A part warps because one region of it shrank more than another. That single sentence contains the whole diagnosis. If every part of a moulding shrank equally in every direction, you would get a part that is uniformly smaller and perfectly straight. You get a bent one when shrinkage is unequal.
So the useful question is never “why is it warping” but “which region is shrinking more, and why.” Everything below is a way of answering that.
This article goes deeper on warpage specifically; for the wider troubleshooting picture see our guide to injection moulding defects and their solutions.
What shrinkage actually is
A polymer occupies more volume as a melt than as a solid. On cooling it contracts, and the amount depends on the polymer.
| Polymer | Typical mould shrinkage |
| Amorphous – PS, ABS, PC, PMMA | roughly 0.4 to 0.8% |
| Semi-crystalline – PP, PE, PA, POM | roughly 1.0 to 2.5% |
| Glass-filled grades | lower overall, and strongly directional |
Semi-crystalline polymers shrink far more, and that is why they warp more. As they cool through the crystallisation range, chains pack into ordered regions that occupy less volume than the amorphous melt. Anything that changes how much crystallinity develops – cooling rate, mould temperature, pressure – changes local shrinkage. A PP part cooled fast on one face and slowly on the other has genuinely different crystallinity across its thickness, and it will bow.
Amorphous polymers have no crystallisation step, shrink less and warp less. If a design is marginal and the material can change, this is the single biggest lever available.
The four causes, in the order they are usually to blame
### 1. Uneven cooling – the most common by a wide margin
If one half of the tool runs hotter than the other, the hot side stays molten longer, develops more crystallinity and shrinks more. The part curls towards the hot side.
Do not assume the cooling circuit is doing what the drawing says. Measure both mould halves with a surface probe at several points during a production run. In Indian plants running hard water without treatment, scale build-up in cooling channels is routine and progressive – a tool that ran straight two years ago may be warping now for no reason other than deposit in the channels.
Check for: blocked or scaled channels, a baffle or bubbler that has come loose, a chiller undersized for the current cycle, hoses connected in series when they should be parallel, and simple asymmetry – core and cavity fed at different temperatures because that is how the hoses reach.
Corner geometry is a known cooling trap. The outside of a corner has more surface area to lose heat through than the inside, so the inside cools slower and shrinks more, pulling the corner closed. Conformal or well-placed corner cooling addresses it; process settings largely cannot.
### 2. Non-uniform wall thickness
Thick sections shrink more than thin ones, because there is more material to contract and it cools more slowly.
The standard design rule is to hold wall thickness as uniform as possible, and where a change is unavoidable, to transition gradually rather than in a step. Ribs should sit at roughly 50 to 60% of the wall they attach to – a rib at full wall thickness both sinks and warps.
This is a design cause, which means process settings can only mitigate it. If a part warps identically across every machine, every cavity and every material lot, stop adjusting the process.
### 3. Orientation – molecular and fibre
Polymer molecules and fibres align with the flow direction. Shrinkage then becomes anisotropic: different along the flow than across it.
In unfilled semi-crystalline polymers this effect is moderate. In glass-filled grades it is severe, and it works in the opposite direction to intuition: fibres restrain shrinkage along their axis, so a glass-filled part shrinks much less along the flow than across it. A flat glass-filled plaque gated at one end will bow predictably, and the fix is gate position and flow path, not holding pressure.
Changing gate location changes the flow pattern and therefore the orientation pattern. On parts where orientation dominates, that is the fix.
### 4. Ejecting too hot
A part ejected before it has cooled enough to be rigid will deform under ejector load and continue to shrink unevenly outside the tool.
This is the cause to suspect first if warpage appeared after cycle time was reduced. It is also the one most often introduced deliberately, by a production team under output pressure, and then investigated as though it were a mystery.
Post-mould fixtures and cooling jigs are the standard countermeasure where cycle time cannot be extended – the part is held to shape while the last of the shrinkage happens.
Fixes in cost order
Work down this list. Do not start in the middle.
Free – process settings
– Increase cooling time, or lower ejection temperature – Equalise mould half temperatures – Reduce melt temperature, which reduces total shrinkage – Increase holding pressure and time – more material packed in means less volumetric shrinkage, though excessive packing creates its own residual stress – Adjust injection speed to change the orientation profile – Reduce or eliminate regrind if it is shifting flow behaviour
Cheap – fixturing and handling
– Cooling fixtures that hold the part to shape after ejection – Controlled stacking, so parts do not creep under their own weight while warm – Annealing, where the material and application permit it
Expensive – the mould
– Rebalance or descale the cooling circuit – Add cooling to corners and thick regions – Move or add gates to change the flow and orientation pattern – Modify the ejection system to spread load
Most expensive – design and material
– Uniform wall thickness, correct rib proportions, added stiffening geometry – Move from a semi-crystalline to an amorphous polymer, or to a lower-shrinkage grade – Change filler loading
Where colour does and does not come into it
Being precise about this saves a lot of misdirected argument.
