How Do You Hit a Flatness Spec on a Coil Line? Roll Count, Pitch, Gap and Penetration

The reason the unit matters is that a straightener does not remove curvature, it trades large curvature for small curvature spread over many rolls. If you state a target as "flat" you have no way of knowing whether the machine in front of you has one more useful adjustment left or none.

Before anyone touches a handwheel, five values belong on a single sheet of paper: the flatness figure and the span it is measured over, the material grade and yield strength, the thickness range you actually run rather than the nominal, the strip width, and where in the line the check happens. Everything later in this article is downstream of those five.

• The number and its span. "8 I over 1 m" is actionable. "8 I" is not, because the same physical wave reads differently over 500 mm and over 2 m.
• Yield strength, not just grade name. Two coils both described as "high tensile" can sit 150 MPa apart, and that gap moves the penetration setting more than a full roll of adjustment.
• The real thickness range. If the line runs 0.8 mm to 3.0 mm, the straightener is being asked to work two different machines. Settings that suit one end will mark or under-work the other.
• Strip width. Width sets how much of the distortion is edge-driven. Narrow strip tolerates settings that would ripple a 1,250 mm coil.
• Where the check happens. Flatness measured at the straightener exit and flatness measured after the feed rolls and the die are two different numbers, and the second one is always worse.

The Settings That Actually Decide Flatness

Seven things influence what comes out of the straightener. Two of them are bought, five are set. Confusing the two is the most common reason a line that used to run flat stops running flat after a job change.

How Many Rolls You Actually Need

Roll count is the single most consequential decision and the one buyers shop on hardest, because it is visible on a specification sheet and it moves the price. The rule behind it is straightforward: each pass through a roll reverses the bending direction, and every reversal removes some of the residual curvature. More reversals, flatter strip.

The catch is that the return per roll falls away quickly. Going from five rolls to seven buys a large improvement. Going from eleven to thirteen buys a small one, and costs you a longer machine, more drive power and a harder threading job.

Two practical warnings. First, roll count cannot be added later, so if the flatness target is close to the machine's limit, the honest answer is a different machine rather than a better setup. Second, a high roll count does not compensate for a bad roll pitch, which is the next subject and the one that catches buyers who compare specifications line by line.

Roll Diameter, Pitch and the Geometry Trap

Roll pitch is the centre-to-centre distance between adjacent rolls. It matters because rolls have to overlap or nearly overlap to bend the strip, and two rolls of a given diameter cannot sit closer together than their own diameter without colliding.

High-yield strip wants more penetration, more penetration needs more room, and more room means either fewer rolls in the same body or a longer body.

Roll diameter then sets how hard each bend is. Thin strip needs a small bend radius to be worked at all, because the strain a roll imposes is proportional to thickness divided by radius. A 60 mm roll bends 0.5 mm strip properly and barely touches 4 mm strip.

Strip thickness	Typical roll diameter	Implication for pitch	What breaks if it is wrong
0.3 to 1.0 mm	Small, often 40 to 60 mm	Pitch can be short, so many rolls fit in a compact body	Rolls too large simply skate over the strip. Penetration goes in and nothing changes at the exit
1.0 to 2.5 mm	Medium, around 60 to 90 mm	The mainstream compromise between bend radius and machine length	Too short a pitch leaves no adjustment travel, so the last few rolls cannot contribute
2.5 to 4.5 mm	Large, 90 to 140 mm	Pitch must grow with diameter, so the machine gets long	Undersized rolls deflect under load and the flatness drifts as the coil runs down
Above 4.5 mm	Large and heavily supported	Usually a leveller rather than a straightener feeder	A three-in-one machine stretched to this gauge loses the compactness that made it worth buying

The trap is that two machines can both be described as "nine-roll" and behave completely differently, because one has 90 mm rolls on a long pitch and the other has 60 mm rolls packed tight. If you are comparing quotations, ask for roll diameter and roll pitch alongside the roll count. A supplier who cannot quote those two numbers is quoting a category, not a machine.

