Warpage is dimensional distortion that shows up after the part cools — a twist, a bow, a dished face, a corner that lifts. It frustrates people because the part often looks fine the moment it ejects and goes wrong afterward, in the cooling fixture or the shipping box. That timing is the clue to the whole problem: warp is a shrinkage story, and shrinkage keeps happening after the part leaves the mold.
The one idea that explains nearly all of it is differential shrink. Every molded part shrinks as it cools. If all of it shrank by the same amount at the same time, the part would come out smaller but true. It warps when one region shrinks more than another — because of uneven cooling, uneven wall thickness, fiber orientation, or locked-in stress. The region that shrinks more pulls the part toward itself, and the molding bends. Control warp by making shrinkage uniform, or by holding the part rigid until it’s set.
The drivers of differential shrink
Warp almost always traces to one or more of these, and they often stack:
| Driver | Why it warps the part | Direction of the fix |
|---|---|---|
| Uneven mold temperature | One mold half or one area hotter; that side cools slower and shrinks differently | Balance cooling so both halves and all areas run even |
| Non-uniform wall thickness | Thick sections shrink more than thin ones, pulling the part | Uniform walls; gradual transitions |
| Fiber/filler orientation | Fiber-filled resin shrinks less along the flow than across it — built-in directional shrink | Gate and orient flow with the warp in mind; expect anisotropic shrink |
| Ejecting too hot / too soon | Part isn’t rigid yet and distorts as it finishes cooling unsupported | Extend cooling time; eject when the part can hold its shape |
| Asymmetric cooling or runner layout | One side of the tool removes heat faster; flow and pack aren’t symmetric | Symmetric cooling circuits and balanced runners |
| Molded-in stress | High pack, fast fill, or cold flow freezes stress that later relaxes into warp | Reduce stress: tune pack, fill, and temperatures |
| Post-mold handling | Hot parts stacked or boxed creep and set into a warped shape | Cool flat/fixtured; don’t stack parts hot |
The reason warp is hard is that several of these usually act at once, and they can partly cancel or compound. That’s why chasing it with a single setting rarely works — you have to find which drivers are dominant for this part.
Cooling is the biggest lever
Because warp is differential shrink and shrink is driven by cooling, the cooling system is usually where the biggest gains are. A tool whose two halves run at different temperatures, or whose circuits aren’t balanced, bakes a shrink difference into every shot. Even, balanced cooling — both mold halves at their intended temperatures, every circuit actually flowing and turbulent, no blocked or starved lines — is what makes a part shrink uniformly. Uneven cooling is one of the most common hidden causes of warp, and it’s invisible from the press unless you measure flow and temperature circuit by circuit.
The companion lever is cooling time. A part ejected before it’s rigid finishes cooling unsupported, and it distorts as it does. Giving the part enough time in the tool to set up — so it can hold its own shape when the ejectors push — prevents a lot of warp that no amount of process tweaking will fix afterward. Ejecting hot to save cycle time is a false economy if it warps the part.
Wall thickness and material: shrink built into the design
Two warp drivers are decided before the press ever runs.
Wall thickness. Thick sections shrink more than thin ones. A part with uneven walls therefore shrinks unevenly by design, and bends toward the heavier sections. Uniform walls with gradual transitions are as important for warp as they are for sink.
Material and fillers. Shrink rate is a material property, and it varies widely — an unfilled polyolefin like PP can shrink on the order of one to a few percent, which is a lot of dimensional change to manage. Fillers reduce shrink: adding glass fiber can cut a resin’s shrink substantially (a heavily glass-filled grade may shrink only a fraction of the unfilled value). But fillers introduce a new problem — directional shrink. Glass fiber tends to align with the flow and shrinks less along its length than across it, so a fiber-filled part shrinks different amounts in different directions and can warp from that anisotropy alone. Choosing a material is partly choosing a shrink behavior, and it should be a conscious decision, not an afterthought. (Treat any specific shrink figures as illustrative; the datasheet is the source of truth.)
The part that warps in the box
Here’s the effect that surprises people and costs real scrap: a part can be flat when it ejects and warped hours later. Shrinkage continues well after the part leaves the mold, and whatever shape the part is held in while it finishes shrinking is the shape it sets into. Stack hot parts in a bin and their own weight presses them into a warp that becomes permanent. Box parts before they’ve stabilized and they cool against each other crooked.
The fixes are unglamorous and effective:
- Cool warp-prone parts flat or in a fixture until they’ve stabilized, so they set true instead of crooked.
- Don’t stack hot parts under their own weight; the load deforms them while they’re still soft enough to take a set.
- Don’t reduce turnaround so aggressively that parts are handled and stacked before they’ve finished shrinking. The minutes saved come back as warped parts.
A cooling fixture is cheap next to a chronic warp reject, and for some geometries it’s the only practical way to hold tolerance.
A practical sequence
- Characterize the warp — twist, bow, dish — and note where it’s worst.
- Check cooling first: confirm both halves and all circuits are even, flowing, and balanced. Uneven cooling is the usual top suspect.
- Confirm cooling time is long enough that the part ejects rigid, not soft.
- Look at the geometry and material: uneven walls and directional (fiber) shrink are designed-in causes.
- Check post-mold handling: are parts stacked or boxed hot? Fixture or cool them flat.
- Reduce molded-in stress (pack, fill speed, temperatures) once the bigger drivers are ruled out.
FAQs
What actually causes warpage?
Differential shrink — one region of the part shrinking more than another as it cools. Every molding shrinks; it warps only when that shrinkage is uneven. The common drivers are uneven mold temperature (one side cools slower), non-uniform wall thickness (thick sections shrink more), fiber orientation in filled resins (less shrink along the flow than across it), ejecting the part before it’s rigid, asymmetric cooling or runner layout, and molded-in stress that later relaxes. Usually several act together, which is why warp resists single-setting fixes.
Why is my part flat at the press but warped later?
Because shrinkage continues after the part leaves the mold, and whatever shape the part is held in while it finishes shrinking is the shape it keeps. A part that ejects flat but gets stacked hot will be pressed into a warp by its own weight, or cool crooked against neighbors in a box, and set that way permanently. The fix is to let warp-prone parts stabilize flat or in a cooling fixture before stacking or boxing them — and to not cut handling time so short that parts are packed before they’ve finished shrinking.
How does cooling affect warpage?
Strongly — it’s usually the biggest lever. Since warp is differential shrink and shrink is driven by how the part cools, uneven cooling bakes a shrink difference into every shot. If the two mold halves run at different temperatures, or some circuits are blocked, starved, or running laminar, one area cools and shrinks differently from another and the part bends. Even, balanced cooling — both halves at temperature, every circuit flowing and turbulent — makes the part shrink uniformly. Adequate cooling time so the part ejects rigid is the companion fix.
Does glass fiber reduce or cause warp?
Both, in different ways. Adding glass fiber substantially reduces a resin’s overall shrink, which helps dimensional stability. But the fibers align with the flow and shrink less along their length than across it, so a fiber-filled part shrinks different amounts in different directions — directional (anisotropic) shrink that can warp the part on its own. So a filled grade lowers total shrink but adds an orientation effect you have to design the gating and flow around. Choosing the material is choosing a shrink behavior, and it should be deliberate.