Process Manufacturing

UV Cure Defect Rate: What Good Looks Like, and Why PPM Is the Wrong Alarm

A UV cure defect rate tells you how often the process failed, not how close it is to failing. The cure window is the leading indicator, and PPM only moves once the window has already closed.

Defect rate in parts per million is the standard way to report UV cure quality and a poor way to manage it. The problem is timing. PPM is a count of failures that already happened, and a UV cure process does not degrade gradually in its output, it degrades gradually in its margin. Lamp output falls, the conveyor creeps faster, a reflector fouls, and for weeks nothing shows up in the defect count because the process still lands inside the cure window. Then it does not, and the PPM number moves all at once.

Measure the window, not just the misses

The variable worth trending is delivered dose, in millijoules per square centimetre, measured with a radiometer puck run through the tunnel on the same fixture the product uses. Compare it against the range the chemistry actually requires, which the supplier's technical data sheet specifies. The distance between delivered dose and minimum required dose is your safety margin, and it is a leading indicator in a way that a defect count never is. A shop that trends dose weekly sees a lamp aging out a month before the defect rate does.

Why your PPM and your customer's disagree

Two definitions circulate and they differ by a large factor. Defects per million opportunities counts each cure-critical feature separately, so a part with six bonded joints offers six opportunities. Defective parts per million counts the part once no matter how many joints failed. Supplier scorecards usually mean the second; internal quality systems often compute the first. Before arguing about whether a number is good, establish which one is being quoted, because the same process can honestly report figures that differ several-fold.

Defect rate tells you the process failed. Dose margin tells you it is about to. Only one of those arrives in time to do something.

The four variables behind almost every failure

Check in this order

• Lamp output and age. Output falls with hours long before a lamp fails outright, and a lamp-hours log is cheaper than the defects it prevents.
• Line speed. Dose is inversely proportional to speed, so a conveyor sped up to hit a schedule quietly cuts cure energy on every part.
• Distance and geometry. Dose falls sharply with distance from the source, and any surface angled away from the lamp receives less than the flat reference the radiometer measured.
• Reflector and quartz cleanliness. Contamination absorbs the short wavelengths first, which are frequently the ones the initiator needs.

Shadowed geometry is a separate failure mode

Under-cure caused by shadowing does not respond to dose increases at the lamp, because the affected surface never sees the light regardless of source intensity. It is a fixturing and part-orientation problem, and it is particularly dangerous because a flat-panel radiometer check will pass while the recessed features on the actual part remain soft. If your defect rate is stable overall but concentrated on specific part numbers, shadowing rather than lamp condition is the likelier cause.

Use the UV cure defect rate calculator to convert counts into PPM and see the gap to your target. Calculate your cure defect rate

Published 2026-08-08.