The scale still reads 10.00 g. The opening is narrower. That does not mean one instrument must be wrong: whole-part mass and a local dimension describe different things.
When injection molded parts have the same reported weight but different dimensions, first make the measurements comparable. Check what was weighed, which cavity produced it, where and how the dimension was measured, and how long the part had been out of the mold. If a dimensional difference remains under the specified method, investigate it even when the weight check passes.
This guide helps operators and developing technicians organize that comparison with quality and process staff. It does not set tolerances, conditioning periods, or process adjustments.
What, exactly, stayed the same?
Start by retrieving the individual results. Two values displayed as 10.00 g are equal at the recorded resolution; that does not establish identical mass at greater precision. A group average can also stay constant while individual samples change. Keep the original precision and sample count rather than adding decimal places later.
Confirm the object being weighed. Does the record refer to one trimmed part, all cavity parts together, or the entire shot with runner and sprue? Were gate remnants, flash, or attached material treated consistently? For a cavity-specific dimensional concern, a mold-wide total cannot substitute for the relevant part’s measurement.
RJG’s discussion of part weight and molding quality makes the central distinction: equal weight does not establish equal dimensions or internal conditions. Weight remains a useful process signal, but it cannot independently certify a dimensional requirement. RJG sells monitoring systems; this principle does not imply that buying a sensor replaces inspection.
Ask what conclusion the weight check was designed to support. “Within our weight-control band” and “meets the opening-width tolerance” belong in separate fields unless the approved quality plan explicitly establishes a valid relationship. If a dimension fails its requirement, do not use a passing weight value to erase that result.
Make sure the same dimension is being measured
A note saying “width is down” is incomplete. Identify the drawing revision, characteristic number, datum or locating method, and measurement position. An opening can be checked near one end, near the other, or at a defined section. Those readings may answer different questions.
Record the instrument and fixture IDs, applicable method revision, and person or system performing the measurement. For a flexible part, the agreed contact and support method matters. Ask quality staff to resolve a method disagreement rather than squeezing, clamping, or relocating the part until the number matches.
| Check | What to put in the record |
|---|---|
| Part identity | Part/revision, cavity, sample ID, molding time, material/lot reference, and trial or production group. |
| Characteristic | Drawing feature ID, nominal and tolerance reference, measurement location, and specified datums or support. |
| Method | Instruction revision, instrument and fixture IDs, relevant instrument-status check, and retained individual readings. |
| Sample history | Time out of the mold, storage/conditioning record, measurement time, and relevant environment record. |
| Disposition | Inspection result under the applicable method, uncertain records, sample/product location, and responsible quality owner. |
If two inspectors disagree, use the same identified samples for a quality-led comparison under the specified method. Record each reading and the sequence. Repeatedly remeasuring only until an acceptable number appears destroys the usefulness of the comparison.
A method discrepancy does not automatically invalidate a reported defect. Keep the original observation visible while the responsible owner determines whether the measurement is suitable for the requirement. Preserve both the first result and any corrected result with an explanation.
The clock belongs beside the dimension
Covestro’s guide to shrinkage in thermoplastics, printed page 4, distinguishes dimensional change at and after demolding. It also discusses later change associated with stress relaxation, further crystallization in semicrystalline materials, and moisture uptake in materials such as polyamides. These mechanisms depend on the material and history; they do not supply one waiting period for every part.
- Out of the moldSample ID, cavity, and molding timestamp.
- Storage / conditioningConditions, duration, and any missing interval.
- MeasurementTimestamp, method, instrument, and individual result.
ASTM D618 addresses conditioning plastics for testing. Its public scope and significance statement explains why temperature and humidity conditions matter to comparisons, and that an applicable material specification takes precedence. The full paid standard was not used for this guide. Do not infer that every molded-part inspection must follow D618 or that one generic conditioning routine meets a customer’s drawing.
