QUICK ANSWER
A mold trial is an experiment with agreed requirements, controlled inputs and traceable outputs. Accept the tool against documented geometry, function, process and handover criteria—not a small selection of attractive parts. Define the evidence and responsible approvers before the trial.
Engineering reference, not a material specification or machine-setting recipe. Validate decisions against the released drawing, resin-grade data and equipment instructions. Equipment work must follow the applicable safety procedures.
1. Define the decision before running the trial
T0, T1 and similar labels are project conventions, not universal acceptance standards. One team uses T0 for the first safe mechanical check; another expects dimensional samples at that stage. State the purpose explicitly: mechanical readiness, process development, dimensional correction, appearance approval or production qualification. Combining all of these into a single pass/fail label hides unfinished work.
Freeze the drawing revision, CAD revision, resin grade, color and permitted additive or regrind condition. Identify critical dimensions and functions, cosmetic zones and approved inspection methods. A tolerance written without a datum scheme or conditioning requirement may produce disagreement even when both parties measure carefully. Assign a decision owner to each unresolved requirement rather than allowing an unrecorded shop-floor assumption to become the acceptance criterion.
- List required evidence, acceptance criteria and the person authorized to approve each item.
- Record deviations separately from approved requirements; a concession is not an invisible drawing change.
2. Make samples traceable to the actual process
A sample without context is difficult to use for correction. Keep cavity identity, shot or time sequence, material lot and process revision with the parts. Record the machine and relevant screw configuration, melt delivery system, cooling connection map and the actual observations needed to reproduce the condition. Distinguish measured process responses from controller setpoints.
Do not mix warm-up parts with steady-state evaluation samples without labeling them. Define the stabilization criterion with the molding team and record changes during the run. Keep complete cavity sets where cavity balance or local variation matters. Selecting only the best cavity can conceal a tool problem; combining measurements from unlike cavities can conceal separate distributions.
| Evidence | Minimum question it should answer | Common weakness |
|---|---|---|
| Sample identification | Which cavity, sequence and condition produced this part? | Loose parts with no cavity or time record |
| Process record | Can the trial condition be reconstructed? | A photograph of only one settings screen |
| Inspection report | Which revision, datum setup and conditioning were used? | Dimensions listed without a measurement method |
| Issue register | Who owns the correction and how will it be verified? | Verbal agreement with no closure evidence |
3. Inspect the right condition, not just the easy condition
Define when measurements occur and what environmental or post-molding conditioning is required. Some polymers change dimensions after ejection through cooling, moisture uptake or structural relaxation. The inspection interval and condition must match the product requirement and material behavior. A comparison between freshly molded trial parts and conditioned reference parts can create a false steel-correction request.
Check the measurement system before deciding that a small difference is a molding change. Fixturing can distort a flexible part, contact force can change a reading, and different datum alignment can move a reported feature. Use the customer-approved method for critical characteristics. For a functional problem, retain the mating component or a defined functional gauge so that the team does not optimize a dimension while losing the assembly requirement.
4. Separate sample compliance from process capability
A handful of parts inside tolerance does not demonstrate long-term capability. Capability analysis needs an appropriate measurement system, a stable process, a justified sampling plan and a statistical method suited to the data. The relevant variation may include cavity, time, material lot, restart and machine effects; which of these must be studied is a project decision.
NIST's process-capability guidance treats stability and distribution assumptions as prerequisites to meaningful interpretation of common indices. Do not attach an impressive Cpk number to a short, selectively collected trial and call the mold qualified. Record what the study actually supports. A development trial can provide useful evidence for correction without claiming that every production qualification requirement is complete.
5. Close corrections with a controlled handover
Create an issue register that links each symptom to a sample, a requirement, an investigation and an owner. Agree the proposed action and its expected effect before changing steel. Recheck connected features after a correction: changing a gate, cooling circuit or local section can influence dimensions, appearance and cycle behavior beyond the original issue.
The release package should distinguish the approved configuration from open actions. Include the final relevant drawings, cooling and electrical identification, approved process record, inspection evidence, spare-component information and maintenance instructions appropriate to the tool. List remaining restrictions clearly. A shipment date or a purchase-order milestone is not technical evidence that an unresolved performance requirement has been met.
PUT IT INTO PRACTICE
Engineering review checklist
Use these prompts for a project review. This page does not record or certify completion.
- Agree what this trial stage must prove and who can approve it.
- Freeze drawing, material and process revisions; label every retained sample set.
- Confirm conditioning, datum setup, fixtures and measurement-system suitability.
- Keep cavity and sequence information rather than selecting only good-looking parts.
- Separate development results, concessions and capability or qualification claims.
- Close each correction with repeat evidence and a controlled handover record.
Questions engineers ask
Does a T1 trial automatically mean the mold is ready for production?
No. T1 is a project stage label. Readiness depends on the agreed requirements and evidence. A successful dimensional sample may still leave process, appearance, documentation or qualification actions open.
Should every dimensional deviation be corrected by changing steel?
No. First confirm the measurement method, conditioning and process stability. Then determine whether the deviation is a repeatable tool-related effect, a process effect or a requirement interpretation issue. Approve steel changes only after that review.
Sources and application limits
This is independently written educational guidance. The linked manufacturer and engineering references support specific mechanisms or further study; they do not endorse PFMOLD or replace project approval. Supplier recommendations apply to their stated grades, tooling and equipment.
- NIST Engineering Statistics Handbook — What is Process Capability? ↗
Supports the limited discussion of capability assumptions. It is not a mold-acceptance specification or a project sampling plan.
- BASF — Injection Molding Troubleshooter ↗
A manufacturer reference for defect mechanisms and investigation; use grade-specific technical support for an actual trial.