Duplo: LEGO at double scale

A LEGO System stud measures 4.8 mm and mates to a tolerance of hundredths of a millimetre. Get it wrong by a tenth and the part either wobbles or won't seat. On FDM, that tenth is exactly the width of your error band. Duplo changes the rules in the simplest way possible: it makes everything twice as big. It's the range designed for small hands, where the parts are chunky, the edges blunt, and nothing is fine enough to be a choking hazard. That double scale, which exists for safety and motor skills, turns out to help the printer: when every dimension doubles but your process error does not, the fit stops living on the edge of a tenth. This article is about why Duplo is kinder to FDM than its smaller sibling, and where the catch still lies.

What Duplo is: System multiplied by two

Duplo is LEGO at double scale. The grid unit — the pitch between studs — is 16 mm, exactly twice the 8 mm of System, so one Duplo unit in plan covers the same ground as two System units in each direction — that is, four System studs for every Duplo stud. The same holds vertically: a standard Duplo brick is 19.2 mm tall, twice the 9.6 mm of the System brick. The whole system is the same idea scaled by two.

The geometric difference that matters most for printing isn't the size but the shape of the stud. On System the stud is a solid 4.8 mm cylinder that grips by pressing against the walls and tubes of the part above — as LEGO System: the 8 mm pitch and the press-fit clutch explains. On Duplo the stud is hollow, a tube. That isn't arbitrary: a solid stud of twice the diameter would be a mass of plastic that cools badly, sinks on the top surface and shrinks unevenly. Hollowing it turns it into an annular wall of moderate thickness, which cools uniformly and holds its dimension.

The clutch lives in a ring, not in a solid

Here is the heart of the system, and it pays to understand it before touching a single dimension. The friction fit — the clutch — is what makes parts stay put and separate with a deliberate pull, neither loose nor stuck. On System that grip comes from a solid post rubbing against the antistuds of the part above. On Duplo it comes from a tube: the wall of the hollow stud rubs in a ring, pressing simultaneously on the outside against a recess and, depending on the coupling, on the inside as well.

Spreading the clutch around a ring has a direct physical consequence. A thin-walled tube is more elastic than a solid of the same diameter: it gives a little on assembly and springs back, and that radial flexibility is what gives a forgiving clutch rather than an all-or-nothing coupling. It's the same reason System antistuds are tubes and not solid cylinders. When you design a printed Duplo stud, that central void is not material you're saving: it's what sets the stiffness of the grip. A stud you fill in "to make it stronger" grips worse, because it loses the elasticity that absorbs the tolerance.

Why double scale forgives FDM

This is the part that makes Duplo a good candidate for printing. FDM bias is absolute, not proportional: holes come out small and studs come out fat because of first-layer squish, shrinkage and bead width, and those effects add up to roughly the same amount — a fraction of a bead, a few tenths — whether the part is 5 mm or 10 mm. Real printed clearances spells it out in detail: you compensate for a fixed offset, not a percentage.

When the error is fixed but the dimension doubles, it weighs half as much in relative terms. A clearance of 0.15 mm per side on a 4.8 mm System stud is 3 % of the diameter; that same 0.15 mm on a Duplo stud of twice the diameter is 1.5 %. The same machine imprecision that shifts the System fit two rungs towards tight shifts the Duplo fit only one. In practice: a printed Duplo stud grips reasonably well with less fine calibration, because the tenth the machine takes is a smaller share of what's at stake.

The same goes for elephant's foot. Squish on the first layers widens the bottom edge of the brick by a fixed few tenths of a millimetre; on a small System brick those tenths deform a noticeable fraction of the face that receives the studs below, whereas on the wide base of a Duplo brick they're a smaller percentage and, on top of that, easier to shave off with a chamfer on the bottom edge without sacrificing useful geometry.

Dimensions: what's known and what has to be measured

Duplo's large dimensions are mostly deducible from the 2× relationship with System, a relationship that is well documented. Pitch and height are firm. It's with the fine geometry of the hollow stud — outer diameter, inner void diameter, height and wall thickness — that it pays to be honest: LEGO doesn't publish drawings, and the values in circulation are hobbyist measurements of real parts. They are listed below with that distinction made explicit.

Bridges between the two scales

The top of a Duplo part usually accepts System studs: you can rest and clip ordinary LEGO parts on top of a Duplo brick, and that opens the door to adapters between the two scales — a Duplo base with a System surface, for instance. It's a very useful property if you print bridging parts, but don't take it for granted in your design without checking it.

The reason is that this compatibility depends on the exact geometry of the hollow stud and on how System antistuds settle onto it, and that's precisely what no drawing specifies. If you're going to design an adapter, measure the real reference part, print it and verify the clutch in both directions before signing off on the dimension. Honesty here saves reprints: it's faster to measure a stud with callipers than to iterate blind on a number inherited from a forum.

Printing it: studs up, with deliberate clearance

The orientation rule is simple: print the part studs up, just like System. That way the tube of the stud grows in clean horizontal layers, with its inner void as a well-formed vertical hole, instead of coming out oval and drooping if you lay it down. A stud printed on its side loses the circular section the clutch needs and the ring stops pressing evenly.

Account for anisotropy too. With the stud printed vertically, the layer lines lie horizontal and perpendicular to the assembly-and-disassembly effort, which is exactly the direction in which FDM tends to split the layers. Unlike a moulded part, which is isotropic, a printed stud loses grip over cycles if the layers peel apart. Turn the extrusion temperature up to weld well between layers and give the tube a couple of extra perimeters: a ring with good interlayer bonding survives many more assemblies.

Give deliberate clearance in the fit, starting from the numbers in Real printed clearances, but remembering that here you're playing with an advantage: the large dimension absorbs error better, so the tenth that sat right at the limit on System leaves you margin here. Start on the loose side — it's always easier to tighten by reprinting a touch than to rescue a stud that won't go in — and adjust from there.

And remember where elephant's foot falls in this orientation: not on the studs, which come out at the top in the last layers, but on the bottom edge of the brick, which rests on the bed. A small chamfer on that bottom edge, two or three tenths, takes away the first-layer squish without touching the face that mates over the studs below. Duplo's wide base tolerates that tweak well; on a smaller System part the same squish removes a more critical fraction of the bearing contour.

Duplo is, fundamentally, the same clutch problem as System, solved with a few more millimetres of margin. If you've already understood why at the small scale a tenth decides the fit — you have it in LEGO System: the 8 mm pitch and the press-fit clutch — here you only have to apply the same reasoning, knowing that the process forgives more. The next step is always the same: measure a real part, print a test coupon of the hollow stud and confirm the clutch before committing a whole set.