What Causes Camber in an NC Straightener Feeder?

Two Edges, Two Lengths

Take a strip that has been bent through the straightener and imagine it as two independent ribbons joined along the centre. If the straightener treats both ribbons identically, they come out the same length and the strip runs straight. If the treatment differs even slightly, one ribbon ends up longer and the strip curves.

How slight is slight? On a 300 mm wide strip, a length difference of 0.03 percent between the two edges, which is 0.3 mm over a one-metre length, produces visible camber. That is a difference the eye cannot see on the machine and a feeler gauge can barely find. It is also why camber problems so often survive several rounds of adjustment: the operators are looking for something visible when the cause is a tolerance-level asymmetry.

The practical consequence is that camber cannot be corrected by changing the leveler gap overall. Raising or lowering the whole upper roll bank changes both edges equally. Camber responds to differential adjustment: roll tilt, roll crown, or a difference in the material itself.

Five Sources, Ranked by Frequency

Across the lines I have helped commission, camber traces back to a short list of causes. The order below reflects how often each one turns out to be the real answer after the obvious checks are exhausted.

Note the shape of this table. Only one of the five sources lives inside the straightener. The other four come from the coil, the brake, the mandrel or the entry guide. That distribution is the reason camber gets misdiagnosed: the adjustment that would fix a roll tilt is the one most operators reach for, and it does nothing when the cause is a strip profile from the mill.

How Roll Crown Turns Into Camber

A straightener roll is not a cylinder. It is machined with a slight crown, meaning the diameter at the centre is a few hundredths of a millimetre larger than at the ends. The crown exists to compensate for roll deflection under load, so that the strip sees uniform pressure across its width when the machine is working.

That compensation is only correct at one load. Run a narrower strip and the roll deflects less, so the centre of the strip receives more pressure than it needs. Run a wider or harder strip and the roll deflects more, so the edges are under-worked. Both conditions create a differential bend across the width, and differential bend creates a length difference, and a length difference creates camber.

The number that matters is the crown value. A typical straightener roll carries a crown of 0.03 to 0.08 mm over its face depending on roll diameter and bearing span. If that crown is ground to 0.02 mm, the roll under-compensates at full width and the strip edges receive less work than the centre. If it is ground to 0.12 mm, the centre is over-worked and the strip curves the other way. Neither error is visible on a straightedge; both show up in the strip.

This is why camber problems often appear after a roll regrind that was performed correctly to the drawing but with the wrong crown. The rolls look perfect, measure round, and produce a strip that curves.

Why Camber Changes Through a Coil

If camber is constant from the first metre to the last, the cause is mechanical and static. If it drifts, the cause is geometric and changing. Most camber complaints fall into the second group and get treated as the first.

Three things change as a coil unwinds. The coil radius falls, which changes the payoff geometry and the entry angle into the straightener. The back tension required to hold the loop falls with it, which changes the tension the strip carries through the roll bank. And the mandrel's grip on the inner wraps is different from its grip on the outer wraps, which can bias the strip sideways by a fraction of a millimetre as it leaves the coil.

On a line running 1.0 mm material at 300 mm width, I have measured camber going from 1.2 mm per metre at the coil head to 3.5 mm per metre at the tail on the same job with no setting changed. The cause turned out to be mandrel runout of 0.35 mm combined with a falling back tension, so the strip entered the straightener at a slightly different lateral position at the head and the tail. Fixing the mandrel brought the tail figure down to 1.6 mm per metre, and the remaining drift was handled by tightening the brake control band.

This is why a camber measurement taken only at the coil head is not evidence that the job is under control.

Isolating the Source in One Shift

You can separate the five sources with four measurements and no specialist equipment. Work in this order, because each step removes a variable.

• Cut three 1,000 mm samples at the coil head, middle and tail. Lay each on a surface plate and measure the maximum deviation from a straight line
• Micrometer the same samples at five points across the width. A thickness variation above 0.02 mm makes the strip a suspect before the machine is
• Feeler gauge every straightener roll gap at both ends with the machine stopped and the rolls closed to a known setting
• Dial-indicate mandrel runout with a loaded coil, then repeat with an empty mandrel to separate mandrel error from coil winding error
• Run the same coil twice, once at normal tension and once at half tension. Camber that improves at low tension points at the roll crown, not the alignment
• Record all six numbers in the job file. The next time the job runs, the comparison tells you which measurement moved

Step five is the one that saves the most time and gets skipped most often. Roll crown and roll tilt produce similar camber figures but respond to opposite corrections, and the tension sensitivity test separates them in fifteen minutes.

Correction Options and What They Cost

Once the source is known, the fix is usually straightforward. What varies is the cost and the downtime.

Roll tilt is the cheapest to correct. Re-shimming and re-levelling the roll bank is a half-day job with the machine stopped, and the only tooling required is a precision level and a dial indicator. Entry guide misalignment is cheaper still. Unequal tension is a control and maintenance matter: brake pads, loop sensor calibration and mandrel runout all fall into a normal service budget.

Roll crown is where the money goes. Regrinding a straightener roll requires removing it, which means dismantling the roll bank, and the roll must be reground to the correct crown rather than simply made round. A roll pair reground flat will produce camber that no amount of tilt adjustment removes. If the rolls are near the end of their diameter tolerance, replacing rather than regrinding is the better decision, and the replacement should be specified with the crown value for your strip width range rather than the builder's default.

Strip profile from the mill is the case with no machine-side fix. If the strip arrives with a measurable thickness crown, the straightener can only compensate, not correct. The right move is a conversation with your coil supplier about a crown specification and a check on whether the mill certificate reports the profile. On the lines we have delivered, more than 200 of them, the camber cases that ended in a mill conversation rather than a machine adjustment were about one in five, and they were the ones that had been chased longest on the shop floor.

Twelve years of building coil handling equipment has taught me that camber is rarely the straightener's fault and almost always the line's problem. Measure the strip before you adjust the machine, and you will find it faster.