Setting Roll Gap From Strip Thickness

Roll gap is the clearance between the upper and lower roll sets. It is the first thing a new operator reaches for, and the setting most often left too tight. Tightening the gap feels like doing more work because the machine sounds busier and the ammeter climbs.

What is actually happening is that the rolls are pinching the strip rather than bending it.

A workable starting point for a fresh job is the nominal strip thickness, then remove a small amount and close up in steps while watching the exit strip rather than the load meter. The exact figure depends on roll diameter, so treat the numbers below as orientation and confirm them against the strip in front of you.

Material	Starting gap	Adjustment direction	First symptom of going wrong
Mild steel, SPCC or DC01	Close to nominal thickness, slightly under	Close up in small steps until the bow disappears, then back off a quarter turn	Surface marking and a bright polish line down the strip
Semi-hard steel, SPHC	Modestly under nominal	Close up more gradually, and expect to need more penetration rather than a tighter gap	Roll load climbs faster than flatness improves
HSLA 340 to 550	Near nominal, do not over-tighten	Bias towards penetration and roll count instead of gap	Edge cracking and a dull, over-worked surface finish
Stainless 304	At or just under nominal	Work with light contact. Stainless work-hardens under repeated pinching	Strip gets harder as it runs, and later passes stop having an effect
Aluminium 5052 or similar	Near nominal	Prefer fewer, gentler passes. Soft alloys mark easily	Roll pickup and transfer marks that then print through the die
Pre-painted or coated strip	At nominal, never below	Use penetration and pass line, never gap, to do the work	Coating scuffing that cannot be adjusted out and becomes a reject

A rule that saves arguments on the floor. The gap is a clearance, not a setting for flatness. Once the gap is at a sensible value for the thickness, further flatness comes from penetration and pass line. If a fitter is still closing the gap to chase a wave, the wave is being caused by something else.

Roll Penetration and How It Is Distributed

Penetration is how far a roll pushes past the strip's neutral line, expressed as a depth or as a percentage of the roll's adjustment range. It is the setting that does the actual straightening work, and it is the one most often reduced to a single number when it should be a pattern.

On a multi-roll straightener the penetration is not uniform. The entry rolls take the strip's incoming curvature and need the most work. The middle rolls continue the reversal. The exit rolls are doing a much lighter job, mostly relaxing the bend the previous roll left behind.

If all rolls are set to the same depth, the strip leaves the last roll with a residual curl in the direction of that final bend, which is why operators end up with strip that is flat in the middle and curled at the tail.

Penetration pattern	Effect on the strip	Where it stops working
Entry rolls deepest, falling towards the exit	The standard arrangement. Incoming curvature is flattened early and the exit rolls relax it	Needs enough rolls to hold the gradient. On a five-roll machine there is little room to taper
Uniform across all rolls	Easy to set and repeat, and works acceptably at loose flatness targets	Leaves a residual curl at the exit matched to the last roll's direction
Decreasing then slightly increased at the last roll	Used to fine-tune a stubborn long bow that survives the main taper	Easy to overdo. A small error here reintroduces the defect you just removed
Heavy penetration across the board	Flattens aggressively, and hides a wrong pass line	Over-works the surface, consumes drive power, and can crack high yield or stainless strip
Very light penetration across the board	Leaves the machine essentially acting as a guide	The strip passes through unchanged. Operators then tighten the gap, which is the wrong correction

The practical check is to look at the exit strip immediately after the straightener, before the feed rolls touch it, and then look again at the tail of the coil. If the head is flat and the tail curls, the penetration gradient is wrong rather than the overall depth.

If both head and tail curl in the same direction, the last roll is doing too much work and the pass line is the next thing to examine.

Pass Line Height and Back Tension

The pass line is the height at which the strip travels through the roll stack. It has to line up with two things at once: the height the strip leaves the decoiler at, and the height the feed rolls or the die want to receive it at.

