Uniform injection molding wall thickness: the one rule that matters most

Injection molding is a pressure-driven process. Melt takes the path of least resistance. That means thick sections fill first while thin areas hesitate. The imbalance shows up as weld lines, air traps, or short shots.

Cooling punishes variation even harder. Plastics conduct heat poorly. A thick section stays molten long after thin walls freeze, then shrinks more, builds internal stress, and warps the part after ejection.

Erik Foltz of Plastics Technology ranks the nominal wall decision among the first an engineer makes. It drives fill pressure and cooling time, and it decides how much press the job needs. Get it right and everything downstream gets easier.

Uniform does not mean identical everywhere. It means every wall stays within a tight band around the nominal value, and any change between sections is gradual. Section three covers the exact limits.

Recommended wall thickness by material

Each resin has a workable injection molding wall thickness window. Below it, the melt freezes before the cavity fills. Above it, cooling time and sink risk climb fast. Use this table as a starting point, then confirm against flow length and gate layout.

For most thermoplastics, 1.5 – 3.0 mm is the safe middle ground. A published molder guideline puts the absolute floor at 1.0 mm for common resins. High-flow PP or nylon can reach 0.6 mm on short flow lengths with good gating. Stay under about 5 mm unless the design really needs the mass.

Thin is not automatically better. Below the window you trade material savings for pressure-limited fills and scrap. Our DFM rules checklist covers wall thickness alongside the other early decisions that shape tooling cost.

How much variation your part can tolerate

Sometimes the design needs two thicknesses. A mounting pad must be beefier than a cosmetic shell. That is fine within limits. Keep variation under 25% of nominal for amorphous resins like ABS and PC. For semi-crystalline resins like PP and nylon, hold it under 15%. They shrink more.

Where thickness must change, taper it. A 3:1 transition gives 3 mm of length for every 1 mm of thickness change. That keeps the melt front stable and avoids a stress riser at the step.

The reason is cooling math. Cooling time grows with the square of wall thickness. Plastics Technology published a clear example. A 2 mm wall cools in about 1.76 seconds, while a 4 mm wall needs roughly 7 seconds. Put those two walls side by side and the thick one shrinks far more, which is exactly how warpage starts.

Sudden steps also trap air. The melt races through the thick side, hesitates at the thin edge, and folds air into the corner. A gradual taper costs nothing in CAD and saves a defect in production.

Thin walls: where filling fails first

Thin walls save material and cycle time, but they demand pressure. As a rule of thumb, filling should use 50 – 75% of the injection unit’s pressure capacity. That leaves headroom for normal variation. Most conventional machines generate 20,000 to 30,000 psi.

That is why injection molding wall thickness minimums come from the press, not the part. If filling needs more than 75% of the press’s pressure capacity, the wall is too thin for that flow length. Push past that and the process turns brittle.

Thin-wall work routinely needs fill pressures above 25,000 psi with fill times under a quarter second. Cut thickness 25% and you may need to halve the fill time to beat freeze-off. The payoff is real when the design supports it. Each millimeter shaved off a typical 2–4 mm wall can cut cycle time by roughly 15–25%.

But if your simulation shows pressure or tonnage exceeding the press, act early. Change the wall, shorten the flow length, add a gate, or pick a freer-flowing resin. One published case trimmed a container wall 10%, which raised injection pressure 21% and doubled the predicted tonnage.

Gate placement matters more on thin parts. Gate into the thickest section so thin areas pack from thick, not the reverse. Details sit in our mold design guide.

Thick sections hide inside ribs and bosses

The nominal wall can be perfect while the part still sinks. The culprit is usually a feature, a rib or boss thicker than the wall it stands on. Thick spots cool slowly, shrink more, and pull the opposite surface into a dimple. That is a sink mark.

That is why injection molding wall thickness discipline extends to every rib and boss. Size features to the wall, not to the load alone. Ribs should run 0.5 – 0.6 times the wall thickness at the base. Keep them no taller than 3 times the wall, and space them at least 2 wall thicknesses apart. Bosses follow the same logic at 0.6 times the wall, cored from below. I’ve seen more sink complaints from oversized bosses than from any other single feature.

If one rib is not stiff enough, add parallel ribs instead of thickening it. Three thin ribs beat one fat rib on stiffness, and they cool evenly. Gussets follow the same ratio as ribs.

Sink shows worst on glossy, dark surfaces. Texture and lighter color hide it, but that is camouflage, not a fix. Fictiv’s sink mark troubleshooting guide documents the same rib and boss ratios from the molder’s side. Our rib design notes go deeper on placement.

Check the design before anyone cuts steel

Simulation is cheaper than steel. Moldflow-style analysis weighs resin viscosity, wall thickness, flow length, and gate layout. It then predicts fill pressure and clamp tonnage. If the numbers exceed the intended press, adjust the design now, when CAD changes cost minutes.

We run this check on every RFQ with thin walls or tight tolerances. It catches the classic traps, like a long flow path at 1.2 mm or a thick boss on a cosmetic face.

One more rule for the road. Gate into the thickest region. Thick sections pack better, and packing from thick to thin gives the holding phase something to work with. Gate into thin and the thick areas starve.

Uniform cooling deserves the same attention as uniform walls. Our cooling channel guide shows how to aim water at the thick spots that drive cycle time.

Frequently asked questions