Why does the minimum matter more than the average?

Because cure is local. Each square centimetre of part either received enough dose or it did not, and no averaging happens between one part of the surface and another. A line whose average comfortably clears the requirement while its cold edge sits 20% below is producing parts that are properly cured in the middle and undercured at the edge, and those parts ship looking identical.

What non-uniformity is acceptable?

The question is slightly wrong, which is why this page answers a different one. Any non-uniformity is acceptable if the cold spot clears the requirement with margin, and no non-uniformity is acceptable if it does not. A very tight lamp whose whole profile sits below requirement fails; a wide profile whose worst point still has 30% margin is fine. Uniformity is a diagnostic for how much dose you must deliver, not a specification in itself.

How is this different from the dose mapping page?

Same measurement, different subject and therefore different fix. This page surveys the LAMP's cure surface. The spread comes from reflector condition, lamp alignment, lamp end effects and working distance, and the remedies are optical. The mapping page surveys a shaped PART, where the spread comes from faces at different angles and distances and the remedies are fixturing, rotation or a second lamp. In the earlier version these were literally the same calculation; separating what they diagnose is the point.

My survey grid was coarse. Does that matter?

Yes, and it biases in the dangerous direction. The true minimum is almost certainly lower than the lowest point you happened to measure, so a coarse survey overstates the cold spot and understates the spread. If the cold spot margin is close to zero on a coarse grid, re-survey around the low region before concluding the line passes.

Should I raise the dose or restrict the width?

It depends where the cold spot is, which this page cannot tell you. If it is at the extreme edges of the illuminated width, restricting the fixture is usually cheaper and costs only usable width. If the cold region runs through the middle of where parts must sit, the profile has to come up. Accepting that the hot spot rises with it, which may run into overcure, yellowing or substrate heating before the cold spot clears.

Does travel along the belt even out the profile?

Along the direction of travel, yes. A part passing through integrates the whole zone, so variation along the belt largely averages out and does not appear as part-to-part variation. Across the width it does not: a part running at one edge stays at that edge for the whole pass. That asymmetry is why cross-web uniformity is the one that matters and why a survey should prioritise it.

UV Curing calculator

UV Dose Uniformity Calculator: Does the Cold Spot Cure?

Measure how evenly a lamp delivers dose across its cure surface, and find out whether the coldest point still cures. Enter the minimum, maximum and average readings from a dose survey, plus the material's requirement. The calculator returns the non-uniformity, the spread, the hot-to-cold ratio, and the margin the cold spot has against the requirement. That last figure is the one that decides anything: a part is cured where it received enough dose and undercured where it did not, and the average is an arithmetic convenience that no part ever experiences.

What this calculator does

Measure dose uniformity across a lamp's cure surface and, more importantly, say whether the coldest point still meets the requirement.
Use it for commissioning a lamp and confirming the whole working width cures, diagnosing why parts at one edge of a belt fail while centre parts pass, deciding how much of an illuminated width is genuinely usable, checking uniformity after a reflector clean or a lamp change, setting the boundaries of a fixture so parts avoid a known cold region.
Measure dose uniformity across a lamp's cure surface and, more importantly, say whether the coldest point still meets the requirement.

Formula used

Dose spread = maximum reading − minimum reading
Dose non-uniformity (%) = spread ÷ average × 100
Cold spot margin (%) = (minimum − required) ÷ required × 100
Hot-to-cold ratio = maximum ÷ minimum
Dose that governs cure = the minimum reading

Inputs explained

Minimum dose reading (cold spot): Lowest dose measured anywhere a part passes. This is the reading that decides whether parts cure.
Maximum dose reading (hot spot): Highest dose measured across the same surface.
Average dose across the cure surface: Mean of the readings. Useful for quantifying spread and experienced by no part.
Required cure dose: The material's requirement, in a named band. Without it a uniformity figure is a measurement with no decision attached.

How to use the result

Best suited to commissioning a lamp and confirming the whole working width cures, diagnosing why parts at one edge of a belt fail while centre parts pass, deciding how much of an illuminated width is genuinely usable, checking uniformity after a reflector clean or a lamp change, setting the boundaries of a fixture so parts avoid a known cold region.
Three summary numbers cannot describe a profile. Two lamps with identical min, max and average can have very different shapes, and shape decides how much of a part sits in the cold region. Says nothing about WHERE the cold spot is, which is what determines whether it can simply be avoided. A coarse survey grid understates the spread, because the true minimum is likely below the lowest point measured. Ignores the direction of travel. A part moving along the belt integrates across the profile, which averages some of this out. Cure is not the only limit; the hot spot may be overcuring or overheating even when the cold spot passes.

Common questions

Why does the minimum matter more than the average? Because cure is local. Each square centimetre of part either received enough dose or it did not, and no averaging happens between one part of the surface and another. A line whose average comfortably clears the requirement while its cold edge sits 20% below is producing parts that are properly cured in the middle and undercured at the edge, and those parts ship looking identical.
What non-uniformity is acceptable? The question is slightly wrong, which is why this page answers a different one. Any non-uniformity is acceptable if the cold spot clears the requirement with margin, and no non-uniformity is acceptable if it does not. A very tight lamp whose whole profile sits below requirement fails; a wide profile whose worst point still has 30% margin is fine. Uniformity is a diagnostic for how much dose you must deliver, not a specification in itself.
How is this different from the dose mapping page? Same measurement, different subject and therefore different fix. This page surveys the LAMP's cure surface. The spread comes from reflector condition, lamp alignment, lamp end effects and working distance, and the remedies are optical. The mapping page surveys a shaped PART, where the spread comes from faces at different angles and distances and the remedies are fixturing, rotation or a second lamp. In the earlier version these were literally the same calculation; separating what they diagnose is the point.
My survey grid was coarse. Does that matter? Yes, and it biases in the dangerous direction. The true minimum is almost certainly lower than the lowest point you happened to measure, so a coarse survey overstates the cold spot and understates the spread. If the cold spot margin is close to zero on a coarse grid, re-survey around the low region before concluding the line passes.
Should I raise the dose or restrict the width? It depends where the cold spot is, which this page cannot tell you. If it is at the extreme edges of the illuminated width, restricting the fixture is usually cheaper and costs only usable width. If the cold region runs through the middle of where parts must sit, the profile has to come up. Accepting that the hot spot rises with it, which may run into overcure, yellowing or substrate heating before the cold spot clears.

Last reviewed 2026-08-25.