Why Does the Same Mould Fail on a Different Machine?

Home News

Why Does the Same Mould Fail on a Different Machine?

Why Does the Same Mould Fail on a Different Machine?

2026.11.18

Why Does the Same Mould Fail on a Different Machine?

A familiar scene: a mould runs steadily on the original machine — dimensions, appearance and cycle time all on target — then, once it moves to another machine, perhaps a different plant, a different workshop, or a machine the customer specified, problems start. Dimensions drift, sink marks and flash take turns appearing, and no setting seems to recover the original rhythm.

The first reaction is usually one of two: either the mould was badly made, or the setter on the new machine is not skilled enough. Both conclusions come too quickly.

Here is the thing: what a mould delivers was never "a set of parameters" — it is a process window. That comfortable, repeatable state you found on the original machine is the combined result of machine, mould, material and surroundings; it is not a property the mould carries inside itself. Change the machine and you have swapped out much of what held that window together, so of course the window moves. Calling that "the mould is broken" points you in the wrong direction.

So where exactly do machines differ? Not in brand or age, but in a few very concrete things.

One is temperature control. The way the temperature controller is plumbed in, the layout of the water circuits, and how stable the oil or water temperature is all shape the temperature field across the cavity wall. A mould's thermal balance is designed around a certain set of cooling conditions; change the machine and the stability of the water temperature, or how freely the circuits run, changes with it, and the thermal balance breaks down — thin sections freeze first, thick sections last, and dimensions and sink marks follow. That is why the same mould behaves differently the moment you re-plumb the water lines.

Another is plasticising. Screw design, barrel condition and back pressure decide how uniform the melt actually is by the time it reaches the nozzle. It may be the same batch of material, but put it through different plasticising conditions and its viscosity and temperature consistency differ, so it behaves differently once it enters the cavity. A machine that plasticises steadily gives you a wider window; one that fluctuates on its own will drag down even a good mould.

Then there are the "hard" conditions: clamping rigidity, platen parallelism, and the repeatability of injection response. None of these appear on a parameter sheet, yet all of them bear directly on flash, short shots and dimensional stability. A machine that has seen years of service can develop worn tie bars or a distorted platen — changes the eye rarely catches, but enough to turn a once-stable mould into one that is fussy about which machine it runs on.

Once you see that, it is clear why "copying the parameters across" barely works. Parameters are an outcome, not a cause. That neat injection-speed and holding-pressure profile you had on the original machine was realised by that machine's own hardware; record it unchanged into a new one and the melt behaviour it produces is already a different story. It is using someone else's key on your own lock. What should be transferred is the set of criteria for judging "what counts as acceptable", not the numbers themselves.

So a machine change should be handled as a small re-validation, not "a couple of trial shots". Fix the critical dimensions and measurement points first, and check whether their response on the new machine runs in the same direction as before. Use mould flow analysis to anticipate which areas are sensitive to process variation and watch those first. Then work through the problem list one item at a time — temperature control and plasticising first, then injection and holding pressure, and only then the structure. Reverse that order and you land in a loop where fixing one dimension breaks another. Plant transfers, customer-specified machines and bringing a backup machine online should all follow this route.

At the end of the day, the question is not "is this mould any good" but "how wide is this mould's window". A wide window shrugs off machine differences and stays stable wherever it goes; a narrow one is fussy, propped up by the state of one particular machine — and that kind of stability is not real. The width of the window is largely settled at the design stage: how even the thermal balance is, how small the cavity-to-cavity variation is kept, whether the gates and venting make sense. These decide whether a mould can absorb the gap between machines. Trying to compensate with adjustment once you are in production treats the symptom, not the cause. Specific conditions are subject to the agreed technical agreement.

One line: a mould that tolerates machine differences is the one that is genuinely stable; a yield propped up by a single machine does not count once you move it.

At Jingmu, we write the sensitive dimensions and measurement points into the record before the mould leaves, and on a machine change or plant transfer we re-validate against the same criteria, so the window is handed over clearly.