Profile buckling roll forming produces a shape that looks almost right. The section is there, the dimensions are close, and the flange has a wave or a lean that no gauge explains.
Operators usually answer this with more roll pressure. That makes it worse, because the buckle is rarely a pressure problem. It is a strain problem, created two or three stands earlier and revealed at the exit.
Seven causes cover most profile buckling roll forming cases. The shape of the buckle tells you which one to check first, and the fix order matters more than the fix itself.
What Buckling Looks Like at the Exit of a Line
Cold formed sections fail in three different ways, and they are often called the same thing on the floor.
Local buckling is a short wavelength wave in a flange or a web, with a half wave length close to the flat width. Distortional buckling rotates the flange at the web junction, so the corner angle opens or closes along the length. Global buckling moves the whole section, producing bow or sweep rather than a local wave.
Distortional buckling is the one that appears during forming and not in service. It starts where the bend meets the web, and it is driven by residual stress from over-forming rather than by applied load.
There is a simple field test. Lay the section on a flat table and look along the length at eye level. A wave that repeats every 200 to 400 mm is local. A lean that runs the whole length is global. A corner that opens only on one side is distortional.
Cause 1 to 3: Pass Design, Gap and Guide Clearance
Cause 1: Too much angle in one pass. Each pass should bend a flange by no more than 20 to 25 degrees in mild steel, and no more than 15 degrees in 550 MPa high strength steel. Exceed that and the outer fibre strain goes past what the flange can carry without a wave. A pass schedule that needs 90 degrees in four passes is asking for 22.5 degrees per pass before springback, which is already at the limit.
Count the passes you actually need with a simple rule: total bend angle divided by 20, plus two finishing passes. A 90 degree flange needs about 6 passes, not 4.
Cause 2: Uneven roll gap across the strip width. A gap that is tight on the operator side and open on the drive side bends the flange asymmetrically. The tight side forms first, the loose side lags, and the difference shows as a lean. Most profile buckling roll forming cases that appear on one side only trace back to this. Check gap with a feeler gauge at both ends of every bottom roll, not just at the centre.
Gap itself should follow thickness plus 5 to 10 percent per side. The full method is in this roll gap clearance guide, and it matters more on thin high strength strip than on heavy gauge.
Cause 3: Guide clearance too loose, or guides doing the forming. Entry guides should centre the strip, not bend it. A side guide with 5 mm of clearance lets the strip walk, and the roll then forms an off-centre flange. A guide with zero clearance forces the strip and creates a different buckle at the same stand. Set side clearance at 0.5 to 1 mm total for strip under 2 mm thick.
Cause 4 to 5: Material and Springback
Cause 4: Yield strength higher than the pass design assumed. A pass schedule designed for 250 MPa steel will over-form 550 MPa steel, because the same gap produces a larger residual stress in the stronger material. Grade substitution without a pass review is the most common cause of profile buckling roll forming on a line that used to run fine. Confirm yield strength from the mill certificate to ASTM A370 testing rather than from the grade name.
Thickness variation inside one coil does the same on a smaller scale. A coil that runs from 1.45 to 1.55 mm produces a section whose flange wave appears and disappears along the length, which is the signature to look for.
Cause 5: Springback not compensated in the roll design. Mild steel springs back 1 to 3 degrees. High strength steel springs back 3 to 6 degrees, and the number rises with yield strength and with the bend radius to thickness ratio.
If the rolls are cut to the finished angle, the profile leaves the last stand under-formed and the operator closes the gap to compensate. That gap change is what starts the buckle. The effective width method in AISI S100 explains the same instability from the design side, and the pass schedule has to respect it. Roll design has to over-bend by the measured springback of the actual material, which means a test bend on the coil lot before the rolls are cut.
For a wider view of how grade choice drives the whole line, see this guide to roll forming material selection.
Cause 6 to 7: Tension, Tooling Condition and Handling
Cause 6: Strip tension fighting the last stands. A coil car or a pre-cut line that pulls back on the strip adds a longitudinal stress the pass design never accounted for. The result is a section that looks correct on the run-out table and buckles as soon as it is cut, because cutting releases the tension that was holding it straight.
The same mechanism appears at the other end when a run-out conveyor runs faster than the line and drags the section. Edge wave from tension control has a separate signature, covered in this strip tension guide.
Cause 7: Worn rolls and dirty roll faces. A worn entry radius no longer matches the pass, so the strip enters the next stand at the wrong angle and gets corrected by force. Buildup on a roll face does the same in a smaller area. Both produce a buckle that appears on one side only and moves when the roll is cleaned.
