Packaging Testing: Leak, Drop, Closure Fit, Torque and AQL 2.5
What is inspected, how it is inspected, and what the acceptance basis is — from incoming resin to the retained sample kept after the lot ships.
Inspection is a three-stage system, and the tests that matter run on the assembled unit
Inspection runs at three points: incoming (resin and closures), in process (on the line, continuously) and outgoing (on the finished lot). The tests that decide whether a pack actually works — leak, closure fit and torque — are run on the bottle, closure and dispenser assembled, because that is the object the user receives. A bottle that passes alone and a closure that passes alone have not been tested as a pack.
The test register
Every item below is a check that runs somewhere in the process. The acceptance basis column is the part that matters in a dispute — a test without an agreed acceptance criterion is not a test.
| Test | What it verifies | How it is run | Acceptance basis |
|---|---|---|---|
| Leak test | That the assembled pack does not leak with the closure fitted | Air pressure at 0.6–0.8 MPa, unit inverted, held | Zero leakage on the assembled unit |
| Drop test | That the pack survives handling and distribution | Vertical free-fall, three times, on packed units where the route justifies it | No cracking or loss of seal after the drops |
| Closure fit | That the closure seats without cross-threading | Neck finish and thread fit tested against your actual cap, pump or sprayer | Seals without cross-threading or weeping |
| Torque | That the closure can be applied and removed as intended | Application and removal torque measured on the assembled unit | Within the torque window agreed for that bottle and closure pair |
| Spray pattern | That the dispenser delivers the intended spray | Pattern and output checked on the assembled unit | Within the pattern and output agreed for the sprayer |
| Dimensional inspection | That the bottle matches the approved drawing | Height, diameter, neck and thread dimensions measured against the specification | Within the tolerances on the approved specification sheet |
| Weight inspection | That the wall has the material it should have | Bottle weight sampled and compared to the approved specification | ±2% against the approved specification |
| Wall thickness | That the wall is distributed correctly, not just right on average | Measured at four points per cavity | Within the wall tolerance on the specification sheet |
| Visual inspection | That appearance defects are caught inside the run | Every 30 minutes in process, plus AQL sampling on the finished lot | No defect outside the agreed defect classification |
| AQL 2.5 | That the finished lot is accepted on a defined sampling plan | Acceptance sampling agreed with you up front and applied to every batch | AQL 2.5, with the actual inspection results shared for your order |
| Batch traceability | That a shipment can still be traced after delivery | Resin lot → work order → finished-goods lot, walkable in both directions | A question about a shipment eighteen months later can be answered |
In-process inspection also covers print position, colour density and decoration adhesion where the pack is decorated, because a decoration that lifts is a quality failure even when the bottle is perfect.
Incoming inspection: the resin and the closures
Incoming inspection is where most downstream problems are prevented, because a lot of resin that arrives wrong cannot be corrected at the machine.
- ▸Resin grade and supplier batch number logged against the purchase record, so the material identity is documented rather than assumed.
- ▸Moisture content checked before drying — critical for PET clarity, because wet resin hydrolyses during processing and the bottle loses impact strength.
- ▸Melt flow index, density and ash content verified against the data sheet for PE and PP; intrinsic viscosity and ash for PET and PETG.
- ▸Certificate of Analysis per lot, held against the batch record rather than filed separately.
- ▸Masterbatch colour matched against the approved standard chip before the run starts.
- ▸Recycled content confirmed against the recycled-material supply-chain documents where PCR is specified.
- ▸Incoming cartons inspected so packaging defects are caught before they reach the line.
Drying parameters are set per material: PET and PE resin at roughly 170–180 °C for 3–4 hours, masterbatch at 100–120 °C for around 2 hours. Drying is a process control that shows up directly in clarity and strength.
In-process inspection: what runs continuously on the line
- ▸First-article inspection signed off before mass production starts, so the run begins from an approved part rather than from an assumption.
- ▸Visual defect inspection every 30 minutes on the running line.
- ▸Wall thickness at four points per cavity, because a single thin spot at the shoulder is a leak or a drop failure later and will not appear on a finished-goods weigh-in.
- ▸Weight sampled against the approved specification within ±2%.
- ▸Capacity and fill height spot-checked against the approved specification with the filling line’s headspace allowance.
- ▸Neck finish and thread fit checked against your closure — not against a nominal gauge, but against the cap, pump or sprayer you will actually use.
- ▸Print position, colour density and decoration adhesion checked where the pack is decorated.
