The expensive failures in UV printing are rarely caused by a machine that stopped working. They are caused by a change nobody classified - a new ink batch, a replaced lamp, a supplier that adjusted a coating - leaving a proven recipe quietly invalid.

This guide sets out when a UV print recipe requalification is required: what counts as a recipe change, how minor and major changes are classified, which material, consumable, fixture, maintenance and environment events act as triggers, what test sequence to run, and how first-article approval and deviation records keep the decision traceable.

What counts as a UV print recipe change?

Anything recorded at approval is part of the recipe.

If a change can alter adhesion, colour, cure or registration, it is a recipe change - even when the material name and the job file stay the same.

The scope is wider than most teams assume on the first pass. Substrate supplier, manufacturing site, coating, protective film and storage conditions all count. So does anything in the consumable chain: ink, primer, cleaning agent, varnish, and the batch they came from. Then there is the process itself: layer order, print mode, pass count, resolution, curing configuration, and the file-preparation profile that drives them.

Finally, the machine's own condition belongs to the recipe. A replaced printhead changes drop placement, a replaced lamp changes the cure profile, and alignment work changes the geometric relationship between head and substrate. Where the machine uses standard industrial piezo printheads, the technical documentation published by the component maker - Ricoh's industrial inkjet head documentation is one example - explains why head geometry affects output, and therefore why a head change is a process change rather than a spare-part swap.

Minor and major changes are classified differently

Classify changes by whether they can affect a recorded result, not by how large they feel. Minor changes are recorded and reviewed; major changes have to be tested before production continues.

A minor change is one with no credible effect on adhesion, colour, cure or registration: correcting a transcription error in a recorded setting, updating a supplier contact, or a documentation change. A major change touches a variable that was part of the approved outcome. The classification rule should be written down, and the authority to classify should sit with the person who owns the recipe rather than with whoever notices the change.

Write the rule in a form an operator can apply at the machine: "if this changes, run the strip" or "if this changes, stop and ask". That single sentence prevents most unexplained quality drift, because it removes the judgement call from the busiest person in the room.

Supplier, lot and surface-preparation triggers

Any change of substrate supplier, manufacturing site, coating, film or lot is a trigger. So is a change in the preparation step, because preparation is what makes the surface printable in the first place.

Lot changes are the most frequent trigger and the least visible. A new delivery that looks identical can differ in coating weight, film adhesion or surface energy, and those differences change how the ink wets. Handling that with a defined reduced check - a qualification strip printed with the approved settings on the qualified material - keeps the response proportionate: minutes on a routine delivery, a full investigation only when the strip disagrees.

Preparation changes deserve the same treatment. A switch from one cleaning agent to another, a change in abrasion step, a different drying time or a different primer age are all modifications to the process that produced the approved sample. Record them in the recipe, and treat any later change as a trigger rather than as a workshop preference.

Ink, primer, varnish and consumable triggers

A new batch of ink, primer, varnish or cleaning agent is a trigger, and so is a change of product, even within the same supplier. Consumables are the variable buyers most often assume is constant.

Treat a new batch as a reduced check rather than a full requalification. Print the approved strip on qualified material with the current recipe, compare it with the retained standard, and record the result against the recipe. Where the result matches, release the batch; where it does not, hold the batch and investigate before production is committed to it. The record then shows exactly which批次 were used on which jobs, which is what a customer complaint requires.

Where the ink changes, the layer strategy may need to change with it, which is why ink substitutions are major changes rather than consumable swaps. Confirm the substance documentation too: where the destination market restricts substances found in inks and primers, the requirements summarised by the European Commission under REACH are part of the evidence the recipe has to support.

Fixtures, RIP and profile triggers

Fixture changes, file-preparation changes and profile changes are all triggers. They alter geometry, ink volume or both, and neither shows up in a machine log.

A new or repaired fixture moves the part relative to the head, which changes registration and, on curved or tapered parts, the effective print distance. A change in the RIP profile or in the curve settings changes the ink laid down per pass. A change in the source file's layer structure - a different white ink strategy, a different varnish build - changes what the machine is being asked to do even when nothing on the machine has moved.