When the entry is higher or lower than the coil pay-off point, the strip enters the rolls already bowed, and the straightener spends its first few rolls removing a curve it created itself.

Pass line errors are quiet. Nothing alarms, nothing looks broken, and the machine simply needs more penetration than the specification suggested to reach the same flatness. That extra penetration then costs drive power and roll wear for the life of the job.

Symptom	Likely pass line or tension cause	Correction	The cost of leaving it
Persistent long bow in one direction	Entry height does not match the coil pay-off height	Set the entry guide height to the natural pay-off line before touching roll settings	Every job on the line needs extra penetration, so power and roll wear are permanently above what was quoted
Strip snakes from side to side at the entry	Combination of low back tension and a pass line that lets the strip wander	Raise brake tension slightly and re-check the entry guide clearance	Edge damage and guide wear, and a coil that will not track straight through the feed rolls
Loop collapses between decoiler and straightener	Back tension too low for the line speed	Increase decoiler brake torque, or slow the acceleration ramp	The strip hits the floor, picks up grit, and that grit prints through the die as a defect
Coil is pulled off the mandrel rather than unwound	Back tension higher than the decoiler brake can hold smoothly	Reduce brake torque and let the straightener do the work	Coil slip and mandrel wear, which is the subject of the companion guide on how a decoiler mandrel grips a coil
Flatness good at low speed, worse at full speed	Tension control responds faster than the straightener setting can absorb	Add acceleration ramp and check whether the loop control is hunting	You lose the top of the speed range, and the line runs below the throughput it was bought for

Matching Settings to Yield Strength and Grade

Yield strength is what makes a straightener either work well or work hard. A higher yield material resists bending, so it needs more penetration to be straightened, and it springs back further once the rolls release it.

Those two effects pull in opposite directions, and the balance point is what a good operator is actually looking for.

The consequence is that settings do not transfer between grades. A recipe that produces flat SPCC will under-work HSLA 420 in the same thickness and leave it bowed, and a recipe that flattens HSLA 420 will over-work and mark SPCC. If a line runs a mixed coil diet, the practical answer is a written recipe per grade rather than an operator's memory.

Material class	Typical thickness	Setting implication	Cost when the setting is copied from a softer grade
Mild steel, 200 to 250 MPa	0.5 to 3.0 mm	Moderate penetration, fewest rolls needed, forgiving of small errors	Little. This grade tolerates a recipe borrowed from almost anywhere
Semi-hard, 300 to 350 MPa	0.8 to 3.0 mm	More penetration and a slightly wider gap than mild steel	Residual bow that shows up as a leaning part after the press, traced late to the straightener
HSLA 340 to 420	1.0 to 4.0 mm	Needs roll count and penetration reserve. Gap stays near nominal	Under-worked strip, plus a temptation to close the gap, which cracks edges instead of flattening
HSLA 550 and above	1.5 to 5.0 mm	Treat as a different process. Expect to reduce line speed and monitor roll load	Roll surface damage and edge cracks that only appear after forming
Stainless 304	0.5 to 3.0 mm	Light contact, more passes, avoid repeated pinching because the material work-hardens	Each subsequent pass does less, and the operator keeps adding penetration to no effect
Aluminium, 5052 or 3003	0.5 to 3.0 mm	Gentle settings, clean rolls, attention to surface finish above flatness	Roll pickup marks that transfer through the die and become the defect that matters

One more variable sits underneath all of these: incoming coil quality. Straightening removes curvature that the coil supplier created and that your own coil handling may have added. If the coil arrives with real crossbow or a heavily cambered edge, the straightener will correct what it can and pass the rest downstream.

That is why coil telescoping and edge damage in the warehouse end up as flatness complaints at the press.

Defect, Cause and Correction Matrix

Most flatness problems arrive with a name already attached, usually because a previous operator had the same shape and called it something. The value of a matrix is that it forces the diagnosis back to a cause before anyone reaches for a handwheel, because the same visible shape can come from three very different places.

