An Expert in Decoiler Feeder & Metal Coil Processing

Home / All / NC Servo Straightener Feeder Guide / How Does Strip Crown Change Straightener Settings?

How Does Strip Crown Change Straightener Settings?

Nov 18,2026

Strip crown turns a straightener into an uneven machine, because the roll gap is straight across the width and the strip is not.

On a 0.8 mm CR job last spring, a 0.012 mm crown produced a 7 mm wave over one metre, and the setting sheet had nothing on it that could fix that.

The case is worth walking through in order, because the diagnosis took three days and the fix took ten minutes once it was found.

A Wave That Appeared Only in the Middle

The plant ran 0.8 mm cold-rolled sheet, 1,250 mm wide, on an NC straightener feeder at 45 SPM with a 380 mm feed.

The part was an appliance side panel with a 4 I-unit flatness callout, roughly 4 mm of wave over a metre.

Finished panels came off with a visible ripple in the middle 400 mm and dead-flat edges. The wave measured 7 mm over 1,000 mm, about 12 I-units.

Every conventional answer had already been tried. Penetration was increased, then reduced. Roll gaps were checked with feeler gauges at both ends.

The wave got worse at the end of each coil and better at the start, which sent everyone looking at the decoiler brake.

Strip crown measured across a coil before an NC straightener feeder setting change

What the Roll Gap Sees Across 1,250 mm

A straightener sets one gap. The strip arrives at a different thickness in the centre than at the edges, and that difference is the whole problem.

Measured with a micrometer at five points across the width, this coil ran 0.812 mm at the centre and 0.804 mm at 50 mm from each edge.

The gap was set at 0.780 mm. Penetration at the centre was therefore 3.9 percent of thickness, and 3.0 percent at the edges.

A third more work in the middle of the strip, repeated across five or six bend reversals, elongates the centre relative to the edges.

The arithmetic predicted about 30 I-units of differential elongation over a metre. The strip showed 12.

That gap between prediction and measurement is normal. The edges drag the centre back, so only part of the difference becomes visible wave.

The direction was right, the magnitude was three times optimistic. That is enough to identify the cause and not enough to size the fix.

Measure crown with a micrometer and a light touch, never with a caliper across the edge. The burr alone reads 0.01 mm on a slitted edge.

Take readings at 50 mm, quarter width, centre, and the mirrored positions. Six points describe the profile well enough to set a gap.

Three Measurements That Settled It

The plant had been measuring flatness after the line. Nobody had measured the strip before it.

  • Micrometer at five points across the width, on strip from the coil head, middle and tail
  • Cut a 1,000 mm sample before the straightener, lay it on a plate and check for existing centre wave
  • Read the mill certificate for crown and profile tolerance, and compare it with what the micrometer says
  • Run one coil with the gap opened to 0.800 mm and re-measure the wave height
  • Photograph the wave with a straight edge and a scale in frame, so the number survives the meeting

The certificate said crown 1.2 percent, which matched the micrometer. The sample cut before the straightener was already rippled by 3 mm.

So a third of the wave was in the coil and two thirds was being added by the line. Both facts mattered for the fix.

The opened-gap run dropped the wave from 7 mm to 5 mm, which confirmed that penetration was amplifying a material problem rather than creating one.

Four Fixes, Priced by What They Cost

Once crown is confirmed, the options are narrow and each one has a price that is easy to state.

CorrectionEffect on the centre waveWhat it costs you
Reduce penetration to the minimum that still flattens7 mm down to 5 mm on this jobFree, ten minutes, but residual bow rises and the part may not sit flat
Fit a crowned roll set matched to the strip profile5 mm down to 2 mmUSD 6,000 to 9,000 and three to four weeks, and the crown is fixed for one material
Buy strip to a tighter crown, 0.8 percent or betterRemoves about a third of the wave at sourceTwo to four percent on coil price, and eight to ten weeks of mill lead time
Add back tension at the entryHelps 30 to 40 percent on light gaugeUSD 2,000 to 4,000 for a brake and control, and it needs a loop to work against
Engineer comparing crowned straightener roll profiles for coil straightening

This plant took the first and third options together. The gap went to 0.790 mm and the mill crown spec went to 0.8 percent.

