How to predict warpage before cutting steel — and get a good part on the first trial shot.
The misread common sense
Warpage in injection molded parts has always been a headache — especially for engineering components with tight dimensional and GD&T requirements. And glass-fiber-reinforced parts tend to warp even more.
Many people, including me early in my career, hold a misconception: adding glass fiber makes a part fuller and flatter. In fact, it's the opposite — glass fiber makes warpage worse.
The root cause: fiber orientation effect
Why does glass fiber make warpage worse? The deepest reason is anisotropic shrinkage. Glass fibers align along the flow direction during filling. The result: shrinkage perpendicular to flow is greater than shrinkage parallel to flow. This is the fiber orientation effect.
Injection warpage also relates to injection pressure, cooling, and corner effects. But for glass-filled materials, fiber orientation is the dominant factor.
A real lesson
Early in my career, I had a spacer part. The customer required good flatness on the bottom bearing surface, and the material was not yet fixed. At the time we bought molded parts from an outside supplier, who suggested PA6+15%GF — with glass fiber, the part comes out full and won't warp. So we recommended PA6+15%GF to the customer.
The mold was built and trialed — and the sample's bottom face warped badly. The supplier was puzzled too. His fix: hold the part in a fixture after ejection. It didn't work — the part sprang back once removed from the fixture, and it cost labor.
When the project stalled, I dug through injection molding books and searched online — no AI back then. The conclusion was exactly the point above: glass fiber makes parts warp more, not less.
Option 1: change the material — when you can
We switched to unreinforced PA6, and the part came out flat. The customer tested and accepted pure PA6.
But what if unreinforced PA6 can't meet the mechanical requirements? If the customer needs the strength and stiffness, you can't drop the glass. Then what?
Option 2: pre-deformation — the key method
My answer is pre-deformation: at the product and mold design stage, compensate the predicted warpage in the opposite direction, so the part shrinks back to the print dimension after cooling. In practice, this works very well.
Timing is everything. Many people do pre-deformation only after the first trial shot — by then you've already burned time and cost, the customer and mold shop both suffer, and delivery slips.
The right way is to do it at the mold design stage: use Moldflow to predict the warpage trend and magnitude — and if the part demands very tight dimensional accuracy, leave a bit of steel stock on the mold for fine-tuning after the trial shot.
Is mold flow analysis accurate?
Flow analysis isn't accurate, it's only useful for weld lines, fill pressure, and air traps. Many people think this, which is why mold flow reports end up as a formality.
In my own hands, warpage analysis in Moldflow is quite accurate. A few conditions make it accurate:
- Optimize the mesh — thick-walled parts need a 3D mesh;
- Model the runner, gate, and cooling channels exactly as your real mold design, with correct parameters;
- Choose the right material for the flow analysis — most important — use the exact grade you'll run in mass production, and make sure it has measured shrinkage data;
- Set the process parameters, and run the trial shot with the same process settings as the analysis.
Hit these four, and your molded part will closely match the simulation. If you pre-deform the mold drawing based on the analysis data, you essentially get a good part on the first trial shot.
Case study: a PPO+20%GF U-shaped open part
Here's a real case, material PPO+20%GF — SABIC Noryl GFN2F. When I got the drawing, I noticed several critical dimensions in the opening direction. With a U-shaped open part, experience says the opening will shrink smaller. So I decided to pre-deform it right away, using Moldflow to simulate the deformation and then applying pre-deformation to the mold drawing.
Fig. 1 — Case part structure, de-identified — PPO+20%GF
Fig. 2 — 3D mesh with gating and cooling systems built per the mold design
Fig. 3 — Customer-specified material grade, with measured shrinkage data
Fig. 4 — Warpage analysis result — Y direction
Fig. 5 — Opposite pre-deformation applied on the mold drawing
The result: the part passed on the first trial shot.
Next up
Next article: high-temperature engineering plastic PEI — properties, applications, injection molding challenges, and mold design and machining notes.