The 12-Point Roll Shop Audit Checklist

X-axis servo drift accumulates silently. Without IO log review, a grinder can be 15–20µm out of zero-point for months before anyone investigates. This is the most common hidden source of roll profile error we find on audits.

Servos degrade. Backlash in the X-axis ball screw increases. Without a baseline comparison, you don't know what "normal" was — so you can't detect drift. Most mills we visit have never done this check.

Two operators on the same machine produce different outcomes because they set up differently — different traverse speeds, different dressing frequencies, different infeed parameters. If this isn't logged, you have no data to diagnose shift-to-shift quality variance.

Wheel imbalance is the single most common undetected cause of chatter marks we find on roll surfaces. Almost no roll shop measures it. Almost every roll shop has it. The chatter is always blamed on something else — machine wear, material, coolant — when the real cause is on the spindle.

Manufacturer defaults are calibrated for a generic roll type. Your ICDP backup rolls and your HSS work rolls respond differently to the same signal. Default settings mean your system is looking for the wrong thing in the wrong place — and missing real defects as a result.

If you're discovering roll defects at the mill stand rather than in the roll shop, your inspection system has failed. This is a calibration and threshold problem, not a defect frequency problem. The defects were always there — your system couldn't see them.

Spindle bearing degradation is gradual. Without a recorded baseline, you can't detect the trend. By the time the bearing noise is audible to an operator, you're weeks from a failure. A vibration baseline costs an hour. A spindle rebuild costs a week of downtime.

Rolls expand with heat. A taper measurement taken at the wrong temperature produces a correction that creates the wrong taper in service. This is most critical for hot mill backup rolls and aluminum mill work rolls — where thermal crown interaction is significant.

Coolant is a process parameter, not background infrastructure. Concentration too low means poor lubrication and corrosion. Too high means foam, filter clogging, and skin issues for operators. Temperature variation causes thermal crown instability during the grind. Contaminated coolant is a hidden source of surface finish degradation that no inspection system catches.

Without documented thresholds, regrind decisions are operator judgment calls. Operator judgment varies by shift, by experience, and by production pressure. Inconsistent regrind timing leads to inconsistent roll quality into the mill — and is one of the primary causes of unexplained product quality variation.

When a roll fails in the mill and you need to understand why, the first question is: who ground it, on what machine, on which shift, with what wheel specification? Without that record, you have an incident and no diagnosis. With it, you have a pattern you can eliminate.

Shift-to-shift surface texture variation is the signature of an operator-calibration problem or a machine drift problem. Without a time-series view, each reading looks like an isolated data point. As a trend, it reveals the pattern — and points directly to the root cause.