Ink Systems & Formulation — Testing & Validation Protocol

Acceptance Criteria by Test Parameter: How We Define Pass/Fail Before Ink Touches the Press #

Before any ink lot is approved for production use, it goes through what we call our INK-QC-04 incoming validation sequence. This is a structured seven-parameter check, not a spot test. The sequence was developed after a run of colour inconsistency complaints on a folding carton programme in 2022 — fourteen SKUs, 180,000 units, and the root cause traced back to a single batch of process cyan that had shifted in grind fineness without any accompanying test certificate change.

The table below shows our standard acceptance thresholds for lithographic and flexographic ink systems. Gravure has a separate protocol (INK-QC-05) due to the much lower viscosity range involved.

A lot that passes six of seven still fails. There are no partial approvals in INK-QC-04. If a parameter is borderline, the lot goes into quarantine and we request a re-test certificate from the supplier before any disposition decision.

The colorimetric and viscosity thresholds carry the most weight operationally. Colour drift above ΔE 1.5 is perceptible to a trained observer under D50 illuminant and would show up as a brand colour complaint within the first production run. For viscosity, a ±5-second deviation from the approved centre point for water-based flexo systems changes dot gain by roughly 3–5% on mid-tone values — enough to shift a skin tone or a gradient background noticeably on press.

What Causes Batch-to-Batch Drift — and Where Validation Actually Breaks Down #

Pigment dispersion is the variable that causes the most recurring failures in our incoming data. When a supplier changes their grinding media without notifying customers (which happens more often with smaller regional suppliers), the fineness of grind shifts even though the formulation specification on paper stays the same. A lot may arrive with a valid CoA showing 8 µm grind, but our Hegman gauge reads 14–16 µm. The CoA was accurate for the reference batch, not for the current production run. The consequence on press is ink film uniformity loss and mottling in solid areas — and it only shows up after 2,000–3,000 sheets, by which point the issue is already embedded in the production schedule.

Resin ratio drift is a subtler failure mode. Offset inks contain a balance of hard and soft resins that determines tack and setting speed. When a resin component is substituted due to supply chain pressure and the supplier doesn’t trigger a change notification, the tack value can shift by 20–25% while the colour spec remains within tolerance. We catch this via our inkometer check in INK-QC-04, but if a factory skips tack testing and relies only on colour drawdown, this goes undetected until press sheets show picking or linting on coated substrates.

The third failure mode is pH drift in water-based systems, and this one is almost entirely a storage and handling problem rather than a manufacturing defect. Water-based flexo inks are formulated at pH 8.2–9.2, but if stored above 35°C or exposed to CO₂ in poorly sealed containers, pH can drop to 7.4–7.8 within four to six weeks. Below pH 8.0, the resin emulsion destabilises, viscosity climbs unpredictably, and dot sharpness on fine screen work degrades. We’ve seen this specifically with lots shipped by sea freight during summer months in sealed containers without climate control. Our current practice is to test pH on every water-based lot regardless of CoA date — we don’t give pH a “freshness pass.”

Calibration frequency is where validation protocols lose integrity over time. Our pH meters are calibrated against two-point buffer solution (pH 4.01 and pH 7.00 per ISO 3696) every 48 hours of use. The Hegman gauge is verified against a certified block standard at the start of every incoming inspection week. Inkometer calibration follows the manufacturer’s quarterly schedule, cross-checked against a reference fluid with a known viscosity of 1,000 mPa·s. If any calibration check fails, all test results since the last successful calibration are invalidated and the affected lots are re-tested. This has happened three times in the past two years — each time caught at the calibration check, not at a customer complaint.

Should You Require Full Test Certificates for Every Ink Delivery? #

Yes, with a qualifier on scope. A standard CoA from the ink supplier covers colour, viscosity and fineness — which is adequate for stable, long-running jobs where the formulation has been qualified for more than 12 months without incident. For new ink suppliers, new formulations, or any lot arriving for a food-contact compliant job requiring low-migration documentation under EU 10/2011 or FDA 21 CFR §175.300, a supplier CoA is not sufficient by itself. We require independent third-party migration testing before lot release — typically a full compositional declaration and a specific migration test at 10 days / 40°C per the EU framework.

