A bubble-like mark is a symptom, not a root-cause label. It may be trapped air, moisture vapour, volatile material, polymer degradation gas, a shrinkage-driven vacuum void, or an indication that needs further investigation. The same visible feature can carry a different risk when it sits in a cosmetic cover, a sealing land, a snap feature, a pressure boundary or a loaded section. Procurement and quality teams should therefore avoid accepting or rejecting parts from appearance alone.
For an overseas sourcing program, the useful question is: what evidence shows the indication’s mechanism, location, size, repeatability and functional consequence? That evidence should be connected to the approved resin grade, geometry, mold cavity, process revision and inspection method. This guide concentrates on that buyer evidence path—especially how to separate gas bubbles from shrinkage voids and how to set a defensible internal-defect acceptance plan—rather than treating a single process setting or tool change as a universal solution.
Do not write “no bubbles” as a stand-alone requirement. Define the region at risk, the inspection route, the measurable acceptance limit where one is needed, the sampling basis and the disposition path for an indication. A visual observation alone may not distinguish gas from a vacuum void.
01
1. Classify the symptom before assigning a cause
Start with a controlled symptom record. Photograph the part under repeatable lighting, identify the drawing zone and visible face, record cavity number, resin lot, colour or additive system, molding date, tool and process revision, and note whether the condition is visible at ejection, after conditioning or only after a functional test. A raised blister, a subsurface round feature, a surface depression and a CT indication should not be placed in one defect bucket. Their position relative to the gate, weld area, rib, boss, thick section, insert or end-of-fill region is useful evidence, but it is not proof of mechanism.
The first working split is between gas-related bubbles and shrinkage-driven vacuum voids. Gas can be associated with entrained air, inadequate venting, moisture or volatiles, contamination, excessive residence history or material degradation. A vacuum void can form as a heavier section cools and contracts after the surface has become too rigid to follow the shrinkage. These are hypotheses, not interchangeable diagnoses. Resin family, moisture sensitivity, filler content, local wall mass, gate freeze behaviour, mold temperature balance, machine response and the age of the part can all change what is observed.
Ask the supplier to propose confirmation checks before corrective action. Depending on the part and defect size, that can include cavity-coded samples, a short-shot flow study, a controlled hold-stage comparison, part-weight tracking, resin moisture verification using a method suitable for the grade, a carefully controlled heat response check, sectioning, or non-destructive CT. Any heat or sectioning check should be authorized and documented because it can alter the sample. The corrective action should follow the confirmed mechanism; increasing pressure to address an unvented gas source, for example, can worsen another condition.
02
2. Investigate gas-related bubbles through material, flow and venting evidence
A repeatable bubble trail near the end of fill, behind a core pin, at a rib tip, around a weld region or near a hot-runner drop can justify a venting and flow-path investigation. A more dispersed pattern may point the team toward resin condition, volatile contamination, excessive decompression, nozzle or hot-runner leakage, or an unfavorable melt history. These patterns are starting clues only. Similar-looking marks can result from different mechanisms, and one program can have more than one mechanism at the same time.
A buyer can ask for a bounded evidence set: the exact material designation and approved regrind rule; dryer or material-handling records; an agreed moisture result or test method where the resin requires it; evidence of purge and changeover practice; relevant barrel and residence-time review; cavity-specific occurrence data; and a tool inspection record for vents, pins, plate interfaces and hot-runner components as applicable. The supplier should also state which variables were held constant during a trial. A settings printout without material and sample traceability is not enough to establish cause.
Corrective actions may include restoring material handling to the resin supplier’s requirements, removing a contamination source, cleaning or repairing venting features, correcting a mechanical air-ingress path, or revising a flow, gate or process-control approach. The correct combination is program-specific. Do not impose generic drying times, vent depths, barrel temperatures, back pressure or decompression values in an RFQ. Those values depend on the polymer grade, allowable residence time, geometry, mold design, machine, screw and measurement system; their suitability needs confirmation on the actual part.
03
3. Separate shrinkage voids from other hidden integrity concerns
Shrinkage voids are commonly associated with a local mass that cools more slowly than its surroundings, such as a boss base, rib intersection, thick transition, pad or enclosed section. The procurement concern is not merely whether an internal cavity exists. It is whether its position, dimensions and repeatability reduce sealing, strength, fatigue life, electrical isolation, optical performance or another defined product requirement. A void away from a critical function may have a different disposition from a smaller indication on a pressure-retaining wall or a loaded latch.
Confirm the working hypothesis by correlating the indication with a wall-section and flow-path map. Request cavity-coded baseline samples, specimen age and conditioning state, part weight or cavity-pressure information where available, and an evidence-based comparison of hold-stage and cooling conditions. A gate-freeze study may help establish whether additional material can still reach the region; a sectioned sample can expose location and morphology. If a non-destructive method is selected, agree its capability for the specified material thickness and minimum feature size before interpreting a negative result as proof that no smaller void exists.
Potential corrective actions can include a validated packing or transfer-control adjustment, a gate or runner review, cooling-balance work, or a geometry revision that reduces concentrated mass while preserving function. Each option has trade-offs: a change that improves a local void may affect flash, sinks, residual stress, dimensional stability, weld-line position, cycle time or tool maintenance. Require the supplier to show the baseline, the single-variable or planned comparison, adverse observations and the post-change inspection result. Do not approve steel changes or a broad process-window change solely because one optimized sample looks acceptable.
