Bed adhesion and warping

The most common way a print fails is also the earliest: it lets go of the bed. Sometimes it never grips at all and the nozzle drags a growing tangle around the plate; sometimes it holds for an hour, then a corner curls up and the part shifts under the nozzle. Both are adhesion failures, but they have different causes, and the fix for one won't touch the other. This article reads the three signatures you'll actually see — a part that won't stick, a part that warps, and a base that came out swollen — and tells you which lever moves each. The full mechanism behind the first layer lives in The first layer and bed adhesion; here we go symptom by symptom.

The part never stuck

If the part came loose early, or the first layer looks like loose round threads that barely touch, the bead never keyed into the bed. In FDM the first layer has no plastic underneath it — it prints onto glass or coated steel, a surface molten plastic won't grip on its own — so the printer squashes that first bead flat against the plate to make it stick. Too little squish and it stays round and unstuck.

The machine fixes come first because this is usually a setup problem: lower the z-offset until the beads come out flat and just touching, level the bed so the squish is even everywhere, and check that the part-cooling fan is off on the first layer — cold air chills the bead before it can weld. But the design side matters too. Orient the part so its widest flat face sits on the bed; a part balanced on a small footprint has little to grip with. And where the footprint is genuinely small or the part is tall and narrow, a brim — a flat skirt of perimeters fused to the base — buys the gripping area the geometry didn't.

The corner curled up — warping

If the part stuck at first and then a corner peeled and lifted, that's warping, and it's a different enemy. As plastic cools it contracts, and that pull concentrates at the corners of the footprint, curling them up and off the bed. Whether they win is a contest between the shrinking force and the first-layer grip. Warping scales with material and with size: PLA barely warps, PETG and especially ABS pull hard, and a big flat footprint accumulates far more contraction than a compact one.

The design levers are real here. Round the corners of the footprint: a sharp corner cools unevenly and concentrates the shrinkage at its tip, while a generous radius spreads it out and stays down — a base fillet is adhesion geometry, not styling. Keep large flat expanses off the bed where you can, since they store the most contraction. On the machine side, keep the bed hot enough to hold the plastic soft and adhered — around 50–60 °C for PLA, 70–85 °C for PETG — add a brim for gripping area at the edge where peeling starts, and shield the print from draughts. A raft is the last resort: it forgives a bad bed and spreads the contraction stress, but it's slow, wastes plastic and leaves a rough underside.

The base came out swollen — elephant's foot

The third signature is sneakier because the part stuck perfectly and still came out wrong: the base is fatter than you drew it, and holes near the plate are tighter than the rest of the bore. This is the elephant's foot, and it's the dark side of good adhesion. The same squish that makes the first bead stick pushes leftover plastic sideways, bulging the contour out past nominal for the first few tenths of height, then recovering. Push harder to stick better and you get more of it — adhesion and accuracy pull in opposite directions.

Because it's local and predictable, the cleanest fix is geometric: a 0.2–0.4 mm chamfer on the bottom edge gives the bulge a void to fill instead of spilling past your dimension. Reach for it wherever the base is a mating surface — the start of a shaft, the face that seats in a pocket, the edge of a lid. On the machine side, raise the z-offset a touch, drop the bed temperature once the part has taken hold, and enable the slicer's elephant's-foot compensation — but keep that value small (0.1–0.2 mm), because it blindly shrinks the first layers and can trim the very footprint that sticks. How the squish moves the size of holes and pegs specifically, and by how much, is covered in Holes, pegs and first-layer squish.

Most adhesion trouble is a foundation problem, and foundations are designed, not tuned. A part with a flat face down, a stable base, rounded footprint corners and a chamfer where the dimension is critical wants to stay on the plate — and a part that wants to stay stuck rarely needs rescuing.