Precision slitting does not create flatness — it reveals and preserves it. The flatness a fabricator receives in a slit aluminum coil is the net result of three sequential controls: the residual stress profile left by rolling and annealing, the corrective work done by a roller or tension leveler before slitting, and the slitting line’s ability to avoid introducing new shape defects during recoiling. The commercial benchmark for precision slit coil is flatness held below 0.03 mm/m, with camber controlled to ≤1 mm/m and i-unit flatness reaching ≤5 I-units for automotive-grade stock and ≤3 I-units for high-precision applications. Achieving those numbers on a slit coil requires the leveling step to plastically stretch the strip beyond yield across its full width, followed by slitting practices that do not reintroduce edge wave or coil set.
Flatness Tolerance Benchmarks for Precision Slit Coil
| Metric | Precision Slit Benchmark | Commodity / General Benchmark | Governing Reality |
|---|---|---|---|
| Flatness (straightedge) | <0.03 mm/m | ≤2 mm out-of-flat (general tension-leveled strip) | The precision number is a per-metre straightedge deviation; the general number is a broader acceptance band |
| Flatness (I-unit) | ≤5 I-units (automotive-grade); ≤3 I-units (high-precision) | Up to 18 I-units before feeding problems appear | I-unit measures length difference across the strip width; lower is flatter |
| Camber | ≤1 mm/m | Up to 3.2 mm per 10 m in poorly controlled stock | Camber is lateral bow, distinct from local wave or coil set |
| Width tolerance | ±0.05 mm (precision) | ±0.125 mm to ±0.25 mm (general slitting) | Width consistency affects downstream tracking and gap errors |
| Edge burr | ≤0.05 mm (precision); ≤5–10% of thickness | 0.05 mm+ creates downstream problems | Burr is not a flatness metric, but it scores adjacent wraps during recoiling |
Why Roller Leveling Is the Flatness Correction Step
Rolling and annealing leave a strip with non-uniform residual stress across its width. Some longitudinal fibers are longer than others; when the strip is uncoiled and tensioned, the longer fibers appear as edge wave, centre buckle, or a combination of both. A roller leveler corrects this by bending the strip alternately over a series of work rolls, subjecting the outer fibers to tensile and compressive stress beyond yield. The result is selective elongation: the longer fibers are stretched plastically so that every longitudinal section has approximately the same length per unit width.
The critical process condition is that the strip must exceed its yield point during the bending cycle. A leveler that only elastically bends the strip does not change the residual stress pattern and does not correct flatness. The depth of penetration of the yield zone through the thickness determines the correction: deeper penetration corrects more severe shape defects, but it also requires smaller work-roll diameters and tighter roll spacing, which limits the gauge range a given leveler can handle.
For thin-gauge aluminum, a roller leveler alone can be limited because the light gauge is less rigid and does not reliably assume the bend radius defined by the work rolls. This is where tension leveling becomes the preferred route for precision slit stock: the strip is pulled under tension while passing over small rolls, and the combination of tensile stress and bending produces a more uniform plastic elongation across the width. Tension leveling is specifically used to minimise residual stresses from rolling and to achieve the flatness required for can-end and precision applications, with reported requirements as tight as 0.2% steepness and 0.5 mm/m wave height.
The Slitting Line Does Not Fix Flatness — It Can Only Avoid Making It Worse
A leveler upstream of the slitter can deliver a flat master coil. Whether that flatness survives to the slit coil depends on the slitting and recoiling operation, and this is where many flatness complaints originate.
Slitting introduces residual stress through the shear zone. The rotary knives shear the strip with a clearance typically 8–12% of material thickness for aluminum. If the clearance is too tight, the knives wear and generate burr; if it is too loose, the material tears rather than shears, producing a ragged edge that can cause coil collapse during recoiling. Either condition alters the local stress state at the slit edge, and that altered stress can manifest as edge wave after recoiling.
Recoiling tension profiles can reintroduce coil set or edge deformation. Even a perfectly flat strip can acquire a wound-in shape defect if the recoiler tension is not matched to the strip’s thickness and width, or if the coil is wound with uneven tension across the width. The precision slit specification for flatness <0.03 mm/m is achievable only when the slitting line’s tension control, edge quality, and recoiling geometry are maintained as tightly as the leveler’s settings.
This is why “tension leveled — slit last” is a recognised process sequence for shape-critical coils: the leveling step is performed immediately before slitting, so that the flatness correction is not degraded by an intermediate handling or recoiling operation. A coil that is leveled, stored, and then slit weeks later is more vulnerable to stress relaxation and handling damage than one that is leveled and slit in a continuous or near-continuous sequence.
I-Units and Steepness: How Flatness Is Actually Quantified
“Flatness” is not a single dimension, and the two metrics used in precision coil specifications measure different things.
I-units measure the length difference across the strip width, expressed as 10 micro-strain: I-unit = 10⁵ × (ΔL/L). A strip with 5 I-units has a 0.005% length variation between its longest and shortest longitudinal fibers. This is the metric that correlates with downstream feeding stability: automotive-grade coil is specified at ≤5 I-units, and high-precision applications reach ≤3 I-units. A coil at 18 I-units is at the threshold where feeding problems, machine stops, and scrap rates become unacceptable in automated stamping or roll forming.
Steepness and wave height measure the geometric amplitude of the shape defect when the strip is uncoiled and allowed to relax. 0.2% steepness and 0.5 mm/m wave height are reported requirements for tension-leveled aluminum thin sheet in demanding applications. The two metrics are related but not interchangeable: a strip can have low I-unit length variation and still show visible wave if the defect is localized, and a strip with a higher I-unit value may appear acceptable under tension but show edge wave when the tension is released.
The specification must state which metric governs, and at what measurement condition — as-wound coil, controlled uncoiled strip under tension, or levelled strip at rest. The shape characteristic must also be named precisely: camber (lateral bow) is a different defect from edge wave or centre buckle, and a purchase order that collapses all three into “flatness” makes supplier and processor measurements incomparable.


