Most failed prints are design problems, not printer problems. The rules that prevent them are short: keep walls at least two to three nozzle widths thick, keep unsupported overhangs steeper than about 45 degrees from horizontal, oversize holes and expect to ream them, never print fine threads directly, leave real clearance between mating parts, and orient the part so layers run across the load rather than along the break line. Follow those and the reprint rate collapses. Below is each rule, why it exists, and how it changes by process.

Wall thickness: the first and most important rule

A wall thinner than the extrusion width cannot be printed as a solid feature — the slicer either skips it or renders it as a fragile single trace. As a working minimum, use about 1.2–1.6 mm for FDM structural walls and no less than roughly 0.8 mm for anything that carries load. Resin printing tolerates thinner walls, but very thin resin sections warp during post-curing and snap easily.

Uniform thickness matters as much as absolute thickness. Abrupt jumps from thick to thin create differential cooling, and differential cooling creates warp. This is the same principle that drives wall thickness in injection molded parts, and designing to it from the start makes a later switch to tooling far easier.

Overhangs, bridges, and supports

Every layer needs something under it. The practical threshold for FDM is roughly 45 degrees from horizontal — steeper walls print cleanly, shallower ones need support and come out rough. Horizontal spans can bridge a limited distance before they sag.

  • Add a chamfer or fillet instead of a flat horizontal overhang wherever the function allows it — a 45-degree chamfer under a boss or flange removes the need for support entirely.
  • Design so support material never lands on a cosmetic or functional surface. Support scars are the roughest area of any print.
  • Remember support has to be reachable. An internal cavity full of support you cannot get a tool into is a scrap part.
  • Resin printing has its own version of this problem: large flat areas parallel to the build plate generate high peel forces and should be tilted.

Holes, pins, and threads

Printed holes come out undersized, consistently. Extruded plastic pulls inward as it cools, and small vertical holes shrink most.

  • Oversize holes that need accuracy, or model them undersized and drill or ream to final dimension.
  • Horizontal holes print as slight ellipses. Where roundness matters, orient the hole vertically or plan on machining it.
  • Small printed threads are weak and dimensionally unreliable. Use a heat-set brass insert, a threaded nut pocket, or a self-tapping screw into a plain boss. Large threads — jar-lid scale — print acceptably.
  • Thin pins snap. Either thicken them or replace them with a metal dowel in a printed hole.

Clearances between parts that move or assemble

Zero clearance in CAD is a fused joint in reality. Give yourself real gaps:

  • Sliding or rotating fits: around 0.3–0.5 mm on FDM, less on resin.
  • Snap-together assemblies: around 0.15–0.25 mm, adjusted after the first test print.
  • Press fits: a slight interference only, and test it — printed plastic will crack rather than deform if you overdo it.

Print a small clearance test coupon on your actual machine and material before committing an assembly. It costs an hour and saves several full reprints. The underlying variation is explained in 3D printing tolerances.

Orientation determines strength

FDM parts are anisotropic: layers bond to each other far more weakly than the plastic bonds to itself within a layer. A hook printed flat on the bed can be several times stronger than the same hook printed upright, purely because of how the layers meet the load.

Decide the print orientation while you design, then shape the part around it. If a bracket must resist bending in one axis, lay the layers across that bending stress rather than along it. Powder-bed processes such as SLS are much closer to isotropic, which is one reason they are used for end-use parts. Background in are 3D printed parts strong enough.

Text, logos, and surface detail

Raised text prints far more reliably than recessed text on FDM. Keep stroke width and depth above roughly 0.8 mm and 0.5 mm respectively, and put text on vertical or top surfaces rather than on supported undersides. Resin printing handles much finer detail — small serial numbers and fine logos are realistic there.

Designing to cut cost without hurting the part

  • Reduce volume, not wall count. Hollow interiors, ribs, and gussets keep stiffness while removing material and print time.
  • Cut supports. Support is paid for twice: in material and in the labor to remove it.
  • Split large parts. Two smaller printed halves bonded together often beat one tall print that fails at hour nine.
  • Combine parts. Printing lets you consolidate an assembly into one piece, removing fasteners and assembly labor.
  • Nest for the machine. If you need dozens of parts, a design that packs efficiently into the build volume prints for much less each.

More options in how to cut prototype costs.

Design for the specific material and process

These rules shift with technology. Resin allows thinner walls and finer text but hates thick solid blocks. SLS needs escape holes so unfused powder can be removed from internal cavities. Filled filaments are stiffer but more brittle, so sharp internal corners become crack starters — fillet them. Choose the material first, using our 3D printing materials guide and the FDM vs SLA comparison, then apply the numbers that belong to it.

And if the part is headed for volume production, design it for that process now rather than converting later — see design for manufacturing and the 3D printing pillar.

Want a design review before you print? Send us your model through the contact form and we will flag the features likely to fail and suggest fixes that keep the part functional.