TL;DR: Most substrate failures in packaging production are detectable before press — the board specification, not the printing process, is usually where the root cause lives.
TL;DR: In our experience, delamination and crease cracking account for roughly two-thirds of the substrate-related rejections we log under our QC-07 material failure tracking procedure, and both are preventable with caliper and moisture checks at goods-in.
When the Packaging Comes Apart: Failure Modes That Start at Board Selection #
A beverage brand sent us a reorder brief last year for a 350gsm SBS folded carton — same print spec, same structural dieline, same finisher. The first production run had been clean. The reorder failed crease evaluation at 30% of packs in final QC. Nothing on the brief had changed. The problem wasn’t print. It wasn’t scoring depth. It was a 4% moisture content variance in the incoming SBS board — the new mill lot had arrived at 9.2% MC versus the original 8.8% — and that was enough to make the crease fibre fracture rather than compress cleanly under the same die rule pressure.
The brand’s team assumed board was a commodity input. Swap the supplier, hit the same GSM spec, and everything runs. That’s the assumption that causes these failures. Board is not a commodity — it’s a system component. Its caliper, moisture content, fibre orientation, and surface treatment interact with your scoring tooling, your adhesive, and your print process simultaneously. When one variable drifts outside tolerance, the failure usually shows up somewhere else entirely, which is why root cause analysis on packaging failures is harder than it looks.
Crease cracking is the most visible failure, but it’s often not the first thing that goes wrong. What typically precedes it is a board specification that was accepted at goods-in without caliper verification. Our goods-in protocol requires caliper measurement at 10 random points across each delivered lot using a flat-anvil gauge calibrated to ISO 534. If caliper variance within a single lot exceeds ±5%, we flag the lot for hold. A lot that tests at 350gsm nominal but runs 0.38mm–0.47mm in actual caliper within the same pallet is not a uniform substrate — and you cannot run uniform scores and folds from non-uniform board.
The Parameters That Actually Predict Crease, Delamination, and Print Adhesion Failures #
Four measurable board properties determine whether a substrate will perform or fail under production conditions. They are caliper consistency, moisture content, Scott Bond internal bond strength, and surface pH.
Caliper consistency has a direct relationship with crease quality. On our folding carton lines, we specify a maximum within-lot caliper variation of ±0.03mm for premium carton work. Above that threshold, the die rule either over-penetrates on the thicker sheets — causing fibre fracture — or under-penetrates on the thinner ones, producing a hinge that whitens on the outer radius. Neither is acceptable for cosmetics or premium food packaging where the fold is consumer-visible.
Moisture content is the parameter most commonly overlooked at goods-in, and it has the broadest cascade effect. Board arriving above 10% MC is prone to cockling during UV coating and to adhesive bond failure in glued joints. Below 6%, the same board becomes brittle through the crease. The target range for folding boxboard on our lines is 7.5–9.0% MC, verified by a resistance-type moisture meter on edge samples from at least 5 boards per lot. This narrow window matters especially in winter shipments where pallets spend extended time in cold container transit.
Scott Bond internal bond strength, measured per TAPPI T569, predicts delamination risk during folding and gluing. We require a minimum Scott Bond value of 180 J/m² for SBS grades used in folding cartons, and 200 J/m² for boards that will receive lamination or foil blocking. Boards that test below 160 J/m² will delaminate at the fold under compression from the folding carton gluer jaw — not always immediately, but reliably within 30 days in humid warehouse conditions. We’ve seen this failure mode appear as late as post-shipment at a retail distribution centre, which is the worst possible place for it to surface.
Surface pH affects UV and water-based ink adhesion in ways that are invisible before press but catastrophic after. Alkaline fillers used in some recycled content boards can push surface pH above 8.5, which disrupts the cure chemistry of UV flexo inks and causes tacky or soft cure even at correct lamp energy. On our QC-07 incoming inspection checklist, we require surface pH confirmation for any board with recycled fibre content above 30%, using a flat-surface pH probe per TAPPI T529. Target range: 6.8–8.0.
| Failure Mode | Root Parameter | Detection Method | Rejection Threshold |
|---|---|---|---|
| Crease cracking / fibre fracture | Caliper variance + MC | ISO 534 caliper gauge; resistance moisture meter | Caliper spread >±0.03mm; MC >10% or <6% |
| Delamination at fold or glue joint | Scott Bond internal bond | TAPPI T569 z-directional tensile | <180 J/m² (SBS folding carton), <200 J/m² (laminated/foiled) |
| UV ink adhesion failure / soft cure | Surface pH (alkaline filler) | TAPPI T529 flat-surface pH probe | Surface pH >8.0 for UV flexo applications |
| Print mottle / strike-through | Surface porosity and coating weight | Bekk smoothness test; IGT printability | Bekk <150s for coated grades used in halftone litho |
| Cockle / warp during coating | Moisture imbalance, fibre anisotropy | Visual + flatness gauge post-UV coat | Any visible cockle radius >3mm at sheet edge |
The most commonly overlooked parameter is surface pH. Caliper is checked routinely by most operations. Moisture is increasingly common. But surface pH testing at goods-in remains rare outside of food-contact compliance workflows, which is exactly why UV ink cure failures on recycled-content boards keep appearing in production without obvious explanation.
