Why is the fast half of the window more dangerous?

Two reasons that compound. Dose is inversely proportional to speed, so a given step in ft/min removes more dose near the fast limit than it adds near the slow one, and the failure mode differs: overcuring usually shows as yellowing, embrittlement or adhesion loss that inspection can catch, while undercuring produces parts that look completely finished and fail later at a customer.

Where do the two limits come from?

A cure trial on the actual material and coating weight. Run slow until the part is damaged and record that speed; run fast until cure fails by whatever test you accept on: solvent rub, gel content, adhesion, hardness, and record that. Limits inherited from another line, another material or a new lamp are guesses, and the fast one in particular is not portable.

How do I find my line's speed variation?

Measure it under production conditions rather than trusting a drive specification. Log speed through a full run, including start-up, load changes and any warm-up drift, and take the spread. Belt slip, worn drive components and conveyor loading all widen it, and none of them appear on a nameplate.

The setpoint is inside but the tolerance is not. How bad is that?

It is worse than being cleanly outside, because it produces mixed output. Some parts in the run cross the limit and some do not, and nothing in the record distinguishes them: the traveller says one speed. Either tighten the speed control or move the setpoint towards the middle of the window; the second is usually available immediately and free.

Does the window move as the lamp ages?

The undercure ceiling does, and it moves against you. It is set by the dose available at speed, so as lamp output decays the highest speed that still cures comes down. A window established with a new lamp will overstate the ceiling by the end of that lamp's life, which is why the window is worth re-checking on the same schedule as the radiometer readings.

Can I always trade speed for dose?

For accumulated dose, yes, that is the 1/speed relationship this page uses. But some cures are limited by peak irradiance rather than by dose: oxygen inhibition at the surface is the standard example, where the reaction competes with atmospheric oxygen and only a high instantaneous intensity wins. Slowing down does not fix that failure, and no amount of extra time will.

UV Curing calculator

UV Cure Process Window Calculator: Speed Margin in Dose Terms

Check whether a UV line speed sits safely inside its cure window. Including the speed variation the conveyor really has, and see what each limit means in dose. Enter the current setpoint, the slowest speed allowed before overcure, the fastest allowed before undercure, and the ± variation the line actually holds. The calculator returns the margin at the worst excursion rather than at the dial, plus the dose change waiting at each limit. Those last two rows matter because dose is inversely proportional to speed: a window that looks symmetric in ft/min is never symmetric in the quantity that cures.

What this calculator does

Check a line speed against its overcure and undercure limits including real speed variation, and translate both margins into the dose change they represent.
Use it for confirming a line speed is safe once real conveyor variation is included, deciding which direction to move a setpoint that has become marginal, explaining why speeding up is riskier than slowing down by the same amount, re-checking a cure window as a lamp approaches replacement, setting an alarm band on a conveyor drive from the cure limits.
Check a line speed against its overcure and undercure limits including real speed variation, and translate both margins into the dose change they represent.

Formula used

Margin to nearest limit = MIN(current − slow limit, fast limit − current)
Margin at the worst excursion = margin to nearest limit − speed variation
Inside the window = 1 when the margin to the nearest limit is not negative
Dose change at the slow limit (%) = (current ÷ slow limit − 1) × 100
Dose change at the fast limit (%) = (current ÷ fast limit − 1) × 100

Inputs explained

Current line speed (operating point): The speed the line is set to run at right now.
Slowest allowed speed (overcure floor): Below this the parts take too much dose. Embrittlement, yellowing, adhesion loss or substrate heat, depending on the material.
Fastest allowed speed (undercure ceiling): Above this the parts do not receive enough dose to cure.
Speed variation (±): How far the conveyor actually wanders from setpoint in production: from a chart or a tachometer, including load and warm-up effects. A setpoint is not a process.

How to use the result

Best suited to confirming a line speed is safe once real conveyor variation is included, deciding which direction to move a setpoint that has become marginal, explaining why speeding up is riskier than slowing down by the same amount, re-checking a cure window as a lamp approaches replacement, setting an alarm band on a conveyor drive from the cure limits.
The dose-change rows assume dose scales with 1/speed exactly. That holds for accumulated dose, but not for chemistries that need a peak irradiance. Oxygen-inhibited surface cure is the common case, and it does not care how long the exposure is. Treats the limits as hard edges. Cure quality degrades gradually, and where the edge sits depends on the acceptance test used to find it. Assumes symmetric speed variation. A conveyor that sags under load errs in one direction. Cannot see a window that has moved: as a lamp ages, the undercure ceiling comes down while the page keeps using the number you entered. Says nothing about what the parts are. A mixed run with different coating weights has more than one window.

Common questions

Why is the fast half of the window more dangerous? Two reasons that compound. Dose is inversely proportional to speed, so a given step in ft/min removes more dose near the fast limit than it adds near the slow one, and the failure mode differs: overcuring usually shows as yellowing, embrittlement or adhesion loss that inspection can catch, while undercuring produces parts that look completely finished and fail later at a customer.
Where do the two limits come from? A cure trial on the actual material and coating weight. Run slow until the part is damaged and record that speed; run fast until cure fails by whatever test you accept on: solvent rub, gel content, adhesion, hardness, and record that. Limits inherited from another line, another material or a new lamp are guesses, and the fast one in particular is not portable.
How do I find my line's speed variation? Measure it under production conditions rather than trusting a drive specification. Log speed through a full run, including start-up, load changes and any warm-up drift, and take the spread. Belt slip, worn drive components and conveyor loading all widen it, and none of them appear on a nameplate.
The setpoint is inside but the tolerance is not. How bad is that? It is worse than being cleanly outside, because it produces mixed output. Some parts in the run cross the limit and some do not, and nothing in the record distinguishes them: the traveller says one speed. Either tighten the speed control or move the setpoint towards the middle of the window; the second is usually available immediately and free.
Does the window move as the lamp ages? The undercure ceiling does, and it moves against you. It is set by the dose available at speed, so as lamp output decays the highest speed that still cures comes down. A window established with a new lamp will overstate the ceiling by the end of that lamp's life, which is why the window is worth re-checking on the same schedule as the radiometer readings.

Last reviewed 2026-08-25.