TL;DR: Tolerance stackup in spiral-wound composite cans accumulates across three independent variables — body wall, flange curl, and end cap seaming — and must be resolved at the design stage, not during first-article inspection.
TL;DR: A ±0.4mm radial tolerance on the body OD combined with a ±0.3mm variance on metal end curl creates a worst-case gap of 0.7mm at the seam interface, which is enough to cause liner delamination under thermal cycling.
Where Dimensional Tolerance Errors Actually Enter the Assembly Stack #
When a design engineer briefs us on a new composite can format, the first drawing we ask to see is not the label artwork. It’s the cross-section at the end cap seam. That single detail tells us more about downstream risk than any other spec on the brief.
Composite can assemblies involve three independently manufactured components — the spiral-wound or convolute-wound paper body, the metal or foil-laminate end cap, and the barrier liner — each produced on different equipment with different dimensional process windows. When those components meet during assembly, their tolerances don’t average out. They stack.
The paper body contributes ±0.3–0.5mm on outer diameter (OD), depending on mandrel wear state and ambient humidity. The metal end cap contributes ±0.2–0.3mm on the curl radius. The liner, if heat-bonded separately, adds ±0.1–0.2mm on its bonded edge height. In a worst-case additive scenario — and we do see this in production runs with multi-source components — total interface gap can reach 0.8–1.0mm. At that point, the double-seam or friction-fit cannot maintain adequate radial compression, and the hermetic barrier at the base seal fails under 40°C heat cycling.
Our standard incoming inspection protocol (logged under IMC-DIM-09 in our composite can dimensional register) flags any paper body where OD variation within a single production lot exceeds 0.4mm. Above that threshold, we quarantine the lot before liner bonding, because post-bond correction is not economically viable.
The diagnostic table below maps observable symptoms to their likely tolerance origin — useful for determining where to probe first when a prototype or production sample is showing seam or closure issues.
| Symptom | Probable Tolerance Source | Diagnostic Measurement |
|---|---|---|
| End cap gaps or lifts after seaming | Paper body OD undersized by >0.3mm | Measure body OD at 3 positions along 100mm length |
| Liner edge visible / proud of flange | Liner bond height out of tolerance (+0.2mm) | Cross-section micrometer at 6 o’clock position |
| Cap pops under >40°C thermal load | End cap curl radius too tight (underformed) | CMM check of curl inner radius vs. spec |
| Body splitting at spiral seam under compression | Wall caliper below 3.0mm for >100mm OD tube | Ultrasonic wall thickness gauge, 8-point radial scan |
| Inconsistent torque on threaded end caps | OD eccentricity >0.2mm TIR | Runout dial gauge on mandrel fixture |
For formats below 60mm OD, these tolerances tighten because the curl geometry scales nonlinearly. We ask designers to reduce the allowable body OD tolerance band to ±0.25mm for sub-60mm diameter formats.
The Root Cause Most Design Reviews Miss: Hygroscopic Growth After Labeling #
Paper body dimensional change during and after labeling is the most consistently misdiagnosed failure source in composite can programs. When a can comes back from first-article review with end cap seam issues, the instinct is to re-check the body winding spec or the seaming tooling. In our experience, the root cause is moisture uptake in the paper layers between liner bonding and final assembly — a window that can span 6–72 hours in factory conditions.
The mechanism: spiral-wound paper bodies are wound at equilibrium moisture content (EMC) of roughly 6–8% for standard kraft plies at 23°C/50% RH (aligned with ISO 187 conditioning). When freshly wound bodies are moved to a label application station that runs at 65–75% RH — common in high-throughput printing environments during summer months in south China — the outer plies absorb moisture and the body OD grows. For a 90mm OD can with a 4-ply wall at 2.5mm total caliper, we have measured OD growth of 0.15–0.22mm over a 4-hour exposure window at 70% RH.
That growth doesn’t reverse cleanly when the can moves to a drier downstream station. Kraft paper exhibits hysteresis: the return-to-equilibrium OD is typically 0.05–0.10mm larger than the pre-exposure OD. Per TAPPI T411, this is expected behavior for kraft at standard grades. The problem is that end cap tooling is set to the pre-exposure OD. So what arrives at seaming is a body running systematically oversize, and the seam tool applies insufficient curl pressure to achieve the minimum 1.5mm engagement depth we specify for hermetic seals.
Confirmation method: measure body OD immediately after winding (baseline), after liner bonding (intermediate), and again immediately before seaming (final). If the final OD is more than 0.15mm above baseline, investigate the humidity gradient along the production route. We use a portable hygrometer logged against each production batch — this data feeds directly into our first-article package and is available to our brand partners on request.
This matters more than most people account for during design simulation. FEA models for composite can compression and seam integrity typically use nominal OD and fixed elastic moduli from dry-state paper data. If the simulation inputs don’t include a moisture-adjusted OD and a humidity-corrected compressive modulus — which for kraft drops roughly 12–18% from dry to 70% RH per TAPPI T822 — the simulation will underpredict seam failure rates in humid shipping environments.
