Thick Plate Rolling: Set and Verify Developed Length

For gradual cold rolling of thick plate, the starting length does not automatically equal either the finished inside circumference or the circumference at the section’s centroid. It follows the material’s longitudinal strain through thickness; the effective neutral axis depends on the elastic-plastic stress distribution and the actual rolling process. Measure the finished geometry and thickness rather than infer blank length from a thin-plate rule.

How should you interpret a mismatch in developed length?

The neutral axis is the location through the thickness where longitudinal bending strain is zero for the bending state being considered. Material on one side lengthens; material on the other shortens. The blank length maps to the length at that zero-strain location only when axial strain, material flow, and subsequent trimming are accounted for.

For a finished circular section with inside radius Ri and neutral-axis distance yN measured outward from the inside surface, the corresponding circumference is L = 2π(Ri + yN). For a partial arc with angle θ in radians, use L = θ(Ri + yN). The inside circumference, 2πRi, is the relevant length only if the neutral axis lies at the inside surface, which is not a general rule. The outside circumference is not the starting-plate length by default either.

These geometric relations describe the final arc at a specified cross-section. They do not predict how much length the plate will retain during rolling. Plastic strain, thickness change, axial forces, and end trimming can change the relationship between starting length and finished circumference. A small-scale test that does not match a simple neutral-axis calculation is a reason to check those conditions, not to select the inside diameter as a universal substitute.

Which measurements distinguish geometry from process effects?

Signal or measurement	Source	Wrong-value symptom
Starting blank length	Measure the plate before rolling, along the intended rolling direction.	An inaccurate initial value makes the calculated developed length unreliable.
Finished inside diameter or radius	Measure the rolled section at several locations after it is stable.	Variation around the shell can make one circumference calculation unrepresentative.
Finished outside diameter and wall thickness	Measure the completed section; compare with the inside measurement.	A changed thickness can shift the estimated neutral-axis position and alter both diameter checks.
Pass history and thickness after passes	Record reduction settings and measure thickness before and after rolling.	Unexpected thinning can be mistaken for a neutral-axis or length-calculation error.
End condition and trim amount	Inspect both ends for flat tangent regions and record material removed.	Trimming tangent ends can shorten the finished part without revealing the strain distribution in the retained shell.

Take these measurements before changing roller settings. A length discrepancy alongside reduced wall thickness points to a different process condition than a discrepancy with stable thickness but variable diameter. Likewise, flat tangent ends are a geometry and trim issue: they affect usable finished length, but do not by themselves locate the neutral axis.

Where does the neutral axis move during thick-plate bending?

In simple elastic bending of a straight, symmetric plate, the strain varies through the thickness and the zero-strain plane is at the centroid. Curved-beam theory places the elastic neutral axis inward from the centroid for a curved section. For a rectangular section under the idealized conditions of elastic pure bending, its neutral radius is rn = h / ln(ro/ri), where h = ro − ri, and ri and ro are the inner and outer radii. This is a theoretical elastic relation, not a direct blank-length recipe for a plate that has yielded during rolling.

As bending increases, the inner and outer portions can yield while material nearer the middle remains elastic; with further deformation, the stress distribution depends on the material’s stress-strain curve and loading history. The neutral-axis location must satisfy force equilibrium across the section. In a simple section model, the longitudinal stress integrated over the section balances any axial force; under pure bending with no net axial force, that integral is zero. Once the material response is plastic, the location cannot be chosen solely from the centroid or a curved-beam elastic formula.

Also distinguish neutral axis from neutral surface and from a stress-free layer after unloading. Springback and residual stress can make the post-roll geometry differ from the loaded shape. A no-crack observation on the outside surface does not establish that the plate remained elastic, that the inside did not yield, or that thickness stayed constant.

Why can many rolling passes change the wall thickness?

Multiple small passes may reduce abrupt deformation, but they do not guarantee preservation of thickness. A reported workshop attempt to roll plate described as “40 something mm” on equipment previously used for a 24 mm plate in two passes resulted in about 3 mm of thinning after many passes. That is one shop experience, not a general allowance or a predictable reduction for another machine, material, or setup.