Defect	What you see	Most likely cause	First correction	When it is not a straightener problem
Camber	The strip curves sideways as it leaves, so the coil wanders across the line	Thickness or yield variation across the strip width, or edge damage on one side	Check the strip across the width and correct the pass line and entry guide before roll settings	If one edge measures consistently thicker, the straightener cannot fix a coil that was rolled unevenly
Edge wave or edge ripple	Short-pitch waves running along one or both edges	Edges are being elongated more than the centre, usually because the gap is too tight at the edge	Open the gap slightly and increase penetration so bending rather than pinching does the work	A wave that survives a full penetration taper is usually incoming material, not setup
Centre buckle	Waves running down the middle of the strip with flat edges	Centre is longer than the edges after rolling, and the straightener is not working the centre hard enough	Add penetration and check that the rolls are parallel across the face	Roll parallelism drift is mechanical. A worn bearing will produce the same shape and no setting will hold
Crossbow	The strip is curved across its width, like a shallow channel	Uneven reduction during rolling, and often a pass line that is off centre	Confirm entry and exit heights are level, then work the bow out gradually over more rolls	Heavy crossbow can need a dedicated leveller. It is difficult to remove with a straightener alone
Twist	The strip rotates about its own axis along its length	Asymmetric residual stress, frequently combined with camber	Correct camber first. Twist usually reduces on its own once camber does	Twist that persists after camber is gone points to a coil defect or a badly set entry guide
Longitudinal curl	The strip coils up like a spring along its length	Penetration gradient ends too deep at the last roll	Taper the penetration so the final rolls relax rather than bend	A curl that reverses when you flip the strip is a setup issue. One that does not may be material
Springback-driven bow	Flat at the straightener, bowed again after the feed rolls or the die	High yield material springs back between the straightener and the forming station	Increase penetration to exceed the elastic limit properly rather than just touching it	If the gap between straightener and die is very long, the strip will bow regardless of settings
Diagonal or zig-zag buckles	Waves running at an angle across the strip	Combined longitudinal and transverse stress, usually a gap that is far too tight	Open the gap and rebuild the setting from the penetration side	If it appears immediately on a new coil and vanishes on the next, suspect the coil
Roll marking	Bright polish lines or a repeating pattern along the strip	Gap too tight, roll pressure too high, or contaminated rolls	Open the gap, clean the rolls, and check roll surface condition	Pre-painted and mirror-finish stock marks at settings other material tolerates. Coating choice is a machine specification, not a setting
Coil break	A line of transverse cracks, often at the coil's tighter bend radius	Too much penetration applied too quickly, or a coil stored in a tight wrap	Reduce penetration on the entry rolls and increase it gradually across the stack	A coil break already in the material cannot be straightened out. It can only be detected before forming

Two habits make this matrix far more useful in practice. Keep a sample of each defect taped to the machine with the setting written on it, so a new operator can match a shape rather than guess. And when a defect appears, change one thing at a time.

Two adjustments at once may clear the symptom while leaving you with no idea which one worked, which means it cannot be repeated next week.

Where Straightening Setup Goes Wrong

The failures below are the ones we see repeatedly across different customers, different countries and different end products. None of them are exotic. All of them are expensive in a way that does not show up on the same day the mistake is made.

