How to Tell When Servo Feeder Rolls Need Replacing?
The first sign of a worn roll set is rarely a mark on the strip. It is a feed length that needs a small correction every second shift, and an operator who stops mentioning it.
On an NC servo feeder, the servo is usually the last thing at fault and the first thing blamed. The rolls are the opposite.
This covers the three symptoms worth acting on, a twenty-minute confirmation method, and how to decide between re-grinding, re-coating and replacement.
Three Symptoms That Point at Worn Rolls
Three field symptoms account for most genuine roll wear, and each one has a different mechanical cause behind it.
Feed length that drifts one way. The line runs short, the operator adds 0.05 mm, and a shift later it needs another 0.05 mm. A drifting correction points at grip, not at the encoder.
A mark at a repeating pitch. Measure the distance between marks. If it matches the roll circumference, the defect is in the head and not in the die.
Slip that only appears on thick or coated strip. A roll that grips 0.8 mm SPCC but slips on 2.0 mm HSLA is losing pressure, usually because the roll face has glazed or flattened.
A fourth symptom gets misread as wear. Noise from the head at speed is more often a bearing or an alignment issue than a roll surface issue, and it is worth separating before any roll is removed.
Confirming Roll Wear in Twenty Minutes
Confirmation needs a dial indicator, a micrometer and a clean section of strip. None of it requires the head to come out.
- Measure radial runout on each roll with the head open and the rolls rotating slowly by hand
- Micrometer the roll diameter at three points across the face to find taper, which shows as a width-dependent feed error
- Inspect the roll face under a light for glazing, pick-up or a polished band where the strip edge runs
- Run twenty parts at the production feed length and log the deviation on the first and the twentieth
- Repeat the same run with the roll pressure one step higher and see whether the deviation closes
- Check the strip for a mark that moves when you change the roll pressure, which separates a roll defect from a die defect
The fifth step is the decisive one. If extra pressure closes the error, the rolls are losing grip and the surface is the suspect.
If extra pressure changes nothing, the problem is upstream or downstream of the rolls, and replacing them will not fix it.
Six Causes Behind a Drifting Feed, Ranked
Roll wear is one of six causes behind the same symptom, and it is not the most common one.
| Cause | Evidence That Confirms It | Where It Costs You |
|---|---|---|
| Roll surface glazing or pick-up | Extra pressure closes the error and the face looks polished | Re-grinding removes the hardened layer if the roll is case-hardened |
| Radial runout above 0.05 mm | Deviation repeats once per revolution, not once per feed | Roll replacement, and a re-check of the whole set rather than one roll |
| Pneumatic or spring pressure loss | Error appears at the start of a shift and improves when warm | Cheap to fix, but it hides behind a roll replacement if nobody measures pressure |
| Encoder coupling or belt backlash | Error appears at direction changes and at high stroke rates | Costs a full servo investigation, and replacing rolls first wastes both |
| Strip contamination | Error tracks a specific coil or supplier, not a shift | Roll cleaning, plus a lubricant review that no maintenance plan includes |
| Material change | Error starts on the day a new grade or thickness was introduced | A setup problem dressed as wear, and it usually ends in an unnecessary roll order |
The bottom row is the most expensive mistake on the list. A new grade of strip changes grip, and ordering rolls before checking the setup spends money on the wrong part.
Work the table from the bottom up when a feed problem starts on a specific day. Work it from the top down when the drift has been growing for months.
A Runout Reading Does Not Prove Wear
A runout reading is a number, and a number taken in the wrong condition is worse than no measurement at all.
Runout measured unloaded is not the loaded condition. A roll that reads 0.03 mm on the bench can read double that under feed pressure if the shaft or the housing is the compliant part.
A single bad part does not condemn a roll set. One short feed in five hundred is a control or sensor event far more often than a surface event.
Flatness of the roll face is invisible to a diameter check. A roll can measure perfectly round and still have a crown that feeds the strip unevenly across its width.
Wear is not always the reason for replacement. If the roll set is right but the line has moved to thinner strip, the answer may be a different roll material rather than a new set of the same one.
There is one case where measurement genuinely misleads. On a line running coated strip, the coating transfers to the roll and the roll measures oversize.
Clean the face before measuring, or the reading will send you looking for a fault that does not exist.
When the Strip Is the Problem, Not the Roll
Two material conditions produce symptoms that look exactly like roll wear and survive a roll replacement.
Coil-to-coil thickness variation. If the supplier runs at the top of the tolerance band on one coil and the bottom on the next, the roll gap that held yesterday will slip today. Measure the strip, not the setting.
Oil film that varies with the coil. A wet coil grips less than a dry one, and the difference is often enough to move the feed length by more than the tolerance.
Both cases share a signature. The error changes with the coil rather than with the shift, and it does not respond to roll pressure.
That signature is why the first step in any feed investigation should be a strip measurement, not a maintenance work order.
These three come up in almost every roll wear conversation we have with a press shop.
At what runout does a 60 mm feed roll need replacing?
Treat 0.05 mm of radial runout as the practical limit on a 60 mm roll. Beyond that, the resulting feed error starts to show up on tight-tolerance work.
Should a roll set be re-ground at 250,000 or 500,000 strokes?
Inspect at 250,000 strokes and decide on the measurement. Case-hardened rolls often cannot be re-ground safely at all once the hardened layer is thin.
Does a 0.05 mm runout really change the feed length?
Yes, by roughly the same order as the runout. On a 200 SPM press that is enough to move a part out of a ±0.05 mm window on a progressive die.
Re-Grinding, Re-Coating or Replacing
Once wear is confirmed, three routes are open and only one of them is right for a given roll.
Re-grinding suits a through-hardened roll with surface glazing. It restores the profile and removes pick-up, at the cost of a smaller diameter that has to be re-datumed in the control.
Re-coating suits a roll whose body is sound but whose surface has lost its grip. Chrome and carbide coatings are both common, and both need a clean substrate to bond.
Replacement is the answer when the roll is case-hardened and the hardened layer is already thin, or when the diameter is below the manufacturer's limit.
Ask the supplier for the minimum permissible diameter before choosing. Re-grinding a roll past that point produces a roll that fits and no longer grips.
For a line holding ±0.05 mm at 200 SPM, the roll set is a precision component rather than a wear plate. Treating it as a wear plate is how a line ends up with four mismatched rolls and no baseline.
Logging That Makes the Next Roll Set Last
Roll life is only predictable if the numbers that ended the last set are written down. Five entries do it.
Strokes and tonnes at replacement. Two counters, not one, because a roll set running thin strip wears differently from one running heavy gauge.
Material grades and thicknesses run. A set that spent a year on 1.0 mm SPCC cannot be compared with one that ran 2.5 mm HSLA.
Roll pressure settings used. Over-pressure is the fastest route to glazing, and it is invisible in the maintenance record unless it was recorded.
Runout readings over time. Three readings across the life of the set tell you the wear rate. One reading at the end tells you only that it failed.
Coating and cleaning routine. Whether the faces were wiped weekly or annually explains most of the difference between two identical roll sets.
Log the strip measurement alongside the feed deviation. Once both sit in the same row, the difference between a roll problem and a material problem becomes obvious on inspection rather than after a replacement.
Where to go next
This article sits under the feed accuracy cluster, one layer below the specification overview. If the drift persists after the rolls are cleared, the error budget from coil to die is where to look next.
The Coil Line Guide Library