Diagnose by Shape Before You Touch a Roll
| What you see | Likely cause | First check |
|---|---|---|
| Flange wave with 200 to 400 mm pitch | Too much angle per pass | Pass schedule against 20 degree rule |
| Flange lean, one side only | Uneven gap or guide off centre | Gap at both roll ends, guide clearance |
| Corner angle opens along the length | Distortional buckling from over-forming | Springback test on the coil |
| Buckle appears after cutting | Strip tension released | Tension between line and run-out |
| Wave that comes and goes along the coil | Thickness or yield variation | Micrometer log across the coil |
| Buckle on one side, moves after cleaning | Roll buildup or worn radius | Roll face and entry radius |
| Twist with no flange wave | Roll alignment across stands | Stand square and shaft level |
Sorting the shape first saves hours. Polishing rolls for a tension problem never fixes it, and adding pressure for a springback problem makes the buckle deeper. Practical pass design notes from The Fabricator follow the same sequence: fix the strain, then the tooling.
Fix and Verify
| Fix | What to change | How to verify |
|---|---|---|
| Split the angle | Add a pass or reduce angle per pass | Flange wave gone at line speed |
| Reset the gap | Thickness plus 5 to 10 percent, even both ends | Feeler gauge at both ends, section measured |
| Correct guides | 0.5 to 1 mm side clearance, no forming force | Strip centred, no side marks |
| Over-bend for springback | Test bend, recut or adjust last stands | Finished angle within tolerance off the line |
| Stabilise tension | Free loop or dancer before run-out | Section stays straight after cutting |
| Restore roll condition | Polish, replace worn entry radii | No one-sided buckle after cleaning |
Verify at production speed, not at crawl. Many buckles disappear below 10 m/min because the strain rate is lower, and they return the moment the line runs at rate. This is also why a commissioning check that only runs at low speed can sign off a line that will fail in the first week.
FAQ
1. What causes profile buckling in roll forming? It is almost always excess strain from too much angle in one pass, uneven gap, or uncompensated springback. Material grade and strip tension are the next most common causes.
2. Is buckling the same as edge wave? No. Edge wave runs along the strip edge in the flat, while profile buckling appears in the formed flange or at the web to flange corner. They can share a tension cause but need different fixes.
3. How much angle can one pass take? Up to 20 to 25 degrees in mild steel, and up to 15 degrees in 550 MPa steel. Beyond that, outer fibre strain in the flange exceeds what the flat can carry without a wave.
4. Why does the buckle only appear after cutting? Because strip tension was holding the section straight. Cutting releases the stress and the buckle shows. Fix tension control before changing any roll.
5. Can I fix buckling by tightening the rolls? Rarely. Closing the gap adds compressive strain and usually deepens the buckle. Measure the gap first, then check the pass angles.
6. Does high strength steel need more passes? Yes. Higher yield strength means more springback and less elongation before yield, so the same profile needs more passes at smaller angles.
7. What is distortional buckling in a formed section? It is rotation of the flange at the web junction, so the corner angle changes along the length. In production it usually comes from over-forming, not from service load.
8. How do I check springback on a new coil lot? Bend a strip sample to the tightest angle in the profile, release it, and measure the angle change with a protractor. Do it for every grade change, and record it with the coil number.
9. Does roll polishing help a buckled profile? Only when the buckle is one-sided and moves after cleaning. For a wave across the whole flange, polishing changes nothing because the cause is strain, not friction.
10. Can a profile be designed to resist buckling better? Yes. Wider flanges at the same thickness buckle earlier, and small stiffening lips raise the critical stress a lot. A lipped channel design is the usual answer when a plain channel keeps waving.
Conclusion
Profile buckling roll forming is a strain diagnosis, not a pressure adjustment. Read the shape, find the pass that created it, then change angle, gap, guide or springback compensation in that order.
Changing the last stand rarely fixes a buckle made three stands earlier. The exit only shows what the entry decided.
Beli Rollforming builds roll forming machines and pass schedules for mild steel, galvanized and high strength grades, with springback measured on the buyer’s own coil lot before the rolls are cut. Since 2005 and across 50+ countries, the lines that hold tolerance are the ones where the pass design matched the material from the start.
Last updated: 2026-11-04
Article Changelog
- 2026-11-04: First publication. Seven causes ordered by fix sequence, three diagnosis and fix tables, pass angle limits for mild and 550 MPa steel, springback figures, and a shape-first diagnostic method.
Next Review Triggers
- AISI S100 revision that changes effective width or distortional buckling rules
- Field data showing a new dominant cause on high strength grades
- A customer reports buckling traced to a specific pass number
English