TL;DR: Most flexographic print defects are traceable to three root causes — anilox cell loading, impression pressure drift, and substrate surface energy drop — all measurable before a job runs long.
TL;DR: In our experience, ink viscosity outside the 18–25 seconds (Zahn #2) window accounts for roughly 60% of density and mottle complaints we receive during press checks.
Dot Gain, Mottle, and Density Shift: Measuring What’s Actually Happening at the Nip #
Flexographic print quality lives or dies at two nip points: the anilox-to-plate nip and the plate-to-substrate nip. When a job comes back from press with washout highlights or plugged shadows, the first thing we do is measure — not adjust. Adjusting before measuring is how you trade one problem for three more.
Dot gain in flexo is expected. The question is whether it falls within the agreed tolerance. Our production standard targets total dot gain of 18–22% at 50% tonal value on LDPE film, referenced against ISO 12647-6 for flexographic process control. If we’re seeing 28–32% gain, that’s a plate or impression problem. If we’re seeing 12–14%, we’re likely under-inking or running a substrate with anomalously high surface tension.
| Failure Symptom | Primary Measurement | Threshold That Triggers Corrective Action |
|---|---|---|
| Highlight dot loss (≤5% tones drop out) | Plate durometer (Shore A) | Below 55°A — replace or re-examine exposure |
| Shadow plugging (85–95% tones fill solid) | Anilox cell volume | Above 5.5 BCM for process work on film |
| Mottle / uneven coverage | Ink viscosity (Zahn #2) | Outside 18–25 sec window |
| Colour density shift mid-run | Impression pressure gauge | Delta >0.05mm from job setup sheet |
| Substrate delamination / ink pick | Dyne level test | Below 38 dynes/cm on PE, 42 dynes/cm on OPP |
The table above reflects the parameter ranges we log under our press setup form PR-04. Values outside these thresholds go to the press operator as a mandatory stop, not a discretionary adjustment. That distinction matters — discretionary calls during long runs tend to drift in the wrong direction.
On the density shift row: a 0.05mm impression change sounds small, but on a 600mm web running coarse flexo (85 lpi), it’s enough to shift L* by 3–4 units on a mid-tone, which is visible. We track impression pressure at job start and at 30-minute intervals on critical colour jobs.
What Actually Causes Press Failures Mid-Run — and Why They’re Often Mis-Diagnosed #
Anilox cell loading is the most common culprit we see, and it’s consistently blamed on something else. A pressman sees density dropping and adds impression. That makes solids look better temporarily, but dot gain in the midtones climbs, and 20 minutes later the highlights are plugging. The actual problem was cell loading — dried ink residue reducing effective cell volume from, say, 4.8 BCM down to 3.1 BCM — not impression. Once impression has been wound in to compensate, you’re now running two compounding problems.
The detection check is simple: pull the anilox, wipe a small section clean with the correct solvent, and remeasure that area’s density transfer against the uncleaned section. A delta of more than 15% transfer efficiency tells you the roll needs an ultrasonic clean before you continue. We schedule ultrasonic anilox cleaning every 8–10 million impressions for standard ceramic rolls, per the maintenance schedule in our equipment log EQ-11.
Substrate surface energy loss is the second most mis-diagnosed failure. OPP and PE films are corona-treated at the converting stage, but surface energy decays — typically from 44–46 dynes/cm at the time of treatment to below 36 dynes/cm after 6–8 weeks of storage at ambient temperature, or faster if films are stored near heat sources or exposed to UV. We’ve received film rolls from the same supplier in the same month where one roll measured 42 dynes/cm and another measured 34 dynes/cm, due to different warehouse bay temperatures. The lower-energy roll ran with visible ink beading on solids within 2 minutes of press start.
The mechanism: water-based flexo inks have a surface tension of roughly 28–34 mN/m depending on formulation. When substrate dyne level drops below that range, the ink no longer wets out across the surface — it beads rather than spreading. No press adjustment fixes this. The substrate needs re-treatment, typically inline corona at 0.5–1.5 kW depending on web speed, or the roll needs to be quarantined. Running impression pressure harder doesn’t improve adhesion; it transfers more ink mass to a surface that still won’t hold it, and you get smear on the next roller contact.
Plate swell and chemical incompatibility is the third failure type we encounter, almost exclusively when a new ink system is introduced without a re-qualification of the plate material. Photopolymer plates — both analogue and digital — are not universally solvent-resistant. Alcohol-heavy cleaning solvents swell the plate surface, increasing effective durometer softness, which increases dot gain, which prints heavy. This typically appears 45–90 minutes into a run when the plate has been in contact with the ink long enough for absorption to occur. The plate looks fine dimensionally; the problem is rheological. If highlight dots measured 10.5% at press start and are reading 14.5% at the one-hour mark without any setup change, plate swell is the first thing we check. Running a solvent drop test against the photopolymer formulation spec sheet (from the plate manufacturer) confirms it in under 5 minutes.
Does Anilox Line Screen Affect Failure Rate More Than Cell Volume? #
Cell volume has more direct influence on failure rate than line screen, but the two can’t be optimised independently.
Running a high line screen (700–900 lpi) with excessive cell volume (above 4.5 BCM) floods the plate with more ink than fine dots can hold without spreading. Conversely, a low cell volume (below 2.8 BCM) at a moderate line screen (360–440 lpi) starves process work of density in shadow areas. The combination we specify for 133 lpi process flexo on film is 600–700 lpi anilox at 3.8–4.2 BCM — this holds for film substrates. For absorbent substrates like kraft paper, we shift the cell volume up to 5.0–5.5 BCM to compensate for substrate absorption.