Masterbatch is not a normal cause of warpage. Pigment loading in a typical coloured part is a fraction of a percent of the total, far too little to change bulk shrinkage.
There are two genuine exceptions, and both are worth knowing:
Nucleating pigments. Some pigments – certain organic pigment classes in particular, and some blues and greens – act as nucleating agents in semi-crystalline polymers. They give crystallisation more sites to start from, which changes crystallisation rate and final crystallinity, which changes shrinkage. This is real and measurable in PP, and it is why the same tool can produce a slightly different dimension in one colour than another. It is not a fault; it is chemistry. Where dimensions are critical, qualify each colour rather than assuming the natural approval covers them all.
Very high dosage. At the loadings used in pipe black, high-opacity white or heavily filled systems, the masterbatch carrier is a real fraction of the compound and can shift both flow and shrinkage. That is a formulation matter to raise with your supplier, and a reason the carrier polymer belongs on your masterbatch specification.
The test that settles it, again: run the tool in natural resin at the same settings. If it still warps, the colour is not the cause.
Measuring warpage properly
“It’s warped” is not data. Before changing anything, establish a baseline you can compare against.
– Measure against a defined datum scheme, the same one every time – Measure at a consistent interval after moulding – parts continue to shrink for hours, and semi-crystalline materials for considerably longer – Record mould temperature on both halves, melt temperature, cycle time and cavity number with every measurement – Measure several cavities, because a single warping cavity is a tool problem and all cavities warping is a process or design problem
That last distinction alone will save you days.
Talk to us about the colour side
We supply colour, white and black masterbatch to injection moulders across India, manufactured under ISO 9001:2015. Where a part is dimensionally critical we will formulate with nucleation behaviour in mind and keep the carrier consistent between lots, so the variable you are chasing is not ours.
Send us your polymer, part geometry and colour references and we will tell you what to expect. Related: household, automotive, furniture.
FAQs
What causes warpage in injection moulding? Differential shrinkage – one region of the part shrinking more than another. The four underlying causes are uneven mould cooling, non-uniform wall thickness, molecular or fibre orientation, and ejecting the part before it is rigid enough to hold its shape.
How do you fix warpage in injection moulding? Work in cost order. Start with free process changes – longer cooling, equalised mould half temperatures, lower melt temperature, adjusted holding pressure. Then cooling fixtures. Then mould changes such as rebalanced cooling and gate relocation. Design and material changes last.
Why do glass-filled parts warp more? Because glass fibres align with the flow and restrain shrinkage along their axis. The part then shrinks much less along the flow direction than across it, and that anisotropy bends it. Gate position and flow path are the effective fixes, not holding pressure.
Does mould temperature difference between halves cause warpage? Yes, and it is the single most common cause. The hotter half produces more crystallinity and more shrinkage, so the part curls towards it. Measure both halves during production rather than trusting the cooling drawing – scaled channels are very common.
Can masterbatch cause a part to warp? Rarely, but there are two real exceptions. Some pigments nucleate crystallisation in semi-crystalline polymers and slightly change shrinkage, which is why the same tool can give different dimensions in different colours. And at very high dosage the carrier becomes a real fraction of the compound.
Why did my parts start warping after we reduced cycle time? Almost certainly because they are now being ejected too hot. A part that is not rigid at ejection deforms under ejector load and continues shrinking unevenly outside the tool. Either restore cooling time or hold the parts in a cooling fixture.
Do amorphous plastics warp less than semi-crystalline ones? Generally yes. Amorphous polymers such as ABS, PS and PC shrink around 0.4 to 0.8 percent with no crystallisation step, while semi-crystalline polymers such as PP and PE shrink 1 to 2.5 percent and are sensitive to cooling rate. Where a design is marginal, material choice is the biggest single lever.
How long after moulding should I measure a part? At a consistent interval, and long enough for post-mould shrinkage to have largely completed. Semi-crystalline materials continue to shrink for many hours. What matters most is that every measurement uses the same interval, so results are comparable.