Wireforms and springs

Springback in wire forming: why the first bend is never the final angle

How elastic recovery changes the angle and the radius of a wire bend, how to estimate it, and how to set an overbend that holds across a production lot.

QKQumek Engineering Desk··9 min read·Updated
R 4.0Ø 3.00

Key takeaways

  • Springback is elastic recovery. It changes both the angle and the radius when the tool releases.
  • Yield strength drives springback, so temper and wire lot matter as much as the tool setting.
  • An overbend is a process setting, not a drawing value. It belongs in the setup record, not the title block.
  • A functional gauge or a load test catches springback drift better than a single angle measurement.

What springback actually is

When a wire is bent, part of the deformation is plastic and part is elastic. The plastic part stays after the tool is removed. The elastic part recovers, and the bend opens slightly. The finished angle is therefore smaller than the angle the tool commanded, and the finished radius is larger than the tool radius.

Both changes matter. A clip that should sit at 90 degrees may finish at 87 degrees. A hook that should have an inside radius of 2 mm may finish at 2.1 mm. On a single bend the error is small. On a part with eight bends, the angle errors accumulate along the part and the last leg can finish well outside its position.

Springback is not a defect. It is the normal behaviour of the material. The process deals with it by overbending: the tool is set past the required angle so that the recovered part lands on the drawing. The skill is in setting the overbend so that it holds across a whole lot, not just on the first piece.

The variables that set the amount

Four variables explain most of the variation. The first is the yield strength of the wire. A higher yield means more elastic recovery, so spring temper wire springs back more than annealed wire of the same alloy. The second is the elastic modulus. Steel sits around 200 to 210 GPa, stainless around 193 GPa, copper around 120 GPa and brass around 100 GPa. A lower modulus means less springback for the same strength.

The third is the ratio of the bend radius to the wire diameter. A large radius bends a longer length of wire, so the elastic recovery has more length over which to act, and the angle loss grows. The fourth is the wire diameter itself. A thicker wire is stiffer in bending, so the same tool force produces less curvature and the recovery is a smaller share of the total.

Tool geometry and machine compliance sit on top of those four. A tool that flexes, a gripper that slips or a bend pin that is slightly undersize all add to the apparent springback. That is why a calculated overbend is a starting point, not a final setting.

Springback drivers and the practical response
DriverEffect on springbackWhat to do
Higher yield strengthMore recovery, larger angle lossSet the overbend by trial on the lot
Lower elastic modulusLess recovery for the same strengthExpect copper and brass to need less overbend than steel
Larger radius to diameter ratioMore angle lossCheck the angle tolerance on large-radius bends
Larger wire diameterSmaller share of recoveryStandard settings are more stable on heavy wire
Tool or machine complianceAdds to the apparent recoveryCheck the tool and the setup before blaming the wire

An estimate you can use at the drawing stage

A full springback model needs the stress strain curve and the tool geometry. A useful approximation for a round wire bent into full plasticity is that the change in curvature is about 3.4 times the yield strength divided by the product of the elastic modulus and the wire diameter. Curvature is one divided by the radius.

For an illustrative example, take 1.2 mm music wire with a yield strength of 1,600 MPa and a modulus of 207,000 MPa. The curvature change is about 3.4 times 1,600 divided by 207,000 times 1.2, which is about 0.022 per millimetre. If the tool radius is 1.8 mm, the tool curvature is 0.556 per millimetre. The finished curvature is 0.534 per millimetre, which is a radius of about 1.87 mm.

For the angle, the arc length stays roughly constant, so the finished angle is the commanded angle multiplied by the ratio of the finished curvature to the tool curvature. In this example a 90 degree bend finishes at about 86.5 degrees, a loss of about 3.5 degrees. The real loss may be larger because of tool compliance, so treat the number as a floor.

The estimate is for planning, not for setting

Use it to see whether a bend is sensitive, to compare two materials and to sanity check a supplier's overbend. Set the machine on the actual wire lot, then record the setting.

Overbend and the setup record

Overbend is the difference between the angle the machine commands and the angle the drawing asks for. On a CNC former it is a program value. On a manual or four-slide machine it is a tool or cam setting. Either way, it belongs in the setup sheet, not on the drawing, because it changes with the wire lot, the tool wear and the machine.

A good setup sheet records the overbend at each bend, the tool radius, the wire lot number and the first-off angle. When a new lot arrives, the operator checks the first piece against the record. If the angle has moved, the overbend is adjusted before the run continues. That is a controlled response, not a guess.

