Three parts look complete. The fourth is missing the end of a rib. Before changing a setting, keep those four parts together: the useful comparison is already in your hands.
A short shot in one cavity is a pattern to investigate, not a diagnosis. The first job is to establish whether the same cavity repeatedly produces the same incomplete feature, whether the other cavities change with it, and whether the samples really came from the same cycles. This guide shows how to make that comparison without losing the evidence in a mixed parts bin.
Give the defect an address
“Top right” is ambiguous when one person looks at the mold and another looks at the ejected parts. Use the tool's documented cavity identification and preserve the viewing direction. If a part has no readable cavity mark, use the approved collection method to maintain identity; do not infer a cavity number from the order in a tote.
Record the incomplete feature as well as the cavity: for example, “C3, terminal rib opposite the gate, missing end segment.” Include an oriented photograph with a scale and a complete comparison part. The label should make sense to someone who was not standing at the press.
A visible cavity mark identifies the source of a part; it does not describe the flow path that fed it. Before comparing neighboring positions, obtain the actual cavity and runner map. Adjacent cavities in a tray are not necessarily neighbors in the feed system.
Read across the shot, then down the cavity
Start with a defined sample window. Keep every available cavity result from each sampled cycle, including incomplete parts. If a sample is absent or its identity is uncertain, say so. A missing sample is neither a complete part nor a short shot.
| Cycle | C1 | C2 | C3 | C4 |
|---|---|---|---|---|
| 101 | Complete | Complete | Incomplete | Complete |
| 102 | Complete | Complete | Incomplete | Complete |
| 103 | Complete | Complete | Incomplete | Complete |
| 104 | Complete | Complete | Sample missing | Complete |
| 105 | Complete | Complete | Incomplete | Complete |
Reading across each row compares cavities exposed to the same cycle. Reading down a column asks whether a cavity's observation repeats. Here, the rib end is incomplete in all four available C3 samples. It is incorrect to report five C3 failures, or a proved 100% long-term reject rate. The window is small and one observation is missing.
Now imagine a different record: C3 is incomplete in cycle 101, C1 in 102, and every cavity is incomplete in 103. That calls for a different investigation from a stable C3-only pattern. Preserve the sequence instead of summarizing both records as “short shots in a four-cavity mold.”
Link the window to the press, mold revision, material identity, approved process reference, and any interruption or recorded intervention. Compare it with a traceable earlier window when possible. A bag of good parts from an unknown shift is a weak baseline.
The total can stay the same while a cavity changes
Weighing all parts together answers a different question from weighing them by cavity. The following arithmetic exercise uses four nominally identical parts, with runners excluded and the same weighing method assumed. It does not specify an acceptable part weight.
| Sample set | C1 | C2 | C3 | C4 | Total |
|---|---|---|---|---|---|
| Earlier | 10.00 | 10.00 | 10.00 | 10.00 | 40.00 |
| Later | 10.20 | 10.20 | 9.40 | 10.20 | 40.00 |
The total difference is zero. C3 nevertheless differs by −0.60 g, offset by +0.20 g at each other cavity. Neither the total nor the per-cavity mass proves why the rib is incomplete. The exercise shows why an aggregate can hide redistribution. Record appearance, identity, and measurement results together; do not call the heavier parts acceptable simply because they are heavier.
For a family mold producing different components, equal mass across cavities is not the objective. Compare each cavity against the appropriate part reference. Also keep runner inclusion, trimming, conditioning, scale resolution, and sampling time consistent when those affect the comparison.
Use the pattern to narrow the question
Beaumont's discussion of collecting diagnostic short shots explains that controlled short-shot studies can reveal cavity filling differences, and that a poorly conducted study can mislead. Its guidance also preserves cavity and shot identity. A production reject collection is useful evidence, but it is not automatically a controlled filling study.
| Observed pattern | Next question | What remains unproved |
|---|---|---|
| Same feature in C3 repeatedly | Does the tool history or an approved inspection identify a local difference at this location? | A blocked vent, restricted gate, or tooling fault has not been established. |
| Several cavities change together | Which cycle records changed in the same window, and were there interruptions? | A shared change does not identify the responsible material or machine condition. |
| Cavity identity changes or disappears | Can the collection and labeling method be verified first? | Apparent movement of the defect may be a traceability error. |
| Pattern appears after an intervention | What exactly changed, and where is the first reliable before/after pair? | Timing alone does not prove the intervention caused the defect. |
A fixed cavity pattern justifies looking for cavity-specific evidence. It does not rule out a shared process condition making the least tolerant cavity fail first. Likewise, matching controller values do not prove that every cavity experienced identical conditions. These are reasons to preserve alternatives until the process or tooling owner can distinguish them.
The owner of a diagnostic study should define its method, sample handling, limits, and acceptance criteria. This article does not prescribe pressure changes, disabling cavities, or tool modifications. Those actions can change the comparison itself and require the relevant approved procedure.
Write the finding at the strength of the evidence
A useful sample summary
“C3 rib end incomplete in four identified samples from cycles 101–105; C3 sample 104 missing. Corresponding C1, C2, and C4 rib ends appear complete. Samples retained by cycle and cavity. No cause confirmed. Please review the cavity map and local tool history, then decide whether a controlled filling comparison is needed.”
If short shots coexist with flash, add the flash location rather than treating it as a separate success or failure. Continue with mapping short shot and flash together. For the next authorized investigation, use the trial log to connect changes with samples, or prepare a technical-help request when the evidence exceeds your role.