18K Mixed-Alloy Casting Porosity at Thick-Thin Junctions: Why Vacuum Time Alone Is Not Enough
Quick Answer
A practical shift-level workflow for reducing thick-thin junction porosity in mixed 18K yellow/rose casting by coordinating thermal windows, feed-path geometry, and vacuum-pressure timing.
The short version: stabilize gas control, feed-path balance, and pressure timing first, then tune machine parameters. In most workshops, this order reduces repeated defects faster than parameter-only tuning.
What Usually Goes Wrong
- Operators change multiple variables in one shift, making root cause hard to isolate.
- Startup lots are mixed with normal lots, so process drift is hidden by average numbers.
- Thermal window and pressure timing are treated as independent, while they are tightly coupled.
One-Shift Recovery Workflow
- Lock input: freeze alloy/input ratio and tree family for the shift.
- Lock timing: keep burnout-to-pour transfer window consistent for test lots.
- Lock pressure profile: use one validated vacuum/pressure timing profile for the same geometry group.
- Record outcomes: mark each lot with defect type, not just pass/fail.
- Scale after proof: expand only after two consecutive stable lots.
Operational Guardrails
| Checkpoint | Target | Why it matters |
|---|---|---|
| Lot isolation | One product family per validation run | Prevents mixed signals |
| Shift handoff | Single log template | Makes defect recurrence traceable |
| Review rhythm | End-of-shift closure | Stops the same defect from rolling into next shift |
Practical FAQ
Should we replace hardware immediately?
Usually no. For this class of defects, process-window control resolves most recurrence before hardware change becomes necessary.
How many lots are needed before decision?
Use at least two stable consecutive lots under identical settings before promoting a new setting to standard.
What KPI should we track first?
Track first-pass yield and rework rate together. Looking at one KPI alone often creates false confidence.