the part is tight.
• Flange lengths, with any flange below 4.5 × t flagged.
• Bend angle and the tolerance, with the measurement method.
• Whether the outside face must remain unmarked, which drives die choice and film.
• Hole-to-bend distance, since a hole near the bend distorts (the hole-to-bend distance rules give the multipliers).
• Any radius that must be held after coating, not before.

The bending and welding assemblies hub collects the rest of this cluster, from bend allowance and flange minima to springback and inspection.

When the drawing needs a different route

A radius below the table is a material or process decision, not a tooling decision. Ganesh does not anneal or stress-relieve parts after forming, so a design that depends on softening between bends needs a different temper from the mill, a different grade, or a formed-then-heat-treated route with a supplier who does both.

Three other cases leave the press brake. A radius that is smaller than 0.5 × t on thin sheet at high volume belongs in a press tool, where the form is coined rather than air-bent. A constant-section profile a few thousand metres long belongs in roll forming. And a bend longer than 4,450 mm cannot be made in one hit on our bed, so it needs a splice or a different supplier.

Ganesh does not roll-form profiles, does not heat-treat after forming, and does not quote a radius below the material table without saying so in the quotation. If the radius is critical to the function of the part, send it with the drawing and the mill certificate for the heat.

Send the drawing with the grade, temper, thickness and the radius in question. The engineering team will confirm whether the radius is formable on the tooling, whether the punch tip can produce it, and what the bend table will look like for the grade — and the RFQ checklist lists the rest of what helps.