04
4. Build acceptance criteria around function and a capable inspection method
Acceptance criteria should state the product risk rather than a vague visual preference. Define the controlled region on the drawing or a marked image; the defect type being assessed; the largest permitted indication or permitted absence of indications; count, spacing or area coverage if relevant; sample condition; lot or cavity sampling plan; and the decision authority for a deviation. Where a void can affect a seal, load path, electrical clearance or regulatory requirement, add the relevant functional or downstream test. A universal maximum bubble diameter is rarely technically sound because allowable risk varies with resin, wall thickness, local stress, service environment and end-use function.
Use the least complex method that can answer the acceptance question reliably. Visual inspection can screen surface blisters and sinks but cannot prove internal integrity. Sectioning can give direct local evidence but destroys the sample and may miss an indication outside the cut plane. Industrial X-ray CT can reveal and locate internal features without cutting; however, its usable resolution, contrast, artifact control, segmentation threshold, scan setup and measurement uncertainty need to be demonstrated for the part. ASTM E1441 describes CT as a guide and expressly does not establish accept/reject criteria for a new object. Buyers still need to define the product-specific criterion and validate the agreed method.
Request an inspection report that identifies the part and revision, material, cavity and lot, equipment or method, scan or section location, resolution or method capability where relevant, operator or laboratory, raw-data retention approach and result against the stated criterion. For CT, clarify whether the supplier will retain reconstructed data and annotated views, not just a pass/fail image. The goal is traceable evidence that an internal indication was assessed consistently, not an impressive-looking scan with an undefined threshold.
05
5. Turn a finding into a controlled supplier-corrective-action and release decision
When an internal defect is found, first contain the affected population. Ask which cavities, dates, material lots, tool-maintenance events, settings revisions and shipment lots could be involved, and how the supplier has segregated them. Then require a problem statement that separates the observed symptom, the most likely causes under investigation, the confirmation checks, the interim containment and the proposed permanent corrective action. A corrective-action report should not claim closure until the mechanism has been tested and the revised process or tool condition has produced evidence against the agreed acceptance plan.
For supplier selection or a production release, practical buyer questions include: Which CTQs could this indication affect? Can you show cavity-by-cavity occurrence, not only pooled samples? What material and conditioning records link to the samples? What measurement-system capability supports the claimed detection limit? Which process variables were changed, what remained fixed, and what side effects were evaluated? Will the change require a new sample approval, functional test, inspection update or customer deviation? These questions make commercial decisions auditable across OEM, engineering, quality and procurement teams.
Release only against a documented basis: approved drawing and material revision, defined acceptance criteria, representative sample and inspection evidence, any required functional result, disposition of deviations, and change-control responsibilities. Continued monitoring can be proportionate to risk, such as cavity-coded start-up checks, periodic internal inspection, defect trend review or requalification after a resin, tool, machine or process change. The needed control level depends on the application. It should be agreed with the supplier and end-use owner rather than inferred from a successful trial or transferred unchanged from another resin, geometry or mold.
Practical takeaway
Before you move
to the next review.
- Identify the suspected defect by drawing zone, cavity, resin lot, sample age and visible or functional consequence before naming the cause.
- Request a supplier investigation that separately states the symptom, likely mechanisms, confirmation checks, interim containment and permanent corrective action.
- For gas-bubble investigations, review material-condition evidence, flow-path and venting observations, and relevant machine or hot-runner checks rather than relying on a settings sheet alone.
- For suspected voids, link the indication to local wall mass, gate and cooling evidence, then compare controlled samples before authorizing a tool or geometry change.
- Approve acceptance criteria only after defining the critical region, inspection method capability, sampling basis, traceable reporting and any required functional test.
FAQ
Common questions
Can a buyer identify a vacuum void just by its shape or location?
No. Shape and location can guide a hypothesis, but they are not conclusive. Record the occurrence and use an agreed confirmation route, such as controlled trials, sectioning or appropriately capable CT, before choosing a corrective action. Part age can also affect how an indication appears.
Should every internal indication fail an injection molded part?
Not automatically. The disposition should depend on its location, size, repeatability, material, part geometry and effect on the intended function. An indication in a non-critical zone may be evaluated differently from one in a sealing, structural, electrical or cosmetic-critical region. Define that basis before production release.
Is industrial CT a universal replacement for destructive sectioning?
No. CT is valuable for non-destructive three-dimensional inspection, but its ability to detect and size a feature depends on part thickness, material contrast, scan setup, voxel size, artifacts and analysis method. Sectioning can remain useful for correlation or local confirmation. Validate the chosen method against the acceptance question.
What evidence should accompany a supplier’s claim that bubbles were corrected?
Request cavity- and lot-traceable baseline and post-action samples, the confirmed mechanism, material and process revision details, a description of the controlled comparison, inspection results against the agreed criterion, any relevant functional test, and confirmation of whether the change triggers reapproval. A photo of one improved part is not sufficient closure evidence.
References
Source material
- [1] BASF — Injection Molding Problems in Engineering Thermoplastics: Causes and Solutions
- [2] ASTM International — ASTM E1441, Standard Guide for Computed Tomography (CT)
- [3] National Institute of Standards and Technology — Getting the Inside Story on Products with Computed Tomography
- [4] U.S. Food and Drug Administration — Process Validation: General Principles and Practices
This guide is general information for early procurement and engineering discussion. It does not replace product-specific DFM, qualification, validation, regulatory assessment or legal advice.