Mistake	What actually happens	What it costs	How to avoid it
Chasing flatness with the gap	Gap closes, roll load rises, strip gets marked and sometimes cracked, and the wave is still there	Reject coils, roll resurfacing, and a machine running well above its designed load for no benefit	Fix the gap once for the thickness, then work with penetration. Treat the load meter as a warning, not a target
Specifying roll count without roll pitch	Two "nine-roll" machines arrive and only one can reach the target because the other has no penetration room	A machine that has to be replaced or worked around for its entire service life	Require roll diameter and roll pitch in every quotation, and compare them, not just the roll count
Ignoring the pass line during installation	The line runs, the strip is acceptable, and every job needs more penetration than the setting sheet says	Permanently higher power draw and roll wear, and a machine that never quite hits its quoted flatness	Set entry and exit heights against the real pay-off line as a commissioning step, and record the values
Copying a recipe between grades	A setting that flattened SPCC is applied to HSLA 420 and leaves a bow, or is applied in reverse and over-works soft strip	Several coils of suspect material and a quality complaint before anyone suspects the recipe	Hold a written setting sheet per grade and thickness, signed off when it was proven
Trusting light angle instead of measurement	Strip looks flat in the workshop and fails inspection under panel lighting	Rework, repaint, or a customer return on an otherwise correct part	Define the measuring span and the lighting in the specification, then check under those conditions
Adjusting two things at once	The defect disappears and nobody knows which change fixed it, or whether it will reappear on the next coil	The same problem returns within days and the diagnosis starts from zero	One change, one observation, then record it. Slower for a shift, much faster over a month
Blaming the straightener for a die problem	Roll settings are pushed further and further out of range trying to fix a part that is bowing in the press	Roll and bearing wear, plus the real die or feed issue never gets fixed	Measure flatness at the straightener exit and again after forming. If the first is good and the second is not, the straightener is not the cause
Buying rolls to fix incoming coil quality	A higher roll count is ordered because the coil supply is inconsistent, and the extra rolls still cannot remove real crossbow	Capital spent on capability that the material problem will still defeat	Fix what the coil supplier delivers first. Straightening is a finishing operation, not a repair operation

Worth saying plainly, because it affects how you read the tables above. The guidance here comes from our own shop floor and from the coil lines our NC straightener feeders run in. Typical values differ between roll diameters, line speeds and material sources, so treat every number as a starting point to be confirmed on your own strip rather than a specification to be adopted unchanged.

Verifying Flatness at the Machine

You do not need a metrology lab to hold a flatness target, but you do need a repeatable method and a consistent place to stand. Most disputes about flatness come down to two people measuring in two different places under two different lights.

Method	What it detects	Practical resolution	When it is not enough
Cut sample on a surface plate	Overall bow and any visible wave in a defined length	Good for anything above roughly 5 I in the hand	Cannot see fine ripple and depends heavily on the plate and the lighting
Straightedge with feeler gauge	Gap between a straight edge and the strip surface	Down to about 0.05 mm across a 1 m straightedge	Only covers one line across the strip. A wave in the next 50 mm is invisible
Shadow or low-angle light check	Short-pitch waves and edge ripple that the eye misses in flat light	Very sensitive, and the fastest field method	Not a number. Two people will disagree unless the light and angle are fixed by procedure
String or wire across a cut length	Camber, the side-to-side curve of the strip	Reads camber of a few millimetres over a few metres reliably	Only for camber. Says nothing about longitudinal flatness
Five-point thickness check	Thickness variation across the width, which often explains a wave	Enough to prove whether a defect is incoming or generated	Needs a micrometer and discipline. It is skipped far more often than it should be
Check after the feed rolls and die	The flatness the part actually has	The number that matters to the customer	It cannot separate straightener error from feed or die error on its own

Whichever combination you use, write down four things on the setting sheet: the measuring span, the light or plate used, the location in the line, and the person who signed the sample off. That record is what turns a one-off good coil into a repeatable process, and it is what an operator reaches for when a defect appears at two in the morning.

Straightening Knowledge Base

This page covers the setup decisions. The thirty-six guides below go deeper on individual threads: roll geometry and how to set it, what each flatness defect means and how to clear it, how strip behaves under load, and how the straightener fits into the rest of the line. Each one is written for a specific question rather than as background reading.

1. Roll geometry and setup

Roll count, diameter, pitch, gap and penetration. These are the settings that decide how many times the strip is bent, and how hard.

• How many straightening rolls a line needs
• Choosing the roll pitch
• Setting the roll gap
• How much roll penetration to use
• Why pass line height matters
• Whether five rolls can handle 4 mm
• What roll pressure to run
• Which work roll material lasts longest
• Maintaining straightening rolls

2. Defect diagnosis: cause and fix

Camber, edge wave, centre buckle, crossbow, twist, curl and springback. Each guide starts from the shape and works back to the cause.