Line screen influences mottle more than density, particularly on large solids where a lower lpi anilox leaves a visible cell pattern in the ink film. Above 550 lpi, this pattern typically resolves below human visual threshold at normal viewing distance.
Specification Notes for Brand Partners #
When you brief us on a new flexo job, the three details that most directly affect press performance are: substrate type and treatment age, the number of process colours versus spot colours, and whether the job includes large solid coverage areas.
Treatment age matters more than most briefs include. If you’re sourcing your own substrate and shipping it to us for printing, tell us the corona treatment date — or we’ll test incoming rolls ourselves. Our incoming QC protocol covers dyne level on 100% of film rolls received. Rolls below 38 dynes/cm on PE or 42 dynes/cm on OPP are flagged under our material risk procedure and not released to press until they’ve been re-treated or replaced.
The most common brief gap that causes sample iterations is under-specifying ink finish. “Matte” means different things — a client expecting 20–25 GU (60° gloss units) will reject a job running at 45 GU, even if the colour is correct. Specify target gloss range, not just finish type. This eliminates one round of samples in most cases.
Our standard sampling timeline for flexo film packaging is 10–15 working days from confirmed artwork and approved materials. Timeline extends by 3–5 days if the job requires a new plate set, or if substrate needs sourcing from our qualified supplier list.
Frequently Asked Questions #
At what dot gain percentage should we stop the press and investigate?
Any mid-run dot gain increase of more than 4% at 50% tonal value versus the approved proof should trigger a press stop. That’s the threshold we use internally — a 2% drift is within normal run variability; 4% is a signal that something has changed mechanically or in the ink system.
Can we reuse plates across multiple print runs?
It depends on the ink chemistry and how the plates were cleaned and stored between runs. Photopolymer plates cleaned with compatible solvents and stored flat in UV-opaque sleeves typically hold dimensional stability for 500,000–1,000,000 impressions before dot sharpness degrades measurably. Plates exposed to incompatible solvents, or stored rolled under tension, fail much earlier — we’ve seen surface cracking within 3 runs under poor storage conditions. We record plate usage impressions in our plate tracking log and flag plates approaching the 800,000 impression mark for dimensional re-check.
Why does the print look fine on press but shift colour after lamination?
This is an optical effect caused by the change in substrate reflectance. A printed film on press is read against the white backing of the press impression cylinder or against air. After lamination to a white or foil substrate, the light path through the ink changes, and densities can appear 8–12% heavier than the pre-lamination press sheet. For jobs destined for lamination, we use a laminated proof as the colour standard — not a plain film pull — to avoid this mismatch.
What’s the minimum web tension for running 30-micron OPP without registration errors?
Web tension for 30-micron OPP typically needs to hold between 40–65 N/m to maintain consistent registration across the web width. Below 40 N/m, the film relaxes between decks and fan-out becomes measurable — we’ve recorded lateral register errors of 0.4–0.6mm on under-tensioned thin film runs. Above 65 N/m, the film stretches at the print nip and you get elongation error in the machine direction. Both problems look like registration failures but have opposite mechanical causes.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
The dyne threshold point hits close — we switched from standard PE pouches to a 30% PCR-blend film last year and had to requalify our entire surface treatment baseline because recycled content drops corona retention faster, so our “safe” floor moved from 38 to 42 dynes/cm just to maintain ink adhesion through a two-week shelf window.
The “measure before you adjust” discipline sounds obvious until you’re three hours into a press check at 2am and the customer’s QC rep is standing behind you — we’ve lost probably six good production runs over the years to operators chasing dot gain with impression pressure when the anilox was the actual culprit, cells loaded with dried ink from the previous job.
Ran into exactly the dyne level issue on a 3-layer OPP laminate job for a nutraceutical pouch line — substrate tested at 39 dynes on incoming QC, passed, but by the time it sat on the spool rack for 11 days in our Memphis facility the surface had dropped to 34. We didn’t catch it until the customer called with delamination complaints on about 8,000 units already in distribution. The incoming test passing gave everyone false confidence; nobody thought to retest after extended spool storage. Now we have a mandatory retest protocol if film sits more than 72 hours unprinted.
Our film supplier in Qingdao was shipping LDPE that tested fine on the cert — 40 dynes/cm — but by the time it hit our press in week three of a long run, we were consistently dropping below 38 and getting ink pick on every fourth reel. Took us two months to figure out they’d switched their corona treatment line and the surface energy was decaying faster than spec, nothing in the paperwork changed.
Switching anilox rolls to hit that 5.5 BCM ceiling for process film work cost us roughly $2,200 per sleeve at our converter in Charlotte, but the scrap reduction on a typical 50k-foot LDPE run paid that back inside four jobs. The “measure before you adjust” mindset from the article is exactly what justified the capital case internally — we had the press logs to show how much waste traced back to impression drift rather than the anilox spec itself.
We track dot gain variance across substrate lots as part of incoming QC now, and the spread on our 40-micron BOPP runs has been wider than expected — same supplier, consecutive orders, and we’ve seen swings from 19% to 26% gain at 50% tonal value without any press-side changes, which points straight back to caliper inconsistency between lots rather than anything happening at the nip.
Plate durometer is the one we kept skipping during makeready because the plates looked fine visually — we’ve standardized on checking Shore A at three points across the plate width now, not just center, after a job on our Mark Andy 2200 showed 57°A at center but 52°A at the operator side and we chased a highlight dropout issue for two shifts before catching it.
Compostable film is where these thresholds get genuinely painful — we ran a PLA/PBAT laminate for a snack brand last year and the dyne floor on that substrate shifts with humidity in ways PE never did, so our 38 dynes/cm trigger point was essentially meaningless without also logging relative humidity at press side. Got through certification on the packaging claim but the print consistency piece took three additional makeready sessions to stabilize.