Why is total energy not a useful answer?

Because nothing in the process is expressed in it. A radiometer reads mW/cm², a material datasheet specifies mJ/cm², an LED array is quoted in watts of UV optical output, and a line is set by speed. Total millijoules connects to none of those, so an engineer holding that number still has every question to answer. It is kept on the page because the arithmetic behind it is right, and demoted because it decides nothing.

Why does the irradiance row ignore coverage?

Because coverage describes what happens between the source and the part, and irradiance is measured AT the part. The surface needs its target dose regardless of how efficiently the optics deliver it; a radiometer at the working position reads what arrives, not what was emitted. Coverage belongs to the optical power row, which is about what the source has to produce.

Can I always trade exposure time for irradiance?

For a dose-limited cure, yes, that is what dose being irradiance times time means. Two limits bite in practice. Oxygen inhibition needs a high instantaneous intensity and does not care how long a weak exposure lasts. And some chemistries show reciprocity failure at very short exposures, needing more total dose than the linear relationship predicts, so a very fast pass can require more than this page suggests.

What is a realistic optical coverage efficiency?

It depends entirely on the geometry and is worth estimating rather than assuming. A part filling a well-designed reflector's focal region captures most of the output; a small part under a long lamp captures very little, because most of the beam lands on the belt. The useful diagnostic is that the losses are recoverable. Masking, better reflector geometry, closer working distance and smaller lamps all convert directly into lower required power.

Should I use the part's footprint or its surface area?

The developed surface being cured. A shaped part can easily have two or three times the surface of the shadow it casts, and using the footprint understates the energy requirement by that factor. The related question of whether every part of that surface actually receives the dose belongs to the dose-mapping and shadowing pages.

How do I check the optical power figure against a quotation?

Compare it with UV optical output, never electrical input. The two differ by a large factor and mixing them makes a marginal array look generous. Then confirm the wavelength: watts emitted where the photoinitiator does not absorb do not contribute to cure at all, so an array quoted at high output in the wrong band can be worse than a lower-output one in the right one.

UV Curing calculator

UV Cure Area Dose Calculator: Irradiance and Optical Power, Not Millijoules

Turn a cured area and a dose requirement into numbers you can specify equipment with. Enter the surface area cured per pass, the target dose, the share of emitted UV that lands on it, and the exposure time. The calculator returns the irradiance the surface must see and the UV optical power the source must deliver, alongside the earlier version's energy totals. Millijoules of total energy are correct arithmetic and not an operating quantity: no radiometer reports them and no lamp is selected on them. Irradiance and optical power are what the rest of the process speaks in.

What this calculator does

Convert a cured area and a target dose into the irradiance the surface must see and the UV optical power the source must put onto it.
Use it for specifying a lamp or array from a dose requirement and a cycle time, testing whether a proposed line speed is achievable with existing equipment, quantifying what poor optical coverage costs in required source power, converting a materials datasheet requirement into equipment terms, comparing a slow high-coverage design with a fast focused one.
Convert a cured area and a target dose into the irradiance the surface must see and the UV optical power the source must put onto it.

Formula used

Theoretical energy = cured area × target dose
Required delivered UV energy = theoretical energy ÷ coverage efficiency
Energy allowance for coverage loss = required delivered − theoretical
Irradiance the surface must see = target dose ÷ exposure time
UV optical power onto the part = required delivered energy ÷ (exposure × 1000)

Inputs explained

Cured surface area per pass: Surface being cured in one pass. Use the developed area of the actual surface, not the footprint it occupies.
Target UV dose for full cure: Dose the material needs at its surface, in the band the source emits.
Optical coverage efficiency: Share of emitted UV that lands on the cured area. An optics property. Reflector geometry, spill past the part, window absorption, not a chemistry one.
Exposure time per pass: Seconds the surface spends under the source. Turns an energy total into the irradiance a radiometer reads and the optical power an array is quoted in.

How to use the result

Best suited to specifying a lamp or array from a dose requirement and a cycle time, testing whether a proposed line speed is achievable with existing equipment, quantifying what poor optical coverage costs in required source power, converting a materials datasheet requirement into equipment terms, comparing a slow high-coverage design with a fast focused one.
Assumes uniform irradiance across the area. Real sources fall off towards the edges, so the required optical power is a minimum rather than a specification. Says nothing about spectral distribution. A watt of UV at a wavelength the photoinitiator ignores contributes nothing. Treats coverage as a single number, when it varies across a shaped part in the way the dose-mapping page describes. Ignores reciprocity failure: at very short exposures, some chemistries need more total dose than the linear relationship predicts. Cannot see oxygen inhibition, which sets a floor on irradiance independent of any dose requirement.

Common questions

Why is total energy not a useful answer? Because nothing in the process is expressed in it. A radiometer reads mW/cm², a material datasheet specifies mJ/cm², an LED array is quoted in watts of UV optical output, and a line is set by speed. Total millijoules connects to none of those, so an engineer holding that number still has every question to answer. It is kept on the page because the arithmetic behind it is right, and demoted because it decides nothing.
Why does the irradiance row ignore coverage? Because coverage describes what happens between the source and the part, and irradiance is measured AT the part. The surface needs its target dose regardless of how efficiently the optics deliver it; a radiometer at the working position reads what arrives, not what was emitted. Coverage belongs to the optical power row, which is about what the source has to produce.
Can I always trade exposure time for irradiance? For a dose-limited cure, yes, that is what dose being irradiance times time means. Two limits bite in practice. Oxygen inhibition needs a high instantaneous intensity and does not care how long a weak exposure lasts. And some chemistries show reciprocity failure at very short exposures, needing more total dose than the linear relationship predicts, so a very fast pass can require more than this page suggests.
What is a realistic optical coverage efficiency? It depends entirely on the geometry and is worth estimating rather than assuming. A part filling a well-designed reflector's focal region captures most of the output; a small part under a long lamp captures very little, because most of the beam lands on the belt. The useful diagnostic is that the losses are recoverable. Masking, better reflector geometry, closer working distance and smaller lamps all convert directly into lower required power.
Should I use the part's footprint or its surface area? The developed surface being cured. A shaped part can easily have two or three times the surface of the shadow it casts, and using the footprint understates the energy requirement by that factor. The related question of whether every part of that surface actually receives the dose belongs to the dose-mapping and shadowing pages.

Last reviewed 2026-08-25.