PET bottle wall thickness measurement is the quality check most often done badly and then trusted. A single reading taken at one point on one bottle tells you almost nothing about a production run, and a gauge used without a defined control section produces numbers that cannot be compared between shifts, machines or suppliers. The result is a plant that measures constantly and still cannot answer a customer complaint about a collapsing bottle.
The consequences arrive downstream. A bottle that is thin at the base fails under stack load in a warehouse rather than on the line, so the batch has already been palletised and shipped. A bottle that is thin at the shoulder splits during filling. A bottle that is thick everywhere passes every test and quietly consumes more resin than the design required, which is the most expensive failure because nobody notices it at all.
Sailwin has delivered PET stretch blow molding machines into 500+ installations across 60+ countries over 15+ years, manufacturing under ISO 9001:2015 with CE marking. Machines run preform heating between 90 °C and 115 °C under PID control holding approximately ±1 °C and blow at 25–40 bar, and the PLC monitors 40+ parameters in real time — which means wall thickness variation can be traced to a machine state rather than argued about. This guide covers the measurement methods, where to take readings, and how to set tolerances a production team can actually hold.
Key Takeaways
- A control section must be defined before any measurement is taken. Without agreed measurement points, readings cannot be compared between shifts, machines or suppliers, and the data is worthless even when it is accurate.
- Different methods answer different questions. Sectioning gives the true distribution, ultrasonic gauges give speed without destruction, and bottle weight gives a fast proxy for total resin — none of them replaces the others.
- Tolerance bands should be set from function, not from the drawing number. Base, body and shoulder carry different loads and should not share one tolerance just because a table lists one number.
Send Your Bottle Specification for a Wall Thickness Control Plan
Share the container drawing, target weight and load requirement — our engineers return a measurement plan, machine recommendation and cavity count within 24 hours.
1. The Four Measurement Methods and What Each Is For
Wall thickness on a PET bottle is not uniform, and it is not meant to be. The base is thickest because it carries stack load, the body wall is thinnest because it carries only internal pressure and handling, and the shoulder transitions between the two. Any measurement method has to be understood in that context, or the numbers will be read as defects when they are simply the design.
| Method | What it gives you | Limitation | Where it belongs |
|---|---|---|---|
| Sectioning and optical measurement | The true wall thickness distribution around the full circumference and height | Destructive, slow, and depends on how accurately the section is cut and presented | Tool qualification, mould acceptance and investigating a complaint |
| Ultrasonic thickness gauge | Point readings on a finished bottle without destroying it | Reads only the point it is placed on; coupling and curvature affect the result | Routine production control at agreed control points |
| Bottle weight | A fast, repeatable proxy for total resin in the part | Distributes nothing: a bottle can hit the target weight and still be thin where it matters | Every-shift monitoring and changeover verification |
| Top-load and burst testing | The functional result the wall thickness is there to deliver | Tells you that a bottle failed, not which section caused it | Design validation and periodic confirmation, paired with sectioning |
The practical combination is bottle weight for routine monitoring, ultrasonic readings at defined control points for verification, and sectioning whenever a number moves. Weight alone will not catch a distribution problem, and sectioning alone is too slow to run every shift.
2. Where to Measure: Defining a Control Section
A control section is the set of named points where readings are taken on every bottle that is measured. It should be written down once, with a height above the base and a clock position for each point, and it should never be redefined casually. The reason is comparability: a reading of a given value means nothing unless the next person takes it in the same place with the same method.
- Base, at the point that touches the pallet and the point that carries the standing ring. This is where stack load concentrates and where a thin section causes a shipped batch to fail.
- Lower body, just above the label panel. Typically the thinnest area of the wall and the most sensitive to stretch ratio and heating profile.
- Shoulder, midway between body and neck. The point where the material is still moving as the mould closes, and where variation appears first when the heating profile drifts.
- Neck and finish. Not a body-wall measurement, but part of the same record, because finish dimensions and body wall together decide whether the bottle works on a filling line.