• Diagnosing and fixing camber
• What causes camber
• Preventing edge wave
• Fixing centre buckle
• Removing crossbow
• Fixing twisted strip
• Fixing strip curl
• Fixing springback in high-strength strip
• Preventing strip scratch

3. Strip behaviour and material response

What the strip does under load: flatness, residual stress, coil breaks, die life, and how high-strength grades change the answer.

• How flatness actually improves
• The flatness a straightener can achieve
• Reducing strip stress
• Preventing coil breaks
• Extending die life
• Whether straightening changes material properties
• Running HSLA steel
• How back tension affects the result
• How servo straightening boosts precision

4. Maintenance, drive power and accuracy limits

The work either side of the setting: roll maintenance, motor sizing, throughput, and where the machine stops being the limiting factor.

• Maintaining an NC servo straightener
• What drive power is needed
• Setting up for fast stamping
• Servo versus hydraulic straightening
• Choosing between two straightener types
• Troubleshooting accuracy problems
• The accuracy a straightener can achieve
• Which part limits accuracy
• Selecting the right model

Part of the 3-in-1 decoiler straightener feeder cluster, alongside the coil handling and feeding accuracy guides.

Frequently Asked Questions

How do I know when a straightener has run out of adjustment?

Watch the penetration reserve rather than the flatness reading. If the defect improves as you add penetration and then stops improving while the roll load keeps climbing, the machine is at its limit.

As a working rule, if you need more than roughly 80 per cent of the available penetration travel to hold the target, there is no room left for the next grade or the next coil. The honest answer at that point is a machine with more rolls.

How many straightening rolls do I need for 3 mm HSLA 420?

Nine rolls is the practical floor for that combination at a visible-panel flatness target, and eleven is comfortable if the target is tight or the grade varies.

Three millimetres of high yield strip needs enough reversals to work the section without cracking it, and a seven-roll machine will typically leave a residual bow that no amount of penetration resolves. Confirm against a strip sample rather than the grade name alone.

What flatness can a seven-roll machine realistically hold?

On mild steel up to about 3 mm, a well-set seven-roll straightener commonly holds in the region of 10 to 15 I over a one-metre span. Push the thickness or the yield strength up and that figure worsens quickly, often to the point where visible waviness returns.

A nine-roll machine typically holds under 10 I on the same material. Treat these as field observations, not guarantees, because roll diameter and pitch change the answer more than roll count does.

Should the roll gap ever be set below the strip thickness?

Only slightly, and only as a clearance correction rather than a flatness adjustment. On mild steel a small amount under nominal is normal. On pre-painted, coated or stainless strip the gap should sit at or marginally above nominal, because anything tighter marks the surface or starts the material work-hardening.

If you find yourself closing the gap well below thickness to remove a wave, the wave is coming from penetration, pass line or the incoming coil.

Why is the strip flat at the straightener but bowed after the press?

Three candidates, in order of likelihood. High yield material springing back across a long gap between the straightener and the die; feed rolls set with uneven pressure pulling the strip into a curve; or a die that is not sitting level. Measure flatness at the straightener exit first.

If that reading is good, the straightener is doing its job and further roll adjustment will only make the first reading worse.

Can pre-painted strip be straightened without marking it?

Yes, but the setting sequence changes. Keep the gap at nominal so the rolls are not pinching, do the work with penetration, keep the pass line accurate so the strip is not dragged across a roll, and use rolls in good surface condition with a finish suited to coated stock.

The remaining risk is a machine specification question: mandrel and roll contact materials designed for coated strip are what prevent the marking that no setting can remove.

Where to go next

The setup order on this page applies to any multi-roll straightener. If you are still deciding which machine the order should be placed for, start with the cluster overview and work outwards from the specifications.

Read the 3-in-1 decoiler straightener feeder guide