Uniform wall thickness

An abrupt jump from a 1.5 mm wall to a 5 mm boss looks harmless in CAD. On the printer it is one of the most reliable ways to introduce warping, sink marks and an internal stress concentration that shows up as a crack weeks after the part is in service. Uniform wall thickness is one of the oldest rules in plastics part design, and it carries over to FDM, but the physical reason it matters is different from injection moulding.

Why abrupt changes cause warping

In FDM, each layer solidifies as it cools, and thicker sections hold heat longer than thin ones next to them. When a thick boss sits beside a thin wall, the boss is still contracting as the wall has already gone rigid, and the mismatch pulls the thin section out of plane. This is the same mechanism behind warping at large flat bases, just concentrated at a local feature instead of spread across the whole part.

The same thermal mismatch also produces sink marks, visible as a shallow dimple on the surface opposite a thick internal feature such as a rib or boss. On an injection moulded part the cause is packing pressure and gate location, on an FDM part it is uneven cooling and the different shrinkage rate of a thick zone against the surrounding shell.

Internal stress and inconsistent cooling

Beyond visible defects, uneven wall thickness locks in internal stress that reduces long-term durability. A part can look dimensionally fine off the printer and still fail early under load or thermal cycling because the thick to thin junction acted like a pre-existing crack initiation site. This is worse with materials that shrink more (ABS, ASA, PA) and less pronounced but still present with PETG or PLA.

Blending transitions: tapers and fillets

The fix is the same one used in mould design: never step directly from one thickness to another. Blend the transition over a run of at least three times the wall thickness difference, using either a linear taper or, better for stress, a fillet radius at the inside corner. A jump from 1.5 mm to 4 mm should be spread over roughly 7 to 8 mm of transition length rather than happen at a single edge.

• Use a taper of 1:3 or gentler between two wall thicknesses
• Add a fillet radius at any internal corner where thickness changes
• Avoid placing a thickness change directly at a screw boss or snap arm root
• Check the transition in slicer preview, not only in CAD, for actual layer consistency

Ribbing instead of thickening

When a section needs more stiffness, the instinct is to thicken the whole wall. A rib achieves the same or better stiffness with a fraction of the added mass and, critically, with a smaller thermal mismatch against the surrounding shell than a solid thick zone would create. Keep rib thickness at 50 to 70 percent of the wall it attaches to, so the rib cools at a similar rate and does not itself become a new sink mark source on the opposite face.

What carries over from mould design, and what does not

Designers with an injection moulding background already know the uniform wall thickness rule, and that habit transfers well to FDM. What does not transfer directly is the underlying justification. In moulding the concern is filling balance and packing pressure through the gate. In FDM there is no gate and no melt flow to balance, the concern is purely thermal: how evenly a section cools relative to its neighbours. That means some mould-driven habits, like adding draft everywhere, matter less here, while others, like avoiding isolated thick masses, matter just as much or more.