Front page / How much do injection moulded parts shrink?
How much do injection moulded parts shrink?
Most unfilled injection-moulded parts shrink by somewhere between about 0.4 % and 2.5 % as they cool, depending mainly on the resin, according to Komaspec's design guide.
That sounds small, but the mould cavity is cut oversize by exactly this amount, so the shrink value you and your toolmaker agree on is effectively built into the steel. This article compares published values, explains why they differ, and sets out what a first-time buyer should pin down before the tool is cut.
Published shrinkage by resin
Shrinkage figures vary between publishers, partly because they are measured or summarised differently. The table shows three published sets side by side.
| Resin | Komaspec design guide | Komaspec materials guide | Fictiv |
|---|---|---|---|
| ABS | 0.4 to 0.7 % | about 0.4 to 0.7 % | 0.7 to 1.6 % |
| Polycarbonate (PC) | 0.5 to 0.7 % | about 0.5 to 0.7 % | 0.7 to 1 % |
| PC/ABS | not in table | about 0.5 to 0.7 % | 0.5 to 0.7 % |
| Polypropylene (PP) | 1.0 to 2.5 % | about 1.5 to 2.5 % | 1 to 3 % (homopolymer), 2 to 3 % (copolymer) |
| Nylon, unfilled | 1.0 to 2.0 % (PA6/PA66) | about 0.8 to 2.0 % (PA6/PA66) | 0.7 to 3 % (nylon 6/6) |
| 30 % glass-filled nylon | 0.3 to 0.5 % | not in table | 0.2 to 0.5 % |
| Acetal (POM) | 1.8 to 2.5 % | not in table | 1.8 to 2.5 % |
| TPU / TPE | not in table | about 0.8 to 2.0 % | not in table |
The sources agree on the broad pattern but not on the detail. Fictiv's range for ABS starts higher than both of Komaspec's, and even Komaspec's two guides give different lower bounds for polypropylene and nylon. Fictiv describes its values as a general range and tells readers to use the material data sheet for best results. Treat every number in this table as a starting point for a conversation with your moulder, not a design value.
What a resin supplier publishes
Resin data sheets are the better reference, and they show how much context sits behind a single number. Covestro's data sheet for its Makrolon 3105 polycarbonate lists:
| Measurement | Covestro value |
|---|---|
| Shrinkage parallel to flow, 60 by 60 by 2 mm plaque at 500 bar (ISO 294-4) | 0.70 % |
| Shrinkage normal to flow, same plaque (ISO 294-4) | 0.75 % |
| Range from general practical experience | 0.60 to 0.80 % |
Covestro says these are reference values from plaques moulded under controlled conditions, that real parts vary with size, thickness, geometry, mould design and moulding conditions, and that choosing a shrink value is the customer's responsibility, based on its own testing. In other words, even the resin maker will not tell you the exact shrink of your part.
Amorphous versus semi-crystalline
The biggest single factor is the type of polymer.
- Amorphous resins (ABS, PC, PC/ABS, polystyrene, acrylic) have no ordered structure. Komaspec says they shrink 0.4 to 0.7 % and about equally in every direction.
- Semi-crystalline resins (PP, polyethylene, nylon, acetal, PBT) form ordered crystalline regions as they cool, which releases volume. Komaspec gives 1 to 2.5 % and says they often shrink unevenly, with more scatter from part to part.
- Glass-filled grades shrink less overall, but Komaspec notes that the fibres only restrain shrinkage along the flow. Its example is 30 % glass-filled nylon at roughly 0.3 % along the flow and around 1.0 % across it.
That difference between directions is what bows flat panels in glass-filled and semi-crystalline parts. Komaspec attributes most warpage in PP and nylon parts to it.
Why it matters to the design
Komaspec states that the shrinkage value is the number the toolmaker uses to size the cavity. Its worked example: a 0.5 % error on a 200 mm dimension means 1 mm of error on every part.
Shrinkage also explains three common first-project surprises:
- You usually cannot swap resins later. Komaspec says that once the tool is made, changing to a resin with a different shrink value normally means modifying or re-cutting the mould. Switching from ABS to PP is its example of a change that needs tool work; moving between two grades in the same family may not.
- Thick sections shrink more and later. Komaspec explains that thick areas cool slowly and keep shrinking after the surface has frozen, pulling the part out of shape or leaving sink marks. Uniform walls are the main defence.
- Process settings move the number. Fictiv lists injection pressure, hold time, material density and mould temperature as having a significant effect on shrinkage, which is why the same tool can produce slightly different parts on a different press or day.
Nylon adds a further twist: Komaspec notes that it absorbs moisture and grows after moulding, so its final size depends on conditioning as well as shrinkage.
A checklist before the tool is cut
- Choose the resin, and the exact grade, before tooling starts.
- Get the shrinkage value and test method from that grade's data sheet.
- Ask the toolmaker which shrink value they will cut to, and whether it differs along and across the flow.
- Keep walls as uniform as the design allows.
- Ask whether critical features will be left with spare steel so they can be adjusted after the first trial parts are measured.
- Measure first-article parts against the drawing before approving the tool.
These are typical published values only. The data sheet for your exact grade, and measured trial parts, decide what your part will actually do.
Questions buyers ask
Does the toolmaker or the designer allow for shrinkage?
Normally the designer draws the part at its finished size and the toolmaker scales the cavity by the agreed shrink value; Fictiv describes this as designing to nominal and scaling up the mould. Confirm in writing who is responsible and which value is used.
Can I switch from ABS to polypropylene with the same mould?
Komaspec says no, not without modifying the tool, because PP shrinks considerably more than ABS. Changing between grades of the same resin family may work, but test parts should be measured before accepting it.
Why do two sources give different shrinkage for the same resin?
Values depend on the grade, test plaque, thickness, flow direction and process settings, and some publishers summarise many grades into one range. Covestro's data sheet shows how specific the conditions behind a single figure are. Use the data sheet for your grade rather than a generic table.
Do glass-filled parts warp more?
They can. Komaspec notes glass-filled nylon shrinks roughly three times as much across the flow as along it, which can bow flat panels. Gate position and part geometry both affect the result, so raise it at the design review.
Sources
- Komaspec, Plastic part design for injection molding: https://www.komaspec.com/about-us/blog/plastic-part-design-for-injection-molding/
- Komaspec, Injection molding materials: https://www.komaspec.com/about-us/blog/injection-molding-materials/
- Fictiv, The Ultimate Guide to Injection Molding Tolerances: https://www.fictiv.com/articles/injection-molding-tolerances-an-in-depth-look
- Covestro, Makrolon 3105 product data: https://solutions.covestro.com/en/products/makrolon/makrolon-3105_56978201-00009648