Corrective Actions Ranked by Impact and Feasibility #
-
Establish a controlled humidity zone between winding and seaming (highest impact, medium cost). Maintaining RH at 50% ±5% along the full production flow eliminates hygroscopic OD growth as a variable. For most factories, this means partitioning the winding-to-seaming area with HVAC or dehumidification. Capital cost is real, but it removes the root cause rather than compensating for it.
-
Tighten incoming body OD acceptance band to ±0.25mm for all hermetic-spec cans. This is our standard practice for food-contact and pharmaceutical composite cans. It increases rejection rate on incoming bodies by roughly 3–5% but catches dimensional drift before it compounds at assembly. The cost delta is absorbed in incoming QC, not in rework.
-
Adjust seam tooling to a humidity-corrected nominal. If humidity control isn’t feasible, tooling can be offset to the expected in-process OD (baseline + 0.15mm growth) rather than the dry nominal. This fixes the seam engagement for the majority of production. It does not fix edge cases where humidity spikes above 75% RH.
-
Specify WVTR-rated outer label stock to reduce moisture transmission to paper body. A label substrate with WVTR ≤ 15 g/m²/24h at 38°C/90% RH (per ASTM E96) acts as a partial moisture barrier during the winding-to-assembly window. This is a lower-cost intervention than HVAC control, but it only slows ingress — it does not eliminate it.
-
Redesign end cap curl geometry for ±0.5mm OD tolerance accommodation (high cost, thorough fix). A compound-radius curl profile, rather than a single-arc design, provides 40–60% more radial compliance without sacrificing seam strength. Tooling cost for this modification typically runs to 3–5 weeks of tooling lead time and requires re-qualification under ASTM D4169 drop and vibration protocol.
Prevention — What to Specify Upfront to Avoid This Failure Mode #
On the design brief and PO, specify: body OD tolerance band (we recommend ±0.25mm for hermetic formats), wall caliper minimum (3.0mm for diameters ≥75mm, 2.5mm for sub-60mm), and the humidity conditioning window for bodies between winding and assembly (target 50% RH ±5%). Ask the supplier for their IMC or equivalent dimensional control log format before sampling starts — if they don’t have one, that is your signal to probe quality system readiness further.
For FEA or thermal simulation inputs, request dry-state and 65%-RH-state compressive modulus data for the specific paper grades used. Generic kraft modulus values from literature will not reflect the actual production material.
Request the first-article dimensional report to include pre- and post-conditioning OD measurements at minimum three lot sizes.
Specification Notes for Brand Partners #
When you brief us on a composite can program, the three inputs we need before we can commit to a first-article sample timeline are: target OD and height with tolerance expectations, intended end cap type (metal double-seam, friction-fit foil, or threaded plastic), and the downstream environment the filled can will see — specifically peak temperature and RH during shipping or retail display.
The brief gap that causes the most sample iteration is an underspecified end-use humidity condition. We receive briefs that state “ambient storage” without a temperature or RH range. For a composite can going into a climate-controlled US retail environment, that’s workable. For the same can going into Southeast Asian distribution with uncontrolled warehousing, it changes the seam specification, the liner WVTR requirement, and the body moisture conditioning protocol — all of which affect sample lead time.
Our standard first-article sample timeline for composite can formats is 18–22 working days from confirmed brief and approved dieline. Formats requiring custom seaming tooling add 15–20 working days for tooling fabrication and qualification. If the program requires food contact compliance under FDA 21 CFR §176.170 or EU 10/2011 migration testing, allow an additional 20–30 working days for third-party lab results.
What OD tolerance should I specify for a 90mm composite can going into a food application?
For a food-contact hermetic format at 90mm OD, we specify ±0.25mm on the body OD and a minimum 3.0mm wall caliper. Those numbers are tighter than general-purpose tube spec, but at 90mm the seam engagement depth is sensitive to even 0.3mm of OD drift, and food applications require the barrier integrity to hold through distribution conditions covered by ASTM D4169 Assurance Level II testing.
If I send my CAD model with nominal dimensions, can you just match them exactly?
Not on first article — and any supplier who says yes without qualifying it is telling you what you want to hear. Nominal CAD dimensions are targets, not process guarantees. What we do is review your CAD, map each critical dimension to our process capability (Cpk data for the relevant feature), and flag any dimension where our process Cpk is below 1.33. For most OD and wall features on composite cans, we hold Cpk ≥ 1.33. For liner bond edge height on sub-60mm formats, Cpk typically runs 1.10–1.20, which we disclose upfront.
Can thermal simulation be run before tooling is committed?
Yes, and we’d rather do it before tooling than after. The inputs we supply for thermal/mechanical simulation are: wall layer stack (paper grade, GSM, caliper, dry and conditioned modulus), liner material and bond peel strength (per ASTM D1876 T-peel), and end cap geometry with curl radius. With those inputs, a simulation will predict seam integrity under a 20°C-to-50°C thermal cycle to within roughly ±15% of empirical test results — sufficient to identify high-risk designs before committing to tooling spend.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
