TL;DR: Tolerance stackup in thermoformed packaging is a mechanical problem first — resolve it at the CAD stage before tool steel is cut, not after first samples fail fitment.
TL;DR: A draft angle under 3° on a 120mm-deep PP tray will generate demold forces exceeding 400N on a standard pneumatic ejection system, causing panel distortion visible to end consumers.
Where CAD Geometry Meets Thermoforming Reality #
There is a specific moment in packaging development when a structural designer’s CAD model — drawn to nominal dimensions, clean radii, uniform wall thickness — collides with what a thermoformed part actually is. That collision usually happens at first sample review, and it costs two to four weeks of tool modification time plus the rework fee. Getting ahead of it requires understanding how thermoforming deforms theoretical geometry, and feeding that understanding back into the CAD environment before tooling approval.
Thermoforming is a process that stretches a heated sheet over or into a mold cavity. The geometry you draw determines where that stretching concentrates. Every inside corner, every steep sidewall, every flat base panel with a large footprint — these are areas where wall thickness will be thinner than the nominal sheet input. For a standard packaging design workflow, ignoring this thinning distribution is the single most consistent source of structural failures we see at incoming sample inspection.
The governing variable is draw ratio: depth of draw divided by the smallest horizontal dimension of the cavity. For PP and PETG sheet between 0.6mm and 1.2mm nominal gauge, we consider a draw ratio above 0.8:1 to be high-risk territory for wall thinning below 50% of nominal. At that point, the base corner radius and the sidewall angle in your CAD model are not aesthetic parameters — they are structural specifications that determine whether the produced part meets ASTM D2659 column compression requirements or fails at 60% of the target load.
The Parameters That Predict Dimensional and Structural Outcomes #
Six parameters in a thermoform CAD model have direct, quantifiable production consequences. Most structural designers know three of them. The other three are where samples fail.
Draft angle is the most commonly misunderstood, and its threshold varies by material. For PP thermoformed trays, minimum 3° draft is the floor for reliable pneumatic demolding at cycle times under 8 seconds. PETG is stiffer and requires 4° minimum on female tooling. PVC, which some categories still use, can release at 2° due to its lower coefficient of friction against aluminum tooling — but PVC’s regulatory position under EU REACH and the current trajectory of PPWR 2025 makes it a difficult material choice for any EU-market packaging, draft angle advantages notwithstanding.
Base corner radius directly controls wall thinning concentration. Our standard internal specification, logged as DFM-Check 04 in our tooling approval workflow, flags any inside base radius below 2.0mm as requiring simulation review before tool release. Below 1.5mm inside radius on a draw ratio above 0.6:1, thinning at that corner will typically fall below 40% of nominal — and on 0.8mm input sheet, that leaves 0.32mm of actual wall at the highest stress concentration point.
Wall angle uniformity affects both dimensional stability and stack height. Asymmetric sidewall angles — common when packaging is designed around a product shape rather than a forming constraint — create uneven spring-back during cooling. We’ve measured lateral bow of 1.2–1.8mm on asymmetric trays that measured perfectly flat in simulation, because the simulation assumed uniform cooling. In production, the side with the steeper wall cools and contracts faster.
Flange width and flatness tolerance matters for lidding seal integrity. For MAP or vacuum-sealed food trays, our sealing validation requires flange flatness within ±0.3mm over a 200mm span, per the weld-head contact area requirements in ISO 11607-1. A CAD model with a nominal 8mm flange that tapers to 5mm at corners due to sheet draw will fail seal integrity testing even if the geometry “looks fine” on screen.
Nominal sheet gauge vs. produced wall thickness at critical zones needs to be specified separately. Most incoming briefs list sheet gauge. We need minimum acceptable wall thickness at three defined locations: the base corner, the sidewall midpoint, and the rim. These should be specified in the CAD drawing as minimum values, not derived from nominal gauge.