If the Board Has Already Passed Goods-In — Decision Framework for In-Process Failures #
If you’re already on press or at the gluer and a failure appears, the decision logic changes depending on where in the process the failure surfaces.
If crease cracking appears during the fold-glue step, the first check is scoring rule depth, not board specification — but only to eliminate it. If the rule depth is correct (typically 0.1–0.15mm below board caliper for SBS), the issue is almost certainly board-side. At that point, pull five sheets, measure caliper and MC on each, and compare against the goods-in record. If MC has risen more than 1.5 percentage points since goods-in, the board has absorbed ambient humidity during production staging. The corrective action is to condition the board at 20°C / 50% RH for 24 hours before continuing. This is not always practical on a live production schedule, but running humidity-affected board will generate a scrap rate that costs more than the delay.
If delamination appears at the glue joint rather than the fold, and Scott Bond was acceptable at goods-in, look at the gluing process first: specifically, glue open time and jaw pressure. Cold-crack delamination that appears during transit rather than production is a different scenario. That failure is typically triggered by board MC dropping below 5.5% in a low-humidity shipping environment, combined with marginal Scott Bond. For cartons destined for dry-climate markets (parts of the Middle East, inland US in winter), our standard practice is to specify a minimum Scott Bond of 220 J/m² — 10% above our baseline — as a buffer.
If print adhesion failure appears on UV-coated sheets, and surface pH was not tested at goods-in, the options narrow considerably. You cannot retroactively fix alkaline board chemistry. The practical mitigation is to run a test pass with increased UV lamp energy (raise from standard 120 mJ/cm² to 150 mJ/cm²) and verify adhesion per a cross-hatch tape test after 24h cure. If adhesion is still below 4B on the ASTM D3359 scale, the substrate lot should be quarantined and replaced. This is a last-resort path; the cost in press time and material waste is significant.
The non-obvious recommendation: specify Scott Bond and surface pH as mandatory goods-in acceptance criteria in your purchase order terms with your board supplier, not just GSM and caliper. Most board orders are written with only the basic dimensional spec. That gap in the purchase specification is what allows out-of-tolerance lots to pass receiving. Once those parameters are in the PO, the supplier is responsible for mill test certificates on each lot — and incoming verification becomes a confirmation step rather than a full test programme. This doesn’t apply to a one-time short-run order; the overhead isn’t worth it below roughly 5,000 sheets per SKU. Above that volume, it’s the right call.
Specification Notes for Brand Partners #
When you brief us on a new folding carton or rigid box project, the board specification is one area where we always ask questions before we quote or cut a sample. A GSM number alone isn’t enough.
We need to know: the target caliper range (or we’ll spec it from GSM and application), the surface finish requirement (cast-coated, machine-coated, uncoated), any recycled content requirement or FSC certification requirement, and the end-market destination climate (humidity and temperature range affect our moisture conditioning protocol). For rigid boxes, we also need greyboard core density targets, which affect both structural performance and die-cut edge quality.
The brief gap that causes the most unnecessary sample iterations is an unspecified moisture environment at point of use. A folding carton spec’d and sampled in our factory at 22°C / 55% RH may perform differently in a warehouse in Houston at 90% RH or in a shop display in Dubai at 15% RH. If you tell us the end-use environment, we can adjust board selection and coating spec to match — without that information, we’ll build to our own defaults, which may not be optimal for your market.
Our standard sample timeline for folding carton substrates is 10–14 working days from confirmed spec. Rigid box samples with custom greyboard sourcing can run 18–22 working days depending on mill availability. Rush samples are possible but require all specifications to be confirmed in writing before we start.
What caliper variance within a board lot should I reject at goods-in?
On our production lines, we hold ±0.03mm as the maximum acceptable within-lot caliper spread for premium folding carton work. Above that, you get inconsistent crease depth across the run. For less critical brown-box applications, ±0.05mm is workable, but for any consumer-visible fold or panel, hold the tighter number.
My UV coating is coming off in patches on a new board grade — what’s the likely cause?
Surface pH is the first thing to check. If the board has recycled content above 30%, alkaline fillers can push surface pH above 8.5 and interfere with UV cure chemistry. Test with a flat-surface pH probe per TAPPI T529. If pH is above 8.0, try raising lamp energy to 150 mJ/cm² and retest adhesion via ASTM D3359 cross-hatch after 24 hours. If you’re still below 4B, the board lot isn’t compatible with your UV system and needs to be replaced — there’s no process fix for fundamentally incompatible substrate chemistry.
Does FSC certification affect the board’s physical performance?
