SLS 3D Printing: The Complete Guide to Powder-Bed Limits

• 1
  How SLS 3D Printing Fuses Powder
• 2
  Design Rules That Decide Whether SLS 3D Printing Works
• 3
  Nylon, Fillers, and What Each Grade Buys You
• 4
  Tolerance, Anisotropy, and Where SLS Drifts
• 5
  When SLS 3D Printing Beats Machining and Molding
• 6
  SLS 3D Printing vs CNC Machining vs Injection Molding

Selective laser sintering builds parts inside a heated powder bed. A recoater blade or roller spreads a layer of polymer powder, typically 0.08–0.12 mm thick, across the build platform. A CO2 laser then scans the cross-section that belongs to the part and melts the polymer particles where the beam passes. The platform drops by one layer thickness, fresh powder rolls over the top, and the cycle repeats.

The key detail is that the bed stays hot. Chamber temperature is held just below the material's melting point, often 160–175 °C for polyamide 12. That preheat means the laser only needs to add a small amount of energy to trigger melting. Without it, the part would curl as the melt pool cooled against cold powder. The bed is the process, not the laser.

Because unsintered powder surrounds every layer, the powder itself supports overhangs. You can build a hollow sphere with no internal scaffolding and no breakaway marks on the downward faces. This is the main reason SLS 3D printing suits complex ducting, lattice brackets, and parts with internal channels that would be impossible to support in resin or FDM.

After the build finishes, the entire powder block cools slowly inside the machine, sometimes over 6–12 hours. Rushing this stage is the most common cause of warp. Parts are then excavated, bead blasted to remove clinging powder, and the remaining cake is sieved for reuse. Reclaimed powder is blended with virgin material in ratios that affect surface finish and mechanical strength.

• 1
  Layer thickness0.08–0.12 mm is the usual band; thinner layers improve Z finish, not XY accuracy.
• 2
  Chamber preheatNear-melt bed temperature keeps the part from curling against cold powder.
• 3
  Self-supporting powderOverhangs down to 0° are buildable without dedicated support geometry.

Start with wall thickness. Glass-filled or plain PA12 holds a 0.8–1.0 mm wall, but anything below 0.6 mm risks a porous skin and warped edges. If a wall is load-bearing, keep it at 1.5 mm or more. Thin ribs are fine when they are short; long, unsupported fins warp because the sintered skin cools faster than the core.

Holes are where most first-time designs lose tolerance. A hole smaller than Ø1.5 mm may close up or sinter shut depending on orientation. Even larger holes shrink during cooling, so a nominal Ø6 mm hole can come out 0.15–0.25 mm under. The practical fix is to model holes undersized by that margin, or plan to drill and ream them after sintering.

Clearances matter as much as features. Mating parts need 0.3–0.5 mm of gap if they must move freely after the build. Anything tighter may fuse at the interface because heat conducts across the powder gap. For snap fits and living hinges, print a test coupon first: powder-bed hinges behave differently from injection-molded ones and will crack if the section is too thick.

Escape paths are the silent killer. Any enclosed hollow volume traps unsintered powder that cannot be removed after the build. Add two or more Ø4–5 mm drain holes at the lowest points of the cavity, or design the cavity as an open channel. A sealed internal void is a rattling part and a dimensional variable you cannot inspect.

• 1
  Minimum wall0.8–1.0 mm for PA12; 1.5 mm for load-bearing walls.
• 2
  Minimum holeØ1.5 mm; expect 0.15–0.25 mm shrinkage and plan for reaming.
• 3
  Clearance0.3–0.5 mm between mating surfaces that must move.
• 4
  Powder escapeØ4–5 mm drain holes at the low points of any hollow body.

Polyamide 12 is the default. It has the widest sintering window, good chemical resistance, and predictable shrinkage, which is why it dominates production SLS. Parts come out slightly grainy, with a matte surface that bead blasting turns uniform. It takes a dye bath well, so black and grey finishes are common without paint.