Shrinkage allowance is where I see the most tooling rework. PP shrinks 1.2–1.8% on cooling; PETG shrinks 0.3–0.6%; PET sheet typically 0.4–0.7%. These are directional — shrinkage along the machine direction of the input sheet differs from the transverse direction by up to 0.4 percentage points for some PP grades. A toolmaker who ignores this anisotropy will cut a symmetric cavity that produces asymmetric parts.
| Material | Typical Shrinkage (MD) | Typical Shrinkage (TD) | Min. Draft Angle (Female Tool) |
|---|---|---|---|
| PP homopolymer | 1.4–1.8% | 1.0–1.4% | 3° |
| PETG | 0.3–0.5% | 0.3–0.5% | 4° |
| PET (A-PET) | 0.4–0.6% | 0.4–0.7% | 3.5° |
| rPET (≥50% recycled) | 0.5–0.9% | 0.5–0.8% | 4° |
| HIPS | 0.3–0.6% | 0.3–0.6% | 2.5° |
rPET shrinkage range is wider because recycled content IV (intrinsic viscosity) varies between lots. We track IV on each incoming rPET lot against our QC specification of 0.72–0.82 dL/g. When IV drops below 0.72, shrinkage behavior becomes less predictable and we flag the lot under our incoming material review — Form IMR-12 — before releasing to production.
Simulation Inputs, Tolerance Stackup, and Where Decisions Branch #
If your design team is running FEA or thermoforming process simulation before tooling, the inputs that matter most are material-specific: elastic modulus at forming temperature (typically 170–210°C for PP, 90–130°C for PETG), the sheet gauge tolerance band from your material supplier (typically ±5–8% for commercial-grade reel), and the mold surface temperature differential (we hold aluminum tooling surface temperature between 15–25°C for PP on our pressure-forming lines).
Thermal simulation inputs for wall thinning prediction should use a friction coefficient of 0.25–0.35 for PP on aluminum (dry), 0.15–0.25 for PETG. These matter for plug-assist tool design: an incorrect friction assumption in simulation changes the predicted plug penetration depth by 8–12mm on a 100mm-deep cavity, which flows directly into incorrect wall thickness predictions at the base.
Tolerance stackup decisions follow a branching logic:
If the part is a tray with a formed-in snap lid, and the lid-to-tray closure force must fall within a defined range (typically 15–35N for consumer-acceptable open-close on PP), then the tolerance on the tray rim width and the lid flange thickness must be independently specified and verified. A nominal-to-nominal stack may look acceptable — 0.5mm on each component, 1.0mm total stack — but if the rim width and lid flange tolerances are correlated (both affected by the same sheet gauge lot), the actual worst-case stack is not 1.0mm. It may be 1.6mm or higher, enough to make the snap closure either inoperable or unintentionally permanent.
If the part is an open tray with a heat-sealed lid film, the tolerance that governs function is flange flatness and sealing temperature uniformity, not lateral dimensional tolerance. In this case, investing tighter tolerance budget in flange flatness (holding ±0.25mm vs. a typical ±0.5mm) while relaxing lateral body tolerance to ±0.8mm is the correct engineering tradeoff. Tighter lateral tolerance costs money in tooling and process control; it has zero effect on seal integrity.
If the part must nest or stack, and stack height in a retail display unit is specified to a maximum, then the draft angle and flange height must be co-designed. A 1° increase in draft angle on a 150mm-diameter circular tray changes the nested stack height by approximately 2.5mm per unit — meaning 10 nested units shift total stack height by 25mm. That is a real retail display engineering consequence, and it should be calculated at the CAD stage with a parametric model, not discovered at production trial.
One non-obvious recommendation with a specific boundary: for any part where flange seal integrity is critical (food contact, MAP, vacuum), specify the flange surface roughness Ra value in the tooling spec. We specify Ra ≤ 0.8µm on sealing surface areas of our aluminum tooling. Above Ra 1.2µm, seal peel strength variability increases by 20–30% in our validation testing, and achieving consistent ISO 11607-1 seal integrity scores requires lower sealing temperatures that slow cycle time. This holds for flat-flange trays; profiled or textured flanges require individual sealing validation regardless of Ra value.