For standard commercial jobs, the supplier CoA covers the baseline. Our INK-QC-04 drawdown and viscosity check adds the incoming verification layer. Both together are the minimum for a reliable batch release decision.

Specification Notes for Brand Partners #

When you brief us on a new packaging programme requiring colour-matched inks, we need at least three things to set the validation threshold correctly: your approved Pantone or spectral reference file (Lab* values under D50, 2° observer per ISO 13655), the substrate you’re printing on (coated, uncoated, laminated film), and whether the packaging has any food-contact or child-product regulatory exposure.

The brief gap that causes the most unnecessary sample iterations is an undefined substrate. A ΔE 1.5 acceptance threshold on a 350gsm SBS board does not translate to the same visual result on a matte-laminated PP film — the ink laydown, dot gain profile and surface energy are all different, and the ink formulation needs to be requalified for each substrate class. If you send us a pantone target without confirming the substrate, we’ll build the colour standard on our default stock, and the first physical sample on your actual substrate may look different even if it technically passes the same ΔE threshold.

Our standard sampling timeline for a new ink qualification is 10–14 working days from receipt of approved substrate and colour reference. This extends to 18–22 working days if third-party migration testing is required. Substrate availability on your side is usually what controls the timeline, not press availability on ours.

Frequently Asked Questions #

What AQL level do you apply to ink lot sampling?
For incoming ink lots, we apply AQL 2.5 per ISO 2859-1 at general inspection level II — this means a sample size of 32 units from a standard delivery lot of 800–1,200 kg, with a maximum of two non-conformances allowed before the lot is rejected.

Can you match a colour from a physical sample rather than a Pantone reference?
Yes, we do this regularly using spectrophotometric reverse-engineering — we measure the physical sample under D50 illuminant, extract Lab* values, and build a target. The accuracy depends on the substrate the sample is printed on. If the sample is on a different substrate than your production packaging, expect 0.5–1.0 ΔE of residual offset that is substrate-related rather than ink-related. For brand-critical colours, we recommend confirming the target on your actual production substrate before locking the spec.

How often do you recalibrate the equipment used in ink testing?
pH meters every 48 hours of active use, Hegman gauge weekly, inkometer quarterly — all against certified reference standards. Results since the last successful calibration are invalidated if a calibration check fails, which has happened three times in the past two years.

Does a passing CoA from the ink supplier mean you’ll skip your incoming tests?
No. The supplier CoA covers their outgoing QC at the point of manufacture. Our INK-QC-04 incoming check covers conditions after transit and storage. Viscosity, pH and grind fineness can all shift during shipping, particularly for water-based systems in summer sea freight. The CoA tells us what the ink was; our incoming check tells us what it is when it arrives.

What happens if an ink lot fails one parameter but passes all others?
The lot goes into quarantine. We do not run partial approvals. If the failing parameter is borderline (within 10% of the threshold), we can request a re-test certificate from the supplier and recheck our own sample before final disposition. If it fails cleanly, the lot is returned or destroyed — it does not go to press.

Is ΔE 1.5 tight enough for brand-critical Pantone colours?
It depends on the colour. ΔE 1.5 is our standard incoming threshold, but for high-chroma brand colours — particularly saturated reds, oranges and specific blues — perceptibility can be higher because the human eye is more sensitive to hue shift in those regions. For brand-critical colours with formal brand standards documentation, we tighten the threshold to ΔE 1.0 and note this in the job specification.

Do you test every single ink lot, or just spot-check?
Every lot, without exception, for any job with a colour-critical or food-contact requirement. For repeat jobs using a supplier we’ve qualified for more than 12 months with a clean incoming record, we run a reduced check (colorimetric and viscosity only) but maintain the full INK-QC-04 frequency for any new formulation or new delivery source. Reduced checking is an earned status, not a default.

---

Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.