It depends on which FSC product type. FSC 100% (virgin fibre, certified chain of custody) typically performs identically to non-certified virgin board from the same mill. FSC Recycled grades, however, involve post-consumer fibre and will usually show higher porosity, lower Bekk smoothness, and sometimes elevated surface pH compared to virgin grades. For print-quality-sensitive work, always request a mill test certificate alongside the FSC certificate — the sustainability credential and the print performance spec are separate questions.
How do I know if my board specification is suitable for a humid distribution environment?
This depends on three variables: the board’s moisture content at production, the internal bond strength (Scott Bond), and the laminate or coating barrier. Board MC at 7.5–9.0% with Scott Bond above 200 J/m² and a water-based barrier coating will handle moderate humidity exposure reasonably well. Our dataset on this covers SBS and FBB grades through 18 months of shipment logging to Southeast Asian markets — for boards going to very high humidity environments (above 80% RH consistently), a PE lamination or moisture-barrier coating is worth adding to the specification.
At what point does recycled fibre content start to visibly affect print quality on coated board?
Our press data from coated SBS and CRB (coated recycled board) runs suggests the inflection point is around 30–40% post-consumer content. Below 30%, Bekk smoothness, ink hold-out, and halftone dot gain are typically within 5% of virgin grade benchmarks. Above 40%, you start to see measurable porosity increases that affect gloss and shadow detail in halftone litho — not always visible to the naked eye in flat tints, but detectable in fine gradients and skin tones. For packaging where photograph-quality printing is central to brand identity, we’d recommend staying below 30% PCR content or specifying a higher-coat-weight grade to compensate.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
The 9.2% MC example is close to what broke a whisky gift carton run for us in 2022 — but our failure mode was different. We were running a 400gsm SBS with soft-touch laminate and the crease wasn’t fracturing, it was delaminating at the laminate-to-board interface on inside folds, which our standard crease evaluation completely missed because we were only checking outer fibre. Took three rejected batches before we added a peel test to goods-in specifically for laminated stock.
Had almost the exact same situation with a 350gsm SBS lot from a Shandong mill — reorder came in at 9.6% MC and we didn’t catch it until crease failures started showing up on the gluing line. We’ve since added a resistance moisture check as a hard hold point at goods-in, which our supplier initially pushed back on because it slowed intake, but after we shared the rejection data they built it into their outbound QC instead.
The 4% MC variance point hits close — we had a foil-laminated SBS lot come in at 9.6% last February from our Midwest supplier and didn’t catch it until we were already 4 hours into a treat pouch run, by which point the crease failures were showing up intermittently enough that the line team had blamed the scoring rule twice before anyone pulled the moisture meter out.
On the delamination threshold — is the 180 J/m² Scott Bond floor for unlaminated SBS something you’re holding across the full GSM range, or does that number shift when you’re down at 280–300gsm where the z-directional tensile naturally reads lower anyway?
Watch the fibre orientation marking on the reel or sheet label before it ever hits the cutting table — we’ve had 350gsm SBS pass every MC and caliper check cleanly, then crack on cross-grain scoring because the mill shipped MD where CD was specced and nobody caught it until the creasing matrix was already set up.
Ran into a surface pH issue on a 280gsm SBS job for a sparkling water brand in late 2023 — UV flexo, 4-colour plus a soft-touch OPV. Adhesion looked fine off press but by day two of stacking we were getting ink lift at the score lines, clean sheets pulling apart like a peel test. Sent samples to the lab and the board was sitting at pH 8.4, way above the 8.0 ceiling for UV applications, and nobody had flagged it at goods-in because we weren’t running a pH check as standard on uncoated SBS at that weight. 14,000 sheets. We’ve added TAPPI T529 to our intake checklist since, but that was an expensive lesson in assuming a lighter board meant fewer variables.
Caliper spread is the one we see get waved through most often — had a 350gsm SBS lot last spring where every sheet measured fine at the four corners but the centre-point readings were sitting 0.04mm low consistently, and that’s exactly where the crease failures clustered on the finished carton.
The reorder point is where we’ve been burned most — approval sample from mill lot A, production run from mill lot B, and unless you’ve written mill lot traceability into the purchase order terms explicitly, your converter has no contractual reason to flag the switch. We started requiring lot number declaration on the delivery note and a retained sample from every incoming board pallet in Q3 2022 after a 350gsm SBS reorder from a Portuguese converter came in with a full 5% MC shift that didn’t surface until crease testing on day two of a 400,000-unit whisky carton run.
Snap-lock base geometry is where we found the moisture interaction nobody talks about — we were running a 330gsm SBS watch drawer box with a 6-tab snap base and at anything above 8.5% MC the tab ears wouldn’t seat flush because the board had enough residual spring to push back against the lock geometry. Tightened the die rule pressure trying to compensate, which just shifted the fracture point to the score line on the outer shell. Took us most of a 15,000-unit run in Q3 2022 to isolate it.