Most shops test a difficult material once, produce an acceptable print, and then rely on the memory of whoever ran the test. Six months and one staff change later, the same material is rejected as unprintable and the test is repeated from scratch.

This guide describes how to build a UV substrate qualification library: the record fields that make a past test usable, the status definitions that keep approvals honest, how to capture supplier and lot data, what triggers re-validation, and how to build the first twenty records without stopping production.

Why one-off adhesion tests do not scale

A single test proves a combination worked once. It does not record the conditions, so it cannot be repeated by another operator, and it cannot tell you whether a new delivery of the same material is still within those conditions.

Production exposes the gap quickly. A material that printed well in March arrives in June with a different protective film, and the shop discovers the change through a rejected batch rather than through a record. The cost is not the failed print; it is the hours spent investigating a problem that a recorded baseline would have answered in minutes.

Industrial practice treats qualification as a controlled document set rather than an event. The principle is the same one behind laboratory and inspection records: identify the item, record the conditions, define the acceptance criteria, attach the evidence, and set a rule for when the approval expires. Once the record exists, the question "can we print this?" has an answer that does not depend on who is standing at the machine.

What a UV substrate qualification record must contain

Identity, conditions, results, evidence and validity.

A usable record contains identity, conditions, results, evidence and validity. Miss any of the five and the record cannot be reused, only read.

Identity covers the material name, supplier, thickness, colour, coating and the batch or lot reference. Conditions cover the surface preparation applied, the ink set and primer, the layer order including white ink and varnish, and the curing arrangement. Results cover the tests performed with their measured outcomes. Evidence covers the retained sample, photographs and any certificate supplied. Validity covers the approval date, the approver, the version and the trigger that will require re-validation.

Keep the record on one page where possible. A two-page record that nobody completes is worth less than a one-page record that is filled in every time, and the disciplines that make it useful - identity, settings, result, evidence - fit comfortably on a single form.

Status definitions: qualified, conditional, rejected, expired

Four statuses are enough: qualified, conditional, rejected and expired. The status is the field that turns a collection of test notes into a library, because it tells an operator what to do today.

Qualified means the material passed every criterion with the recorded settings and may be used in production. Conditional means it passed with a named restriction - a specific surface preparation, a minimum primer age, a narrower colour range, a reduced cleaning resistance - and the restriction must be printed on the card, not implied. Rejected means it failed and cannot be used without a process change. Expired means the approval is out of date because a trigger occurred or the validity period passed.

Only the library owner should change a status, and every change should be dated with a reason. This is what stops the familiar drift in which a conditional approval becomes a full approval through repetition rather than through evidence.

Capturing supplier, lot and surface condition

Record the supplier, the manufacturing site and the lot, not just the material name. Two deliveries of "3 mm white PVC" can behave differently, and without the lot reference the difference is invisible until a batch is rejected.

Surface condition deserves its own field. Record whether the material arrived with a protective film, whether it had a coating, how it was packed, and how it was stored before printing. Film-protected acrylic and coated metal behave differently after storage in a cold or damp area, and a record that omits storage cannot explain a seasonal variation.

Where a material carries a supplier certificate or declaration - for example a statement covering restricted substances - reference it in the card rather than filing it elsewhere. Where the destination market imposes substance restrictions, the requirements summarised by the European Commission under REACH are part of what the qualification evidence has to support, alongside the ink and primer documentation.

How should ink, primer, cure and layer settings be recorded?

Record the settings that produced the approved result.

Record ink set and batch, primer, layer order, print mode, pass count, resolution and curing configuration. "Standard settings" is not a record.

The layer order is often the field buyers and operators omit, and it is the field that changes most when a job is re-run. Whether white ink sits under or over the colour, whether varnish is functional or decorative, and how many passes build a raised effect all change both appearance and durability. Where an operator tunes colour between batches, the tunable range belongs on the card with the approved reference.

Component-level data helps where the machine uses standard industrial piezo printheads, because head generation affects drop placement and therefore the setting that passes. The technical documentation published by the component maker, such as Ricoh's industrial inkjet head documentation, is a legitimate reference for that part of the record, and print research bodies such as Fogra maintain published work on how print conditions affect measurable output.