Polyamide 11 is a bio-based alternative with better impact strength at low temperature and slightly more ductility. It costs more and sinters in a narrower window, so it is reserved for parts that flex or see cold environments. Glass-filled PA12 raises stiffness and heat deflection, but abrasion at the recoater shortens machine life and the surface gets rougher.

For parts that need to slide or wear, unfilled PA is a poor choice on its own. Internal lubricants and molybdenum-disulfide-filled grades exist, but they are specialty powders with their own build parameters. If a bushing must hold a running fit, machining from POM or bronze is usually the cheaper path than tuning a filled SLS powder.

TPU is available for flexible SLS parts, typically in a 45–90 Shore A range. It sinters into a porous, springy lattice and is useful for grips and dampers. It is not a replacement for molded rubber: tear strength is lower and the surface retains powder in deep texture. Keep flexible parts simple and avoid thin, high-stress flexures.

• 1
  PA12Default grade: wide process window, predictable shrinkage, easy dyeing.
• 2
  PA11Better low-temperature impact and ductility; narrower sintering window.
• 3
  Glass-filled PA12Higher stiffness and heat deflection; rougher finish and faster wear.

A well-tuned SLS machine holds roughly ±0.3 mm on a 100 mm part, or about ±0.3% of the dimension, whichever is larger. That is a process capability figure for the whole build, not a per-part guarantee. Large thin plates and long beams drift further because they cool unevenly across the bed.

The build is anisotropic. Z-direction strength is typically lower than XY because layer-to-layer bonding is weaker than the sintered material within a layer. A part loaded in tension along Z may fail at 80–90% of the XY value. Orient critical load paths in the XY plane and avoid pulling directly across layer boundaries.

Position in the build envelope changes the result. Parts near the center of the bed see more uniform chamber temperature than parts at the edges, so they distort less. If a batch contains one tight-tolerance feature, nest it near the center and keep the surrounding build light. Packing a dense block of parts around it trades accuracy for throughput.

Post-processing moves dimensions too. Bead blasting removes 0.02–0.05 mm of skin depending on pressure. Dyeing adds no meaningful thickness but can swell thin walls slightly. If a feature is already at the edge of tolerance after sintering, blasting will push it out. Specify the finish before you fix the drawing tolerance.

• 1
  CapabilityAbout ±0.3 mm or ±0.3% on a clean build, whichever is larger.
• 2
  AnisotropyZ strength is roughly 80–90% of XY in tension.
• 3
  Build positionCenter parts hold tolerance better than edge parts.

SLS wins on geometry freedom, not unit cost. There is no tooling, no support removal, and no per-part setup. A single complex bracket with internal channels can ship from a powder bed for less than the programming alone on a 5-axis job. That makes it the right call for prototypes, bridge parts, and low-volume runs where design changes are still likely.

It loses on surface finish and precision. As-built SLS surfaces sit around Ra 8–15 μm, far coarser than a machined Ra 1.6–3.2 μm. If a part needs a sealing face, a bearing bore, or a press fit, plan to machine those features after sintering. Hybrid workflows are common: print the blank, then finish the critical diameters on a CNC.

Volume is the other boundary. Powder-bed builds are priced by envelope space, not by part count, so nested small parts get cheap fast. But once a part is simple enough to mold, injection molding overtakes SLS somewhere in the low thousands of units. For a plain enclosure with no internal features, the crossover can be earlier than most teams expect.

Material is the last filter. SLS is a polymer process. It does not produce structural metal parts, and it will not hold anodized or plated finishes. When the drawing calls for 6061-T6, 17-4PH, or a hardcoat surface, the powder bed is the wrong machine regardless of how complex the geometry looks.

• 1
  Choose SLSComplex internal geometry, low volume, design still changing.
• 2
  Choose CNCTight bores, sealing faces, metal, or a specified surface finish.
• 3
  HybridPrint the near-net blank, machine only the critical interfaces.