Scientific moulding replaces "it worked last time" with a documented, data-driven process. The studies behind it, from viscosity curves to gate freeze, and why it ends shift-to-shift variation and speeds up tool approvals.
On the day shift the part ran at 32 seconds with 0.5% scrap. On the night shift it ran at 35 seconds with 3% scrap. Each setter had their own settings, each certain theirs were right, and when a new setter joined the team, the process changed again. Nobody was wrong. The process had never been defined by data, so it was defined by people.
Scientific moulding, sometimes called decoupled or systematic moulding, replaces that with a process built on measured responses. The result is a robust window that any trained setter can run, and that behaves the same on every shift.
The core idea: separate fill from pack
In a traditional process, filling and packing blur together, controlled by a pressure limit and a switchover nobody has examined. Scientific moulding separates them:
- Fill the part to around 95–98% by volume at a controlled speed, using velocity control.
- Switch over by screw position, not by time or pressure.
- Pack and hold the remaining few percent with controlled pressure until the gate freezes.
Each stage can then be optimised independently.
The key studies
- Viscosity (rheology) study: run injection at a range of speeds and calculate relative viscosity. At low speeds viscosity changes a lot with small speed changes; at higher speeds it flattens out. Choosing an injection speed on the flat part of the curve means normal variation in material and temperature has little effect on fill.
- Cavity balance study: on multi-cavity tools, short-shot the parts and compare the fill of each cavity. Imbalance points to runner, gate or thermal problems that need fixing before anything else.
- Pressure drop study: measure pressure losses through the nozzle, runner and gates to make sure the process is not pressure-limited.
- Process window study: vary hold pressure and melt or mould temperature to find the range that gives good parts, and set the process in the middle of it.
- Gate freeze study: weigh parts at increasing hold times. When weight stops increasing, the gate has frozen; set hold time just beyond it. This directly addresses problems such as sink and dimensional variation.
- Cooling time optimisation: reduce cooling time until critical dimensions or warpage start to change, then add a margin.
What you get
- A process sheet that records not just settings but why they were chosen.
- Settings that transfer between machines, using machine-independent values such as melt pressure, flow rate, fill time and cushion, rather than one machine's screw speed.
- Stability from shift to shift and setter to setter.
- Faster troubleshooting: when something changes, you compare against known responses rather than guessing.
- Better evidence for PPAP and process validation.
Common objections
- "It takes too long." The studies take a day or two on a new tool. The alternative is weeks of trial and error spread across every production run for the life of the tool.
- "Our setters already know what they are doing." They often do, and scientific moulding gives them a method to prove it, document it and pass it on.
- "Our machines are too old." Older machines benefit most, because a robust window absorbs their variation.
Getting started
Start with the tool that causes most problems or most variation between shifts. Run the studies, document the window, train the setters and lock the process. The first one sets the pattern for every tool after it.
