The common intuition that more infill equals stronger parts is approximately half right and half wrong. Infill does contribute to compressive and some bending stiffness, but wall count — the number of perimeter loops — is the dominant factor for tensile strength, part rigidity, and resistance to impact. Understanding which structural variables matter under which load types is the foundation of designing 3D printed parts that perform rather than fracture unexpectedly.
Layers Are the Weak Direction — Always
FDM parts are structurally anisotropic: they're strongest in the XY plane (parallel to print layers) and weakest in Z (perpendicular to layers). The interlayer bond in FDM forms through melt-zone diffusion — molten material from the new layer fuses with the partly-cooled surface below it. Even under optimal conditions, this bond achieves 50–80% of the material's bulk tensile strength. Under typical hobbyist printing conditions (open printer, standard fan cooling, unoptimized temperature), 40–60% is more common.
This anisotropy is not a defect to be eliminated — it's a design variable to be exploited. Orienting a part so that primary load-bearing features have their critical cross-sections parallel to the XY plane instead of perpendicular to it is the single highest-impact decision in structural FDM design. A bracket that will carry tension along its length should be printed standing up (load axis vertical) only if the Z-direction strength of the specific material in use meets the load requirement. More often, printing it flat (load axis horizontal) is structurally superior even though it requires more support material.
Wall Count: The Primary Structural Variable
Perimeter walls (also called shells or contours) are solid material deposited around the perimeter of each layer slice. They form the outer surface of the part and carry the majority of structural load for typical parts. For parts that will see tensile or bending loads, wall count matters more than infill percentage in virtually every realistic use case.
Two walls is the minimum for a functional structural print at 0.4 mm nozzle. Three walls handles most general-purpose functional parts (brackets, frames, housings). Four walls is appropriate for high-stress applications — mounting brackets, tool handles, anything with concentrated loads at specific features. Five or more walls for the cross-section to be substantially all wall (approaching solid), which is the practical equivalent of 100% infill for structural purposes but without the time cost of cross-hatched infill patterns.
The reason walls dominate over infill is that walls run continuously around the perimeter with no void interfaces, while infill lines have start-stop points and interface with walls at their endpoints. Under tensile load parallel to the print plane, the continuous perimeter loops carry load uniformly; under the same load, the infill transfers force at these discontinuities, introducing stress concentrations. For bending loads, the outer perimeter walls are at maximum distance from the neutral axis and therefore carry the highest bending stress — wall count directly determines bending resistance.
Infill: What It Actually Contributes
Infill primarily provides two things: resistance to buckling of thin-wall sections under compressive load, and some additional mass contribution to the structural cross-section under compression. For compressive loads (a leg of a table, a post bearing downward force), infill percentage has real impact — denser infill resists local buckling of long thin sections and increases compressive stiffness. For tensile and bending loads, infill above 20–25% provides diminishing returns relative to its time cost.
Infill pattern matters for specific load cases. Gyroid and 3D honeycomb patterns distribute stress more evenly in three dimensions than rectilinear patterns and perform better under impact loading because the curved geometry has no preferred crack propagation direction. Rectilinear and grid patterns are faster to print and adequate for most general-purpose structural applications. Lightning and organic tree patterns (low-density support-style infills) are for cosmetic or non-structural mass reduction — they provide minimal structural contribution.
Top and Bottom Layers and Structural Decks
Top and bottom layers form solid structural decks that distribute loads from the part's exterior surfaces into the infill and perimeter structure. Too few solid layers (fewer than 3) leaves the infill pattern visible on the top surface and allows localized stress concentrations where point loads pierce through to the infill. Four to five top/bottom layers ensures uniform load transfer for most part geometries.
For parts that will be compressed (gasket seats, press-fit housings), increasing bottom layer count to 6–8 layers with 100% rectilinear infill pattern in those layers improves local stiffness and surface quality at the contact interface.
Temperature and Strength
Printing temperature directly affects interlayer bond strength. For PLA, the difference between printing at 195°C and 220°C can be 20–30% improvement in Z-direction tensile strength. Higher temperatures increase melt-zone diffusion time and produce deeper interpenetration between layers. The practical ceiling is the material's degradation temperature — for PLA, above 230°C, color changes and property degradation begin to occur. For PETG, printing at 240–245°C rather than 230°C produces measurably stronger interlayer bonds without meaningful degradation risk.
Annealing PLA after printing further improves interlayer bond strength (and bulk crystallinity) at the cost of slight dimensional distortion. Parts annealed at 60–65°C for 4–8 hours show 15–30% improvement in Z-direction strength and substantially improved heat deflection temperature. For functional brackets or fixtures in warm environments, annealing is a legitimate performance upgrade.
Material Choice and Its Structural Ceiling
Settings optimization has limits set by the base material's inherent properties. Standard PLA at maximum wall count and optimal print temperature reaches a tensile strength ceiling around 50–60 MPa in the XY direction. PETG reaches 45–55 MPa with better ductility. Nylon PA12 reaches 50–60 MPa with substantially better impact resistance and fatigue life. For parts that must survive repeated load cycling or impact, material choice matters more than wall-count optimization once you're above 3–4 walls.
Fiber-reinforced materials (CF-Nylon, GF-PETG) extend these ceilings significantly but introduce the anisotropy and hardware requirements discussed elsewhere. For most functional FDM work — brackets, mounts, enclosures, tools — well-configured PETG or ABS at 3–4 walls with 20–25% infill and 5 top/bottom layers hits a performance level adequate for household and workshop use loads. The prints that fail unexpectedly almost always fail not from incorrect settings but from incorrect orientation — a bracket printed so that its primary tensile load runs Z-parallel, fracturing along layer lines under normal service.
