Designing 3D-Printed Flexible Lattices and Chainmail: TPU-Free Mesh Structures and Design Rules

Printing TPU gives you a flexible part, but printing a lattice — a repeating structure of interconnected links, hinges, or cells that moves and drapes like fabric even when printed in a completely rigid material like PLA or PETG — is a different trick entirely, and one that catches a lot of makers off guard the first time they see it work. 3D-printed chainmail, woven lattice panels, and interlocking mesh structures don't rely on the filament itself being flexible; they rely on geometry: individually rigid links that are mechanically free to rotate and slide against each other, printed already assembled thanks to FDM's ability to leave a small clearance gap between touching surfaces. This guide covers the design rules that make these structures actually work, print settings specific to interlocking geometry, and where the technique is genuinely useful versus where TPU would just be the simpler answer.

How Rigid-Material Flexibility Actually Works

Two distinct mechanisms get called "flexible 3D printed lattice," and they behave very differently:

• Print-in-place interlocking links (true chainmail) — individual closed loops or links printed already interlocked with their neighbors, each one fully rigid on its own, with the assembly's flexibility coming entirely from the links being free to pivot and slide against each other, exactly like metal chainmail. Nothing about the plastic itself bends.
• Compliant lattice / auxetic structures — a continuous printed structure (not separate interlocking pieces) where thin, deliberately under-dimensioned struts or hinge points flex elastically under load even in a nominally rigid filament like PLA, because the cross-section at the flex point is thin enough to bend without snapping. This is the same underlying principle as a print-in-place living hinge or articulated joint, just applied repeatedly across a grid rather than at a single joint.

Chainmail-style structures give near-unlimited flex life (the links aren't being stressed near their material limit, they're just rotating) but add real weight and bulk per unit area. Compliant lattices are thinner and lighter but have genuine fatigue life limits, since you're cyclically stress-loading the material itself, not just letting rigid parts rotate freely.

Print-in-Place Clearance: The Number That Makes or Breaks the Whole Model

Every interlocking-link design lives or dies on one dimension: the gap left between two surfaces that touch but must not fuse during printing. Too tight, and the links weld together into a solid, immovable blob; too loose, and the links rattle, lack mechanical integrity, or simply fall apart during printing before the surrounding structure sets them in place.

Printer / Process ClassTypical Working ClearanceNotes Well-tuned direct-drive FDM (0.4mm nozzle)0.2–0.3mmStarting point for most consumer FDM printers; tune from here per-printer Loose/older Bowden FDM setups0.3–0.4mmMore ooze and less precise flow control needs extra margin Well-calibrated CoreXY (Voron-class)0.15–0.25mmTighter achievable clearance thanks to better dimensional accuracy and consistent flow Resin / MSLA0.1–0.15mmMuch finer achievable gap, but supports and cleanup inside tight interlocking geometry are a real complication — FDM is generally the easier process for this specific technique

This number is specific to your printer, your flow calibration, and even your filament batch — it is not a universal constant you can copy from someone else's model notes and expect to work unmodified. Always run a calibration test piece first: a small design with 3–5 interlocking rings or links at a range of clearance values (print a staircase of test links at 0.15mm, 0.2mm, 0.25mm, 0.3mm, and 0.35mm gap) and physically test which ones separate cleanly after printing without rattling loosely. This single test, done once per printer/filament combination, saves hours of failed full-size prints.

Calibration Prerequisites

Clearance tolerances this tight don't forgive sloppy dialed-in settings elsewhere. Before attempting a lattice or chainmail model, confirm:

• E-steps and flow rate are calibrated accurately — see our e-steps and flow calibration guide. Over-extrusion is the single most common cause of welded-together links, since it fattens every wall slightly and eats directly into your clearance margin.
• First layer and general dimensional accuracy are dialed in — a printer that's consistently printing holes 0.1mm undersized or walls 0.1mm oversized is effectively subtracting that error from your clearance budget before you've even accounted for it in the design.
• Pressure advance / linear advance is tuned if you're on Klipper or Marlin — blobbing or stringing at direction changes, common around tight interlocking curves, directly interferes with clean separation.

Print Settings for Interlocking Geometry

• Cooling: aggressive, from the first layer up where geometry allows. Interlocking links typically have a lot of short, tight perimeter segments and bridges between adjacent features — without strong part cooling, each new link's wall is still partially molten when the nozzle moves to lay down its neighbor, which both softens overhangs and risks exactly the kind of micro-welding between touching surfaces you're trying to avoid.
• Print speed: moderate, not maximum. Slower travel and perimeter speeds around tight link geometry reduce the risk of the nozzle dragging a still-soft adjacent link out of position before it's had time to solidify.
• Retraction: tuned tight. Lattice and chainmail models have an unusually high number of short travel moves between individual links, and stringing between adjacent-but-unconnected links both looks bad and can physically bridge a gap that was supposed to stay open.
• Supports: avoid if at all possible. Support material generated inside a tight interlocking gap is often impossible to remove cleanly without damaging the link, and support dissolved with a soluble material (PVA via an AMS/ACE-style multi-material system) is the only really clean option if a design genuinely can't avoid needing it. Most published chainmail and lattice models are specifically designed with self-supporting geometry (each link's upper surface oriented to bridge cleanly rather than needing support) for exactly this reason — look for that design detail when sourcing or designing a model.
• Orientation matters more here than almost any other print type — a flat lattice panel printed face-down on the bed prints every link's gap horizontally (bridging, which FDM handles reasonably well with good cooling) rather than vertically (which stacks the clearance gap across layer lines, a much less forgiving geometry for clean separation).

Material Choice

PLA is the easiest starting material for this technique specifically because of its low layer adhesion temperature and tendency to cool and solidify fast, both of which help prevent adjacent links from fusing. PETG's stringier, hotter-flowing behavior makes it noticeably harder to dial in clean separation and is not the best first material to learn the technique on, even though it's generally considered the tougher, more durable choice once you've got settings dialed in on a specific model. Nylon's flexibility and durability make it the ideal material for a finished, wearable, or heavily-used chainmail piece, but its strong tendency to warp and its higher required nozzle/bed temperatures make it a poor choice for learning the clearance-tuning process — dial in your approach on PLA first, then move to nylon for a final, durable version once you know your printer's working clearance number.

Design Applications

Beyond wearable chainmail for costume and cosplay work (a natural pairing with our cosplay post-processing guide), the same print-in-place linked-lattice principle shows up in flexible phone cases built from individually rigid plates, draping light fixtures and lampshades, articulated robot skin/armor plating, and load-distributing padding structures used in some prosthetic and orthotic work. Anywhere you want drape, flex, or multi-axis articulation from a filament that itself has none, this is the technique to reach for instead of switching to TPU — particularly useful when the application also needs TPU-incompatible properties like PETG's chemical resistance or nylon's abrasion resistance that TPU doesn't offer at comparable strength.

The honest tradeoff to weigh before committing to this approach over simply printing in TPU: lattice and chainmail structures take meaningfully longer to dial in (that clearance calibration step is not optional) and print slower than an equivalent solid-flexible part, in exchange for drape behavior, puncture resistance, and material property options TPU simply can't match. For a one-off flexible bracket, TPU is still the faster path. For wearable mail, articulated armor, or anything that needs to move like fabric while resisting abrasion or chemicals like a rigid plastic, the interlocking-link technique earns the extra setup time.