A thin rib is still incomplete, but flash has appeared along the parting line. Another increase in pressure now looks like a trade: fill one feature, make another defect worse.
Short shot and flash can coexist because different regions of a mold do not necessarily fill or resist pressure in the same way. Before choosing a setting, establish what “at the same time” means. Are both defects on one part, on different cavities from one shot, or in samples collected before and after an adjustment? Those are three different records.
This guide helps operators and developing technicians prepare that evidence for the process owner. It does not provide a press recipe. Use the approved trial procedure for process changes and the applicable procedure for tool inspection.
Start with the parts, not the settings screen
Put the defect locations on an approved part drawing or permitted photograph. Identify the incomplete feature precisely: the tip of a rib, a corner, or an entire section. Mark the flash separately, including which edge or tool interface it follows. “Cavity 2 is bad” loses information that the next person needs.
Keep the sample ID, cavity ID, and collection time together. Where your sampling system permits, preserve the relationship between parts from the same shot. If that relationship was lost, write “shot pairing unknown.” Parts from one container are not automatically evidence of one molding cycle.
- Same individual partMap the incomplete feature and the flash location on that sample.
- Different cavities, same shotKeep each cavity’s result separate and retain the shot pairing.
- Different trial groupsReconstruct the change history before calling the defects simultaneous.
Agree on what counts as each defect. Use the drawing, visual standard, or inspection instruction for the job. A missing feature and a damaged feature after ejection may need different investigations. If the classification itself is uncertain, preserve the sample for review rather than building a trial around an unconfirmed label.
Why the two symptoms are not a contradiction
A cavity can have an easy flow path and a difficult one. In a multicavity tool, one cavity may fill ahead of another. Autodesk’s unbalanced-flow guidance identifies flash and short shots among the possible consequences when some paths fill before others. That makes imbalance a candidate explanation, not a diagnosis from appearance alone.
A thin branch can also hesitate while material advances through a less resistant section. If that stalled front freezes, the branch may remain incomplete. Autodesk describes this mechanism in its explanation of hesitation. A photograph of the final part does not reveal the sequence of filling; the process owner may need a controlled fill study or other evidence.
Flash has a separate physical requirement: material gets into a gap where it does not belong. BASF’s troubleshooting brochure, “Flash,” printed page 28, includes sealing-surface problems and excessive cavity pressure among its possible causes. A persistent local flash line therefore deserves a tooling record as well as a process record.
The useful question is whether one mechanism plausibly links both observations or whether two problems overlap. Do not force every defect into a single explanation just because it appeared in the same production window.
Transfer timing is worth recording precisely
Velocity-to-pressure transfer is the change from the filling control stage to the packing/holding control stage. Autodesk’s V/P switch-over documentation explains that switching too early can contribute to incomplete fill, while switching too late can create excessive pressure and flash. Its context is molding simulation; it is not an instruction for your machine.
Record the transfer method, full control label, units, set value, and actual event information available from the press. “Moved transfer back” is ambiguous. Position displays and local terminology vary; a numerically larger value should not be translated into “later” without checking the equipment’s meaning. Retain the original record so a qualified person can interpret it.
Collect evidence that can separate the explanations
Do not start by exporting every screen. Start with the question the record needs to answer, then attach the relevant controlled records. The table below is a proposed review sequence, not a list of changes to make.
| Question | Useful evidence | Limit of that evidence |
|---|---|---|
| Is the pattern local or mold-wide? | Individual cavity counts; feature maps; same-shot sample IDs where available. | A combined scrap total cannot show which cavities or features changed. |
| Did the trial affect filling or transfer? | Before/after process records; actual fill and transfer results available on the press; timestamps. | A changed setpoint does not establish the local conditions at the incomplete feature. |
| Is the short area repeatable? | Matched photographs or retained samples, with a consistent view and inspection definition. | Repeated location narrows the investigation but does not prove trapped air or hesitation. |
| Does flash follow one interface? | Location marked on samples; previous accepted samples; tooling inspection history. | The flash outline alone cannot distinguish tool condition from loading during the cycle. |
| Was the comparison disturbed? | Material/lot records, stops, restarts, alarms, and other changes during collection. | A before/after difference with several changed conditions cannot isolate one cause. |
Short shots have other possible explanations too. Autodesk’s short-shot troubleshooting page lists flow restrictions, inadequate venting, and insufficient material delivery among them. Keep the observed pattern separate from these possibilities. “Incomplete corner; vent condition not checked” is a better entry than “blocked vent” when nobody has verified the vent.
If machine records are unavailable, the sample map is still useful. Say which signals were not recorded. Do not relabel machine-side pressure as cavity pressure, and do not invent a curve from the final part. A tooling inspection should be assigned to the responsible person; identifying an area on a drawing does not authorize access to the mold or a tooling modification.
A lower short-shot count can hide a worse result
Constructed teaching example. These sample IDs and counts are invented. They are not plant results, a recommended sample size, or evidence that a particular adjustment works.
A two-cavity job has an incomplete rib in cavity 2. An authorized trial is recorded as T-24, with its control values held in the plant’s process record. Inspection uses the same criteria in both groups. Twenty parts per cavity are checked in each group.
| Group / cavity | Short shot | Flash | Both on one part |
|---|---|---|---|
| Before / 1 | 0/20 | 0/20 | 0/20 |
| Before / 2 | 8/20 | 0/20 | 0/20 |
| Trial / 1 | 0/20 | 7/20 | 0/20 |
| Trial / 2 | 2/20 | 3/20 | 1/20 |
In the trial group, cavity 2 has four distinct parts with at least one of these defects: two short, plus three flashed, minus the one counted in both. Across both cavities, 11 of 40 trial parts have one or both recorded defects, compared with 8 of 40 before. Other acceptance characteristics have not been evaluated in this example.
A note saying “short shots reduced from eight to two” is numerically correct and operationally incomplete. It hides the new flash and the change in cavity 1. The defensible note is: “Fewer incomplete ribs in the cavity-2 sample; flash appeared in both cavities; no acceptable process established.” The observed counts do not establish long-run defect rates or statistical significance.
The missing link is still the mechanism. These counts alone do not tell you whether transfer, flow balance, tool sealing, or another factor produced the result. Attach the T-24 process history and keep the mapped samples. The next trial should answer a narrower question chosen by the process owner, not simply repeat the direction of the last adjustment.
Leave a decision someone else can continue
Use the paired-defect worksheet to preserve the distinction between sample identity, observed pattern, proposed explanation, and missing evidence. Attach it to the main trial log rather than creating an unrelated process history.
Before handover, record the current authorized process state, sample locations, product status, and next owner. If flash suggests a tooling concern, identify the inspection request and restriction that actually applies. If the main gap is missing cavity identity, better sampling may be the next useful action. A record that prevents the next shift from repeating an ambiguous trial has already done useful work.
For a different apparent contradiction, see sink marks that persist after a hold-pressure increase. If visual improvement comes with a dimensional change, use the weight and dimension comparison guide.
Sources and scope
The linked Autodesk documentation supports the specific flow, hesitation, short-shot, and transfer mechanisms discussed above; simulation guidance does not establish a shop-floor diagnosis. BASF’s supplier brochure supports the brief discussion of flash. The evidence map, comparison table, arithmetic, and worksheet are original MTF teaching material. No supplier has endorsed them, and no forum account is presented as a verified factory case.