For the actual job, cite the required inspection or material specification. Write “conditioning history unknown” when it is unknown. A timestamp showing that both groups were measured at 9 a.m. is insufficient if one group was molded minutes earlier and the other the previous day.
If timing differs, ask whether the retained samples can still support the planned question. Measuring them again later may help, but cannot recreate a missing earlier condition. When a replacement comparison is needed, agree on the collection and measurement schedule before producing another sample group.
Four matching weights, one unresolved comparison
Constructed teaching example. All sample IDs, values, and times below are invented. They are not production results, tolerance recommendations, or a measurement-system study.
Two parts are retained from each of two production windows. All are labeled cavity 2 and weighed as trimmed parts using the same recorded method. The dimension is an opening width. The record contains these values:
| Sample | Weight | Opening width | Age when measured |
|---|---|---|---|
| A-01 | 10.00 g | 20.02 mm | 24 hours |
| A-02 | 10.00 g | 20.01 mm | 24 hours |
| B-01 | 10.00 g | 19.86 mm | 12 minutes |
| B-02 | 10.00 g | 19.87 mm | 12 minutes |
The B samples have lower recorded widths. All four weights match at the reported precision. The dimensional measurements were made at different ages, and the table gives no conditioning or fixture details. There is no tolerance in this example, so it does not establish which samples pass or fail.
It would be premature to write “cooling-water change caused shrinkage,” even if someone remembers a water change that morning. The measurement history needs review, and the process event needs a timestamped record. It would also be premature to claim that waiting will make the B dimensions match A: this example supplies no evidence of the direction or size of later change.
A suitable handover entry is: “Lower opening-width readings in B; weight unchanged at recorded precision. Age at measurement differs. Quality to confirm the specified comparison conditions and sample suitability; process owner to retain the production-window history.” That tells the next person what is observed and what remains unresolved.
If a subsequent comparison at the specified conditions confirms the difference, retain that result alongside the original record. The next question becomes which process or material evidence could explain it. The matching weight values still do not establish dimensional equivalence.
After the measurement check, compare the production windows
Keep cavity identity through this stage. Compare the relevant material and process records for the affected samples, including stops or restarts. Record what changed, what was measured, and what was merely assumed to be stable.
Autodesk’s warpage-analysis documentation separates contributions from differential cooling, differential shrinkage, and orientation. This is a useful reminder that shape change can have several contributors; the software’s list does not diagnose the samples on your bench.
Distinguish a local size change from a change in shape using the specified dimensional checks. One opening-width measurement cannot describe the entire part. If a shape problem is suspected, ask quality or engineering which additional characteristics would distinguish the possibilities. Do not choose extra measurement locations after the fact solely because they support a favored explanation.
For a suspected thermal change, retain the actual records available for the job and their measurement locations. A controller setpoint is not a measured mold-surface temperature. If a water-supply condition was not recorded, write that limitation instead of supplying a remembered value.
The process owner can then define an approved trial with a specific response to measure. Keep its record connected to the troubleshooting log and carry open work into the shift handover. A useful stopping point for this review is a comparable set of measurements and a named next investigation, even when the root cause is still open.
Use the record alongside the quality system
The printable comparison sheet provides space for sample history, weight definition, the dimension’s method reference, observed differences, and the next owner. Attach the detailed individual results from the approved record system. The sheet supplies no acceptance thresholds.
If the open question concerns whether packing reached a thick feature, read the sink-mark evidence review. For operators developing technician skills, preparing a traceable comparison is a practical task to discuss with a supervisor using the skills map.
Sources and scope
RJG supports the distinction between weight and dimensional quality. Covestro supports the brief discussion of time-dependent dimensional change; material-specific behavior still needs the relevant supplier and job specifications. ASTM’s public statement supports conditioning comparability, and Autodesk identifies distinct warpage contributors in analysis. The sample table, record map, measurement-review questions, and worksheet are original MTF teaching material, not a validated inspection method or a verified factory case.