Simulation Says Pass, First Articles Say Short Shots
A simulation predicted complete fill at 92% of the press's injection pressure limit for a glass-filled nylon bracket. First articles show short shots on the two farthest cavities of a four-cavity family tool. The mold has a hot runner with valve gates, and the two short cavities are fed by the longer branch of the runner. Measured melt temperature at the nozzle is 285 °C against a simulated 290 °C. The team suspects the simulation used a generic viscosity curve rather than the supplier's measured data. Cycle time is 22 s, and the customer needs 4,000 parts per week from this tool.
1. What is the most likely cause of the mismatch, and how should the team verify it before changing the tool?
Read the worked answer
The most likely cause is a combination of an optimistic viscosity model and runner imbalance, not a single error.
Evidence pointing that way:
- The short cavities sit on the longer runner branch, which is the expected signature of flow imbalance rather than a global pressure shortfall. If the whole tool were under pressure, all four cavities would show it.
- The 5 °C melt-temperature gap is small but matters for glass-filled nylon, whose viscosity is steeply temperature-sensitive. A generic curve can under-predict viscosity at the actual temperature.
- A generic curve also misses fiber-orientation effects on viscosity in thin sections.
Verification steps before touching steel:
- Run a short-shot study, stepping fill volume up in increments, and record the fill pattern. Compare the actual fill front sequence to the simulated one. If the sequence differs, the model's runner or thermal assumptions are wrong.
- Measure actual cavity pressure in one short and one full cavity if sensors exist, or use a pressure transducer at the sprue.
- Replace the generic viscosity curve with the supplier's capillary data and rerun at 285 °C.
- Check hot-runner tip temperatures on the long branch; a cold tip there mimics exactly this symptom.
Only after these checks should the team decide between runner rebalancing and process changes.
2. Assuming the model is corrected and still predicts marginal fill, what process changes would you try first, and what is the risk to the 4,000 parts per week target?
Read the worked answer
The target is 4,000 parts per week from a four-cavity tool at 22 s cycle, which is about 15,700 shots per week at full rate, so there is real capacity headroom if the tool runs reliably. That headroom should be protected.
Process changes to try first, in order of reversibility:
- Raise melt temperature modestly, 5 to 10 °C, to match the simulated condition. Risk: longer cooling, but cycle has margin.
- Raise injection velocity on the long branch by profiling, so the melt reaches the far cavities before it freezes. Risk: higher shear and potential jetting at the gate.
- Balance the hot-runner tip temperatures so the long branch runs slightly hotter.
- Increase pack pressure only after fill is achieved, since pack does not fix a short shot.
Capacity impact: any change that adds 1 to 2 s to cycle still leaves the weekly target achievable, because the tool has roughly 4x headroom over the required output. The real threat to the target is scrap from intermittent short shots, not cycle time.
Decision: pursue process window first, validate with a 500-shot run, and hold tool modification in reserve. Record the corrected model and the validated window so the fix is reproducible.