Specification Notes for Brand Partners #
When you brief us on a thermoformed packaging project, the most useful document you can send is not a rendering — it is a dimensioned 2D drawing with tolerances called out at functional features, plus a clear statement of the material and the end-use environment (temperature range, product contact type, distribution stress).
The most common brief gap we see is missing fitment data: what does this tray need to fit into, stack on, or seal against? A tray designed in isolation, submitted without the lid film specification or the carton inner dimension it sits inside, typically requires two extra sample iterations. If you have a lidding film spec, send the seal layer material and target peel strength range. If the tray nests in a shipper carton, send the inner dimensions.
Our standard sampling timeline for thermoformed tooling is 18–25 working days from tooling approval to first sample delivery, assuming standard aluminum tool construction. Tooling for pressure-forming or parts with tight flange tolerances (±0.2mm or tighter) adds 5–8 working days. Complex multi-cavity tools for high-volume formats run 30–40 working days.
How much draft angle do I actually need, and can I go below 3°?
It depends on the material and tool surface treatment. For PP on standard anodized aluminum, 3° is the minimum we’ll commit to for consistent demolding at commercial cycle times. If the tool receives a PTFE or DLC coating, we’ve successfully run 2° draft on shallow PP parts (draw ratio below 0.4:1), but we treat those as case-by-case approvals, not a default. For PETG, stay at 4° unless the draw is very shallow.
What’s the minimum base corner radius I should specify in CAD?
For any draw ratio above 0.5:1, specify inside base radius at 3.0mm minimum. At 2.0mm you’re in our DFM-Check 04 review zone. Below 1.5mm on moderate or deep draws, we’ll propose a tooling modification before cutting. The functional reason is wall thinning at the corner — not aesthetics.
Can I run FEA structural simulation on a nominal-wall CAD model?
You can, but the results will be optimistic. FEA on a nominal-wall model assumes uniform thickness that doesn’t exist in the produced part. For any structural application where ASTM D2659 column strength or ASTM D4169 distribution simulation compliance matters, run the simulation with the predicted minimum wall thickness at the base corner, not nominal. That’s the number that determines real-world performance.
What shrinkage value should I build into the CAD model for PP?
Use 1.5% as a starting point for symmetric cavities. For asymmetric part geometries, apply 1.6–1.8% in the machine direction of the sheet and 1.1–1.4% in the transverse direction. Your toolmaker should adjust from first-sample measurement data — we typically reach final tool dimensions within two correction cycles on standard PP grades.
Does rPET require different CAD parameters than virgin PET?
Yes, specifically on shrinkage and flange flatness. rPET at ≥50% recycled content has wider shrinkage variability (0.5–0.9% MD) compared to virgin A-PET (0.4–0.6% MD). We build this into tool offset calculations and track it lot by lot via IV testing. For flange-sealed food trays using rPET, we recommend tightening the flange flatness spec to ±0.25mm rather than the ±0.3mm used for virgin PET — the wider shrinkage band can push you to the tolerance boundary under adverse conditions.
What tolerances are realistic for commercial thermoforming, and when does tighter cost significantly more?
Standard commercial tolerances for thermoformed parts are ±0.5mm on external body dimensions, ±0.3mm on flange width, ±0.25mm on flange flatness over a 150mm span. Tightening to ±0.2mm on body dimensions requires pressure forming (not vacuum), adds tooling cost, and slows cycle time. For most packaging applications, the commercial tolerance is functionally adequate — tighter tolerances are worth specifying only where fitment, seal integrity, or retail stack height is the constraint.
Is there a wall thickness specification I should include in the drawing — or is sheet gauge enough?
Sheet gauge alone is not sufficient for any part where structural or barrier performance matters. We ask for a minimum wall thickness at three points: base corner, sidewall midpoint, and rim. For food trays, minimum 0.25mm at the base corner is our internal threshold below which we flag for structural re-evaluation regardless of nominal sheet gauge. Including these minimums in your drawing prevents ambiguity at first sample inspection and gives us a clear pass/fail target.
Planning a packaging project? Contact our team to request a complimentary specification review and sample quote.