The radius needs the same treatment. If the drawing calls an inside radius, the tool radius may need to be smaller so that the recovered part lands on the nominal. Record both. A tool radius that is copied from the drawing without allowance will produce a part that is consistently oversize on the radius.

Why springback drifts across a lot

Wire is drawn to a tensile range, not to a single strength. Two coils of the same grade and diameter can sit at different points in that range. The higher strength coil springs back more. If the machine is set on the first coil and the second coil is at the top of the range, the angles move.

Diameter tolerance adds a second effect. A wire at the top of the diameter band is stiffer in bending and changes the radius to diameter ratio. A plated wire has a coating that may add to the effective diameter and change the friction in the tool. A change of supplier or a change of drawing route can move the springback even when the certificate says the same grade.

The practical control is lot traceability. Record the coil or lot number against the run, keep the first-off angle and the overbend with it, and check a sample when the lot changes. If the part is critical, ask the supplier to hold the wire to a narrower tensile band, but expect to pay for the selection.

When and how to measure the finished angle

Measure after the part has been released from the tool and allowed to settle. For most wire parts, a few minutes is enough. For a part that has been formed at high strain, or one that will be plated, measure again after plating. Plating adds a layer and can change the fit, and the plating process can also relax residual stress.

Measure the angle the function uses. If the part is a clip, that may be the gap between two legs, not the angle of a single bend. If the part is a spring, the free angle and the load at a stated deflection are more useful than a stack of angles. Use a gauge that holds the part the way the assembly holds it, because a part measured free can sit differently when it is constrained.

Record the temperature if the part is measured in a workshop that moves through the day. Steel expands by about 11 parts per million per degree Celsius, which is small on a 50 mm part but not zero over a 20 degree swing. For most wireforms this is a second-order effect, but it matters when an angle is already at the edge of its tolerance.

Design tactics that reduce sensitivity

A design that depends on a precise angle from a single bend will always be more sensitive than one that shares the load across two bends or uses a stop. If the function needs a leg to sit at a fixed angle, consider a feature that locates the leg mechanically: a hook that engages a pin, a flat that sits against a face, or a second leg that shares the load.

A larger radius reduces the strain but increases the angle loss. That trade is usually worth taking when cracking is the risk. A shorter leg reduces the effect of angle error on the tip position, because the tip moves less for the same angular error. Both changes are cheap at the drawing stage and expensive after tooling.

Finally, put the angle tolerance where the function needs it and let the other angles carry a general tolerance. A drawing that gives every angle the same tight band tells the supplier nothing about which bend matters, and it usually attracts a higher price than the part needs.

  • Add a mechanical stop or a locating feature where the angle is critical
  • Keep the load-bearing leg as short as the function allows
  • Use a larger radius where cracking is the risk, and accept more angle loss
  • Tolerance the functional angle, not every angle on the view
  • Specify the load at a stated deflection for spring parts
  • Agree the overbend record and the lot change check with the supplier

Frequently asked questions

Why does a 90 degree bend finish at 87 degrees?

The wire recovers elastically when the tool releases. The plastic part of the bend stays and the elastic part springs back, so the angle opens. The machine compensates with an overbend. The exact loss depends on the yield strength, the modulus, the wire diameter and the radius.

Does springback affect the radius as well as the angle?

Yes. The finished inside radius is larger than the tool radius. If the drawing calls a nominal radius, the tool may need to be slightly tighter so that the recovered part lands on the nominal. Record both the tool radius and the finished radius.

Can springback be eliminated?

Not in a normal forming operation. It can be reduced with a softer temper, a smaller radius to diameter ratio or a post-form stress relief, but some recovery will remain. The practical approach is to control it with an overbend and to verify it on the lot.

Why did the angle change when we switched wire supplier?

The new wire may sit at a different point in the tensile range, or the diameter may be at a different point in its band. Both change the springback. Check the certificate and the measured diameter, then re-set the overbend on the new lot.

Sources and further reading

  • EN 10270-1, Steel wire for mechanical springs, patented cold drawn unalloyed spring steel wire
  • ISO 6892-1, Metallic materials, tensile testing at room temperature
  • Qumek forming setup and first-off records for wire parts
QK

Qumek Engineering Desk

Engineering team

The engineering team behind Qumek drawings, DFM notes, tolerances and inspection standards. Every guide is reviewed against the drawings and processes used on live programmes.

Chasing an angle that will not hold?

Send the bend details, the wire lot and the measured angles. The engineering desk will comment on the overbend, the tool radius and the material range.

Send the measurements

The Qumek dispatch

One useful email a month

Drawing notes, supplier checks, tariff changes and the questions buyers keep asking. No newsletter filler, no daily drip.

One email a month. No spam, no list sharing.