Outgoing inspection and what AQL 2.5 actually means
AQL is an acceptance sampling plan, not a defect allowance you can spend. It defines how many units are inspected for a given lot size, and how many defective units the plan will accept before the lot is rejected. Two consequences follow from that.
Two things AQL is often misread as
Outgoing inspection covers: AQL sampling on the finished batch; a leak test on the assembled bottle and closure rather than the bottle alone; a drop test on packed units where the route justifies it; carton count, labelling and shipping marks checked against the packing list; and retained samples kept from the batch for later reference.
Wall thickness at four points per cavity, weight tolerance ±2%, visual defect inspection every 30 minutes in process, then AQL 2.5, a 1.2 m drop test, air-pressure leak testing on all closures and a QC report per lot.
Each lot is documented: a QC report per lot, and where the application needs it, resin COA, dimension records and test reports. A test report describes a specific article under specific conditions — it is evidence, and we do not describe it as a certification.
Batch traceability, and why retained samples matter
Traceability runs in two directions: resin lot → work order → finished-goods lot, and back. That two-way walk is what makes it possible to answer a question about a shipment eighteen months after it left, which is exactly when retail customers tend to ask one.
- ▸Records kept against the lot: resin lot and supplier batch, drying record, first-article sign-off, patrol inspection sheets, outgoing inspection result.
- ▸Retained samples kept from the batch, so a question can be answered by comparing the actual part rather than by reconstructing it from records.
- ▸Inspection sheets and batch records archived, so the trace survives well past delivery.
- ▸Documentation selected per market: the report your customer will ask for depends on the destination and the application — tell us which market and we will say what is normally requested.
What testing does not cover
Being clear about the boundary is part of being useful. Four things sit outside what in-house inspection can prove.
- ▸Fill compatibility over shelf life. That is a formulation test against a specific formula, and it belongs to the compatibility programme rather than to lot inspection.
- ▸Full distribution simulation. We drop test and leak test; a complete transit simulation through a specific route is normally a third-party study.
- ▸Regulatory approval. Conformity of the finished article to a market’s requirements is assessed by the party placing it on that market. We supply the underlying reports; we do not approve the product.
- ▸Anything not written into the specification. A parameter that is not on the approved specification sheet cannot be inspected against it, which is why the specification is a document rather than a conversation.
Where a project needs third-party testing, we prepare samples and supply the material and process data the laboratory asks for.
Related engineering and material reading
Packaging testing: buyer questions
What is AQL 2.5?+
AQL 2.5 is an acceptance sampling plan: it defines how many units are inspected for a given lot size and how many defective units the plan will accept before the lot is rejected. It is not a licence to ship 2.5% defective goods. We agree the sampling plan with you up front, inspect every batch against it, and share the actual inspection results for your order.
What tests do you run on every lot?+
In-process checks run continuously: visual defect inspection every 30 minutes, wall thickness at four points per cavity, weight within ±2% of the approved specification, capacity spot-checks, and neck finish and thread fit tested against your closure. On the finished lot we run AQL sampling, a leak test on the assembled unit, and a drop test on packed units where the route justifies it. Print position, colour density and decoration adhesion are checked where the pack is decorated.
How is the leak test done?+
On the assembled unit — bottle plus closure or dispenser — at an air pressure of 0.6–0.8 MPa with the unit inverted and held. Zero leakage is the acceptance criterion. Testing the bottle alone would not prove anything about the pack the user receives.
What is the drop test standard?+
A vertical free-fall of 1 m to 1.2 m, three times, with no cracking or loss of seal. Where a distribution route justifies it we drop test packed units rather than bare bottles, because packing changes how energy is absorbed.
Do you measure closure torque?+
Yes. Application and removal torque are measured on the assembled unit and compared against the window agreed for that bottle and closure pair. Torque targets are set per closure type and per neck finish rather than as a single figure, because a lotion pump and a screw cap behave differently.
How do you check wall thickness?+
At four points per cavity. An average is not enough: a single thin spot at the shoulder becomes a leak or a drop failure later, and it will not show up on a finished-goods weigh-in.
Can I get a QC report for my order?+
Yes — a QC report is issued per lot, and where the application calls for it we add resin COA, dimension records and test reports. The documents your customer will ask for depend on the destination market, so tell us the market and we will say what is normally requested.
Can you trace a batch after delivery?+
Yes. Traceability runs in both directions — resin lot to work order to finished-goods lot — and inspection sheets, batch records and retained samples are archived so a question raised months later can be answered by comparing the actual part rather than by reconstructing it from records.
Tell us what the pack has to survive
Send the fill, the fill temperature, the closure and the distribution route. We will agree the inspection plan and the tests worth running before the first lot is made.