Warpage in Injection Moulding: Diagnosis, Trials and Acceptance

Updated: 2026-10-22

A warped injection-moulded part bends, cups or twists after release because deformation is uneven across its shape. Differential shrinkage, fibre orientation, cooling, residual stress and ejection can interact. A photograph identifies a symptom; it does not establish which setting or tool feature caused it. Diagnose the part in its specified condition before changing the process.

Establish a measurement baseline

Mark the drawing revision, resin grade and batch, cavity, shot sequence and time since moulding. Record drying and moisture condition where relevant. Use the same support points, orientation and measurement method for every comparison. A fixture that forces a flexible cover flat can conceal its free-state distortion; conversely, free-state flatness may not describe its assembled function. Agree which condition governs acceptance.

Measure at a defined interval after moulding and repeat after the agreed conditioning or environmental exposure. Moisture-sensitive grades and parts used hot can change shape after an early inspection. Record temperature and conditioning alongside the result. Do not compare a warm part measured beside the press with a conditioned part measured elsewhere as though process settings were the only difference.

Use the pattern to choose the next check

Observation	Investigate next	What it does not prove
One cavity differs consistently	Local cooling flow, gate restriction, cavity geometry and ejection	That the machine cannot contribute
All cavities change after a restart	Thermal stabilisation, material preparation and process repeatability	That the tool geometry is correct
Twist follows a reinforced material's filling pattern	Fibre orientation, gate strategy and grade-specific shrinkage	That every filled grade behaves identically
Shape changes mainly after ejection or storage	Release forces, residual stress, temperature and conditioning	That longer cooling alone will solve it

A repeatable shape is useful evidence, but it is not proof of a purely structural cause. A repeatable temperature imbalance or pressure distribution is also a process condition. Likewise, a weak response to one trial does not eliminate that mechanism: the change may be too small, the process may not be stable, or another mechanism may dominate.

Separate cooling, packing and release effects

For a cooling hypothesis, inspect actual coolant flow and temperature on the relevant circuits, rather than assuming equal controller setpoints produce equal steel temperatures. Look for restricted circuits, local hot regions and differences between core and cavity. A longer cooling trial can reveal sensitivity to ejection temperature, but cannot establish an optimum production cycle by itself.

For a packing hypothesis, check the filling-to-holding transition, pressure history, cushion repeatability and gate sealing behaviour. Lowering pressure can reduce some stresses while increasing shrinkage elsewhere; increasing it can improve feeding while creating flash or other dimensional changes. Compare the whole drawing and cosmetic requirements, not only the warped corner.

For a release hypothesis, examine whether the part sticks, distorts around ejectors or is handled while insufficiently supported. Check draft and ejection layout with the toolmaker. A fixture, annealing operation or post-forming step needs its own validated method and cost; it is not automatic evidence that the original moulding process is acceptable.

Plan a controlled trial and a release decision

Start from a documented stable baseline. Change one selected factor while holding the material, measurement method and other settings constant; allow the relevant temperatures to stabilise. Compare identified samples from every cavity. A planned designed experiment is appropriate when interactions must be separated; several unrecorded adjustments made together cannot explain which change helped.

An illustrative teaching example: a cover has a drawing limit of 0.50 mm free-state flatness. Cavity 1 measures 0.30 mm and cavity 2 measures 0.80 mm after the agreed conditioning. Their mean is 0.55 mm, but even a passing mean would not release cavity 2. Investigate its local conditions, retain the failed samples, and repeat both dimensional and assembly checks after correction. These invented figures are not Nordmould production results or a recommended tolerance.

When does the design or tool need changing?

Review wall transitions, ribs and asymmetric masses against the wall-thickness guide. Review gate placement for feeding and, in reinforced grades, orientation. A balanced-looking drawing does not guarantee balanced filling or cooling. Adding ribs may increase stiffness but also add mass and alter shrinkage; assess both effects.

Flow and warpage simulation can compare alternatives before a tool change. Its result depends on the material model, geometry, process and cooling assumptions, and should be checked against trial parts. It predicts behaviour; it does not certify an exact final distortion. If the usable process window cannot meet the agreed requirements, compare geometry, cooling, gating or material changes with the appropriate functional tests.

For review, provide CAD, the drawing datums and flatness requirement, per-cavity measurements, conditioning details, process history and photographs showing support points. Record the correction and acceptance evidence in the sample approval pack, or send the project for review.

Engineering background: KEYENCE on measuring moulded-part warpage and Asahi Kasei on defect prediction. Use the selected grade's processing and conditioning instructions for the actual project.