Attachment rules for test strips, photographs and certificates

Attach the physical test strip and a photograph to every record, and file certificates by reference. An unlabelled strip in a drawer becomes unidentifiable within a month, which is why the labelling rule matters more than the storage medium.

Label the retained sample with the record number, the date and the settings version. Photograph the part in consistent lighting against a neutral background, and photograph the adhesion test area rather than only the best-looking region. If you use a tape test, follow the method in ASTM D3359 and record the classification, so the result can be compared with the same test performed a year later on a different lot.

Keep the supplier certificates as linked references rather than duplicates, so a certificate update does not require rewriting every card that uses the material. Where an external laboratory is involved, its report reference belongs in the evidence block, and independent testing providers such as Intertek publish how such reports are structured.

Versioning and change history

Give every record a version and record what changed. A card that is silently edited is worse than a card that is out of date, because it destroys the link between a stored sample and the settings that produced it.

Keep the change history short and specific: version number, date, what changed, who authorised it and whether the retained sample was replaced. Minor edits such as correcting a supplier name should not create a new version; changes to preparation, ink, layer order or curing settings must, because the old sample no longer represents the process.

Print or export the current version for the production area and date it. Operators need the approved setting in front of them, not a folder they have to search, and an exported sheet with a version and date prevents the most common cause of drift: someone following a copy that was superseded six months ago.

What triggers re-validation?

Change triggers re-validation; list them on the card.

Re-validation is triggered by change, and the triggers should be listed on the card so that nobody has to remember them. A record without triggers expires quietly.

Practical triggers fall into four groups. Material changes: supplier, manufacturing site, coating, film, thickness tolerance or storage conditions. Consumable changes: ink, primer or cleaning agent batch, or a change of product. Process changes: print mode, pass count, layer order, curing configuration, fixture or file-preparation profile. Equipment changes: printhead replacement, lamp replacement, alignment work or firmware updates.

Attach a re-validation action to each trigger: a full repeat test, a reduced check on a qualification strip, or a documented review with no test. That way the response is proportionate, and the library owner can see which triggers are causing the most re-work across the shop.

Sample retention and shelf life

Retain one sample per qualification record, and define how long it is kept. Storage conditions matter as much as duration, because a sample kept in sunlight or in a damp corner stops being a usable reference.

A practical rule is to retain the approved sample for as long as the product family runs, or until the record is superseded, whichever is longer, with replacement whenever the settings version changes. Store samples flat, labelled and out of direct light, in the same room as production where possible so that comparing a new print against the standard is a minute's work rather than a search.

Photograph every retained sample when it is created, so that a lost or damaged strip does not invalidate the record entirely. The photograph is weaker evidence than the part, but it is better than nothing, and it is what a new operator will use to orient themselves.

How do you roll it out without stopping production?

Attach it to work you already do, not to a project.

Roll it out as a background task attached to existing work. Every sample test, first article and complaint investigation already produces the information a record needs.

Start with the materials that cause the most problems. For each one, create the card, fill in what is known, run the test that resolves the open question, and mark the status honestly - many first records should be conditional, and that is a useful outcome rather than a failure. Add the card to the release routine for that material so it is used immediately, then move to the next one.

Keep the format stable. Production teams adopt a form they only have to learn once, and a record that changes shape every month will be treated as paperwork. Stability also makes it possible to compare records later, which is the whole point of a library.

A first 20-record build sprint

Twenty records is enough to cover most of a shop's risk and to prove the format works. Build them in a focused sprint, then let the library grow as new materials and jobs arrive.

1. List the materials behind the last twenty rejected batches or customer complaints, and rank them by cost of failure.
2. Create the card for each, filling in identity and preparation from purchase and production records.
3. Run one test per material against written criteria, using a defined sampling approach where you are checking batches - the acceptance-sampling method described in the NIST engineering handbook is a practical reference.
4. Attach the strip, the photograph and the settings that passed, then set the status and the re-validation trigger.
5. Publish the current version to the production area, and add the card to the release routine for that material.

The sprint produces three things beyond the records themselves: a known list of materials that are genuinely conditional, a smaller set of materials that need a process change before they can be used, and a format that has already survived contact with a busy week.