Keep the profile and the file template under the same version control as the recipe. Where colour matters, print research bodies such as Fogra publish work on how process conditions affect measurable print results, which is useful when a deviation has to be explained to a customer rather than corrected silently. Where a first article is approved after a change, judge it with the same method each time - for adhesion, a defined tape test such as ASTM D3359 keeps results comparable across operators.

Maintenance, lamp and environment triggers

Maintenance events are triggers: printhead replacement, lamp replacement, alignment work, firmware updates and major cleaning. So are lasting changes in the production environment, particularly temperature and humidity.

Lamp and head changes are the two most common causes of an apparently unexplained quality shift, because both alter a physical condition the recipe depended on. Treat them as process changes: run the approved strip, compare with the retained standard, re-record any adjusted setting, and note the event against the recipe so that a later investigation starts from the right date.

Environment is slower and easier to miss. Ink behaviour and curing respond to temperature and humidity, so a workshop that changes its extraction setup, moves material storage or starts running night shifts in winter has changed part of the process. Record the conditions at qualification so that a seasonal difference can be recognised as variation rather than as a machine fault.

How is the requalification test sequence run?

Compare first, escalate only if the comparison fails.

Run the test in a fixed order: confirm the material against the qualification card, print the approved strip with the current recipe, then compare against the retained standard.

The fixed order matters more than the individual steps. Comparing against a retained standard before changing anything preserves the ability to say what actually differed; adjusting a setting first destroys the comparison and leaves the team with an impression instead of a result. Record what was printed, with which settings, and what the comparison found.

Where a full requalification is needed, work through the variables in the order that affects adhesion and appearance first: preparation, primer, layer order, print mode, then curing. Change one variable at a time. Where a decision has to be defended statistically - typically when a batch has to be accepted or rejected rather than merely characterised - the acceptance-sampling method described in the NIST engineering handbook provides a documented basis for sample size and interpretation.

First-article approval and deviation records

Every major change should end in a first-article approval: a part printed with the final settings, judged against the criteria, and signed. Without that signature the recipe is provisional, and the next batch is a test.

Keep the approval short and specific: what changed, what was tested, the settings used, the result, the approver and the date. Where the outcome is acceptable but not identical to the previous standard - a slightly different colour, a marginally lower adhesion classification - record the difference as a deviation rather than averaging it away, and decide explicitly whether the customer specification still allows it.

Attach the approved part to the record. The physical standard is what allows a new operator, or the same operator in six months, to compare rather than to interpret.

Customer concessions and documentation

Where a change alters a property the customer specified, the decision belongs in the customer file, not only in the workshop. A recorded concession protects both sides and usually takes one email to obtain in advance.

Describe the deviation in the terms the customer used - colour difference measured against the approved reference, adhesion classification, durability result - and name the affected deliveries or order numbers. Where the change is permanent, update the specification sheet you send with quotations so that future orders are quoted against the current process rather than the original one.

Keep the internal and external documents aligned. A recipe that was updated internally but not in the specification sent to the customer produces a legitimate complaint the first time the difference is noticed, and an argument about which document governs.

What records prevent the same failure returning?

Recipe, change log, retained standard and strip results.

Four records do most of the work: the recipe itself, the change log, the retained standard and the strip results. Together they answer what changed, when and with what evidence.

Keep the recipe as a controlled version with the settings that passed. Keep the change log short and chronological, covering classification and approval as well as the technical change. Keep one retained standard per product family, labelled with the version it represents. Keep strip results logged against the material or ink batch they relate to, so a recurring problem is visible as a pattern rather than as a series of incidents.

Review the log monthly for repeat triggers. A supplier that generates repeated strip failures, or a lamp that shortens its interval, is a maintenance or purchasing decision waiting for a document to appear - and once the record exists, that decision takes minutes instead of an investigation.