Thickness change matters because it changes the distance between inner and outer surfaces and therefore the geometry used to estimate the neutral-axis circumference. If the plate also stretches longitudinally, starting length will not be recovered simply by measuring the final inside radius. Record actual thickness at repeatable locations, and compare it with the initial measurement. Do not convert one reported reduction into a process correction without a trial on the actual material and machine.

Machine capacity and pass count are separate concerns. A machine’s successful use on thinner plate does not establish how it will behave with thicker stock or repeated loading. Monitor the actual result during the trial rather than treating “many passes” as proof that deformation is controlled.

How do you establish a practical blank length?

Use a measured trial to establish the relationship between blank length and the required finished dimension. If the drawing controls inside diameter, that is the dimensional target to roll and inspect; it is not proof that the blank length equals the inside circumference. Include trim stock for end tangents when the process leaves flat regions that must be removed, and record the actual trim rather than hiding it in a developed-length estimate.

1. Record material identification, initial thickness, blank length, and the required finished ID or other controlling dimension. Keep measurements tied to the same rolling direction and reference points.
2. Roll a representative trial using a recorded pass sequence. Measure thickness and the controlling diameter as the work progresses; note any visible change or end tangent that affects the finished part.
3. After the section reaches the specified condition, measure the finished geometry at multiple locations and record material removed at both ends. Separate retained shell length from trimmed material.
4. Compare the measured blank length with the final arc length and trim amount. Use the trial to revise the next blank or setup, then verify the revised result on another part before treating it as a repeatable process value.

If a calculated neutral-axis position is needed, use the actual section dimensions and a material stress-strain model appropriate to the loading range. For substantially plastic bending, equilibrium must reflect the actual stress distribution; if that model is unavailable, measure a trial rather than assign a guessed neutral-axis shift. A cited curved-beam formula can guide elastic analysis, but it cannot replace process measurement when yielding and thickness change occur.

How do you verify the rolled cylinder before release?

Verify the dimension that governs the part function and drawing: for example, measure ID if the ID is specified as the controlling fit. Check the corresponding OD and wall thickness as well, because the three values expose inconsistencies that a single circumference calculation can conceal. Use repeatable locations and record measurements around the shell; one point cannot characterize a section that is out of round or locally variable.

Reconcile the final measurements with the process record. If thickness changed, use the measured finished section in any geometric calculation. If tangent ends were trimmed, keep trim loss separate from the retained shell’s developed length. A length result that only matches after ignoring removed material is not a reliable production blank rule.

For a welded shell, joint fit-up and weld finishing are separate operations from determining the neutral-axis shift. Rolling to a specified ID and checking fit may be appropriate when the drawing requires that ID, but weld preparation or grinding does not resolve an incorrect developed-length assumption. Apply inspection and nondestructive examination requirements from the governing job documents, not as a substitute for dimensional verification.

Which rolling assumptions most often lead to a bad length estimate?

• Equating ID circumference with blank length: this silently places the neutral axis at the inner surface.
• Using a centroid or elastic curved-beam result after substantial yielding: plastic stress distribution and material behavior can shift the strain-neutral location.
• Treating no visible cracking as proof of no permanent strain: absence of an outside crack says nothing conclusive about thickness or longitudinal strain.
• Ignoring thickness measurements: thinning changes the geometry used to interpret the finished ID and OD.
• Counting tangent ends as retained shell: end trimming changes the accepted part length and must be recorded independently.
• Generalizing a single trial: the reported approximately 3 mm reduction is a warning to measure, not a universal correction.

When measurements disagree, first check measurement locations, finished thickness, and trimming records. Then determine whether the calculation assumes elastic bending while the operation produced plastic deformation. Change one setup variable at a time and retain the pass and measurement record so the correction can be verified.

What do engineers ask about thick-plate rolling length?