Wave fell to 3 mm, inside the 4 I-unit callout, and stayed there across coil ends. Total spend was a coil price increase and no capital.

What Changes on the Setting Sheet

Crown work is not a single number change. It is a different way of reading the same five settings.

SettingWhat it should say on a crowned coilWhere it misleads you
Entry gapThickness minus 0.01 mm, not minus 0.03The book value assumes uniform thickness and over-works the centre
Exit gapSet for flatness on a plate, measured at the strip edge as well as the middleMeasuring only the middle hides an edge wave forming
Back tensionEnough to keep the strip on the rolls, no moreExcess tension flattens the wave temporarily and it returns in the part
Line speedReduce 15 percent for the first 30 m of each coilCoil-head crown is often the worst, and speed hides it until the press
Roll conditionCheck crown wear on the rolls every 12 monthsA worn roll set adds its own profile to the strip and reads as material

Row three is where most plants push back. Tension makes the sample look good and does nothing for the part.

Our coil line knowledge base sets flatness out as a roll-geometry problem first, and crown is the input that geometry cannot see.

Telling Crown Apart From a Slack Edge

Both faults put a wave in the middle of the strip. They need opposite corrections, so the distinction is worth ten minutes.

Cut and lay it flat. A 1,000 mm sample taken from the coil head and laid on a plate shows the wave that was already in the material.

If the wave is there before the straightener, crown or a slack edge is upstream. If it only appears after, the line is adding it.

Look at the wave pitch. Crown gives a long, shallow wave, usually one or two across the full width.

A slack edge gives a tighter ripple closer to the centre line, and it often runs at an angle rather than parallel to the edge.

Measure both edges. Crown leaves both edges the same thickness. A slack edge comes from one edge being thinner, and the micrometer finds that within 100 mm of the edge.

Check the slit burr. A heavy burr on one side means the slitter ran badly, and the stress pattern follows the burr into the line.

Roll diameters and available crown profiles for each model are listed in the straightener specification pages, which is the right reference before ordering a crowned set.

Limits of Any Crown Correction

A straightener can redistribute stress. It cannot change thickness, and that boundary decides most of these cases.

Plain cylindrical rolls cannot fix crown. No gap setting makes a 1.2 percent profile behave like a 0.5 percent one.

A fixed roll crown suits one material. A stack crowned for 0.8 mm CR will over-correct 2 mm HR and under-correct 0.5 mm stainless.

Centre wave is not always crown. A slack edge from a bad slit produces the same symptom, and a crowned roll makes it worse.

Coil-to-coil variation defeats any setting sheet. If crown runs 0.8 to 1.6 percent across a delivery, one setting cannot serve both.

FANTY builds to a 45,000 m² plant with more than 80 engineers, and the flatness conversations that end well are the ones where the mill was brought into the specification.

Four questions come back from every plant that finds crown in the middle of a flatness argument.

How much crown is normal in 0.8 mm cold-rolled coil?

One to two percent of thickness is common, so 0.008 to 0.016 mm. Below 0.8 percent is a tight spec and usually carries a price premium of two to four percent.

Does a 9-roll stack handle crown better than a 5-roll stack?

No, and it can be worse. More reversals amplify the differential elongation, so a 9-roll machine on 1.2 percent crown adds more centre wave than a 5-roll at the same gap.

Can back tension remove 0.012 mm of crown effect?

It removes 30 to 40 percent of the visible wave on 0.8 mm strip, which is 2 to 3 mm of the 7 mm measured here. It does not change the thickness profile.

At what width does crown start to matter?

Above about 600 mm. Narrow strip hides the differential because the centre-to-edge distance is short, which is why 300 mm lines rarely see this problem.

Where to go next

Crown is one input among several that decide whether a flatness spec is achievable on a given machine.

The flatness pillar works through roll count, pitch, gap and penetration together, which is where this case should have started.

Read the flatness pillar

Follow us to Get Latest Updates

subscription