Take each control point at least twice, on opposite sides of the bottle, and measure more than one bottle from a cavity. Cavity-to-cavity variation is real, and a single bottle from a single cavity is an anecdote rather than a sample. Sailwin machines span 2-cavity to 8-cavity configurations — from the SW-F2-650 at 2,800 bottles per hour up to the SW-F8H-800 at 16,000 bottles per hour — so the number of cavities to sample follows directly from the machine model.
Two readings from one bottle are a data point. Two readings from each cavity are a sample. Only a sample supports a decision about a mould or a process.
3. Setting Tolerances and a Sampling Plan That Holds
Wall thickness tolerance should be derived from what each section has to do. The base has to survive stack load, so its minimum is set by the top-load requirement. The body wall has to contain pressure and resist handling, so its minimum is set by burst and drop testing. The shoulder is a transition and is judged mainly on consistency, because inconsistency there usually indicates a drifting heating profile or a cooling imbalance rather than a design problem.
Once the minima are known, the tolerance band should be set around the process capability rather than around the drawing. A band your machine cannot hold every hour of every shift is not a quality standard; it is a source of unnecessary rejections and of learning to ignore the data. Sailwin machines hold preform heating under PID control at approximately ±1 °C, which is what makes a tight band holdable — without stable temperature control, a tolerance that works in the afternoon fails on the night shift.
| Check | Frequency | Sample | Decision it supports |
|---|---|---|---|
| Bottle weight | Every shift, and after every start-up and changeover | Several bottles, taken across the run and not only at start-up | Whether the process is producing to the resin target |
| Ultrasonic readings at control points | Daily, and whenever a machine parameter changes | At least two opposite points per section, across cavities | Whether the distribution is still inside the band |
| Sectioning | At tool qualification, and on any unexplained weight or gauge shift | One bottle per cavity, sectioned at the agreed heights | Which section moved and by how much |
| Top-load and drop test | Design validation, then periodic confirmation | Per the agreed protocol, conditioned to the conditions the product will meet | Whether the wall thickness minima are still delivering the function |
Record the settings alongside the readings. Because the PLC monitors 40+ parameters in real time, a wall thickness shift can usually be matched to a specific machine state — a heating zone, a blow pressure, a mould temperature — instead of being attributed to the mould. That is the difference between correcting a process and changing a tool.
Want a Measurement Plan That Matches Your Bottle and Line?
Send the container drawing, target weight and load requirement — we return a control section, sampling plan, cavity count and machine model within 24 hours.
4. Case Study: A Water Bottle Failing Shipment Drop Tests
A bottled water producer had a 1 L container pass in-plant checks and fail drop tests after transport, while bottle weight was consistently inside the target band.
- Bottle weight inside target while bottles still failed drop tests after transport
- Wall thickness readings that could not be compared between shifts because measurement points had never been fixed
- A distribution problem being investigated as a total-material problem
- Control section defined with named heights and clock positions, so readings became comparable between shifts
- Sectioning carried out across cavities at the agreed heights to show where the material actually sat
- Machine state reviewed against the PLC parameter record instead of adjusting the mould
- Sampling plan introduced with weight per shift and gauge readings daily, tied to the machine’s cavity configuration
- The cause was identified as distribution rather than weight, which stopped a mould modification that would not have fixed anything
- Measurements became evidence, because a defined control section allowed readings from different shifts to be compared directly
- The corrective action was validated upstream, at the blowing stage rather than at the warehouse where the failure appeared
Scenario based on a Sailwin customer project; site-specific figures available on request during engineering review.
Frequently Asked Questions
Get a Wall Thickness Control Plan Built Around Your Bottle
Send your bottle drawing, container sample or target output. Our engineering team replies with a machine recommendation, mould assessment and factory-direct quotation within 24 hours.
Related Reading:
• PET Blow Molding Machines — 2-cavity to 8-cavity models
• Lightweighting PET Bottles: Resin Savings Without Losing Strength
• PET Blow Molding Machine Buying Guide
• PET Water Bottle Blowing Machines
• Wide-Neck PET Jar Blow Molding Machines
• Injection Molding Machines for preforms
