Key numbers
UV undercure diagnostics — the physics in five lines:
- UV inks are ~100% solids: monomers + oligomers + photoinitiators. Radiation in the 100–380 nm band splits photoinitiators into radicals that polymerize the film. No solvent evaporates; the film either cross-links or stays liquid.
- Mercury lamps convert only ~25% of power into UV (≈50% becomes IR heat, ≈25% visible light) at 100–120 W/cm — heat management is part of cure management.
- Oxygen inhibition: atmospheric O₂ scavenges radicals at the surface, leaving a sticky, undercured skin; inert-gas (N₂) purging is the industrial counter.
- Cure = dose, not just intensity: accumulated energy (mJ/cm²) and peak irradiance (mW/cm²) are different quantities — a fast carriage pass can deliver high irradiance and still starve the film of dose.
- Thick and white layers absorb/reflect UV: the surface skins over while the bottom stays liquid — adhesion fails at the substrate, not in the ink.
A pallet of printed polycarbonate comes back: the ink flakes off at the corners and the client mentions a chemical smell. Nine times out of ten this is not "bad ink" and not "bad substrate" — it is an undercured film, and undercure always has a findable physical cause.
What curing actually is
A UV ink contains no volatile solvent. It is a liquid mixture of monomers (viscosity regulators), oligomers (the future polymer backbone), pigments, and photoinitiators. Under UV radiation the photoinitiators decompose into radicals; the radicals launch a chain reaction that cross-links monomers and oligomers into a three-dimensional polymer network — within milliseconds.
Two consequences follow. First, there is no "drying later in the stack": whatever did not polymerize under the lamp stays liquid inside the film forever. Second, the reaction needs the right wavelengths — photoinitiator absorption bands must overlap the lamp's emission spectrum, not just receive "some UV."
Lamp-side causes
Mercury gas-discharge lamps emit a broad polychromatic spectrum at 100–120 W/cm of print width — but only about a quarter of the consumed power becomes UV. Half is infrared heat, a quarter is visible light. That is why lamp housings and reflectors need serious cooling, and why thin substrates deform under a lamp that is "working fine."
The practical failure modes an audit checks:
- Aged bulbs: UV output degrades long before the lamp visibly fails. Hour counters must be logged and bulbs replaced on schedule, not on burnout.
- Dirty or degraded reflectors: a large share of the dose reaching the film comes off the reflector; ink mist and dust quietly halve it.
- Spectrum mismatch after ink change: LED units emit narrow bands (typically 365–405 nm region), while many ink systems were formulated for broadband mercury spectra. Switching ink or lamp without re-matching photoinitiator chemistry is a classic silent undercure. (Exact band roles — UVC for surface cure, UVA/UVV for depth — are general photochemistry; verify against your ink's datasheet.)
- Speed vs dose: doubling carriage speed halves exposure time. Peak irradiance (mW/cm²) may look unchanged on a radiometer snapshot while accumulated dose (mJ/cm²) drops below the cure threshold.
Chemistry-side causes
Oxygen inhibition is the most underdiagnosed one: atmospheric oxygen captures radicals at the film surface and terminates polymerization early. The result is a tacky, smearable surface over a cured core. Industrial systems counter it with nitrogen purging of the curing zone (standard in excimer and electron-beam setups); in everyday UV printing, the practical levers are dose, photoinitiator package, and lamp proximity.
Persistent odor is a measurement, not a nuisance. Photoinitiators smell; the odor fades sharply once cross-linking completes. A finished print that still smells days later is telling you it contains unreacted monomers — treat it as a failed cure test, because chemically it is one.
The layer-thickness trap (and white ink)
UV must reach the bottom of the film for the film to anchor. Thick layers — and especially opaque white (its whole job is to reflect and absorb light, see ISO 23498 opacity evaluation) — attenuate UV before it reaches the substrate interface. The visible symptom is a paradox: the surface passes a fingernail test while adhesion fails catastrophically in a tape pull. Operators who respond to poor opacity by "adding one more white pass" are often buying opacity with adhesion.
A 15-minute triage before calling anyone
- Smell test on a fresh vs day-old print — persistent odor = unreacted monomer.
- Cross-hatch + tape pull on the actual substrate (the standard adhesion practice per ASTM D3359-type methods) — failure at the substrate interface points to depth undercure; surface tack points to oxygen inhibition or dose.
- Log lamp hours and compare against the bulb's rated UV-life, not its rated burnout life.
- Halve the carriage speed on a test strip: if adhesion recovers, the problem is dose, not chemistry.
- Check the white: print the same art without the white underbase; if adhesion recovers, the white layer is eating your UV.
Each of these isolates one variable in the chain lamp → spectrum → dose → chemistry → layer stack. An independent audit runs exactly this chain with measurements — radiometer dose profiles, controlled speed ladders, layer-by-layer adhesion — and pins the cause to a component you can fix, instead of a supplier you can blame.