Thermoplastic polyurethane in SLS is the option I reach for when a part needs real flex, not just a little compliance. The best results with SLS TPU come from treating it as a functional elastomer, not as a softer nylon: it can bend, rebound, and seal, but it still reacts to heat, wall thickness, and build layout. This article covers where it fits, how to design it, what usually goes wrong, and what I would check before committing to production.
What matters most before you commit to TPU powder printing
- TPU in SLS is strongest when the part needs repeated bending, shock absorption, or sealing performance.
- Expect a matte, slightly grainy finish and more conservative fine-detail behavior than with rigid nylon powders.
- Use thick-section control, sensible clearances, and test coupons to avoid warp and detail loss.
- In production, SLS usually outperforms FDM TPU when you need nested batches and support-free internal geometry.
- Skin-contact and medical use can be a fit, but final validation still belongs to the manufacturer.
Where TPU powder earns its place in production parts
TPU in SLS is not a general-purpose replacement for every nylon part. I use it when the geometry itself has to move, absorb force, or seal against another surface. That is a narrower job than many teams first imagine, but it is exactly where the material pays for itself.
The strongest use cases are the ones where elasticity is part of the design intent:
- Gaskets and seals that must compress repeatedly without cracking.
- Vibration dampers, bumpers, and cushioning inserts for mechanical assemblies.
- Shoe soles, orthotics, prosthetics, and wearables that need a soft touch and repeatable geometry.
- Flexible covers, cable boots, bellows, and protective sleeves.
- Soft grips and ergonomic pads for tools or consumer products.
A common commercial grade sits around Shore 90A, with printed elongation often starting around 110% in the build direction and rising much higher in-plane. That range is useful because it tells me the material is firm enough to keep shape, but elastic enough to work like a real engineered elastomer. When a part needs to flex dozens or hundreds of times, that balance matters more than a glossy finish ever will. From there, the next question is whether the CAD is actually printable without fighting the process.
How to design parts that print cleanly
I start with the geometry, not the slicer. TPU powder is forgiving in one sense because it needs no support structures, but it is less forgiving than rigid nylon when thin walls, small text, or thick sections trap heat. If I want a softer feel without turning the model into a floppy mess, I reduce solid volume with lattice structures instead of thinning every wall and hoping for the best.
| Feature | Practical starting point | Why it matters |
|---|---|---|
| Unsupported vertical walls | 0.6 mm minimum | Below that, walls can warp or detach. |
| Unsupported horizontal walls | 0.3 mm minimum | Thin roofs are easier to distort during sintering. |
| Pins and wires | 0.8 mm recommended | Very slender features can deform or break during cleaning. |
| Small assembly clearances | 0.2 mm for features under 20 mm², 0.4 mm for larger features | Integrated parts need room to move after printing. |
| Separate parts in the same build | 1.0 mm minimum, 5.0 mm preferred | Spacing reduces thermal warping between neighboring parts. |
| Hole diameter | 1.0 mm recommended | Tiny holes tend to close during printing. |
| Drain holes for cavities | At least 3.5 mm, with two holes per cavity | Powder must escape or the cavity stays packed. |
For TPU specifically, I stay even more conservative with fine text, tiny pockets, and thick solid blocks. Once a section gets into the 3 to 4 cm range, detail on the upper faces can fade because of heat buildup, so I will hollow the form, reorient it, or break it into a more thermal-friendly shape rather than hope the detail survives. Warp usually shows up first on the bottom faces of the part, which is why orientation is not cosmetic. That is also the point where it helps to compare the process with other flexible options instead of assuming all flexible plastics behave the same.
How SLS TPU compares with FDM TPU and nylon
This is where teams often make the wrong call. They compare machine price first, when they should compare the part behavior they actually need. Flexible SLS parts solve different problems than filament-based TPU, and rigid nylon still has its own place.
| Aspect | Laser-sintered TPU | FDM TPU | SLS nylon |
|---|---|---|---|
| Supports | No supports, powder acts as the surrounding support medium | Often possible, but support removal is awkward on flexible parts | No supports, similar powder-bed freedom |
| Geometry freedom | Excellent for nested parts, lattice structures, and enclosed forms | Good for simpler shapes, weaker on fine internal features | Excellent, but less compliant |
| Surface finish | Functional, matte, slightly grainy | Layer lines and stringing are visible | Usually cleaner and more uniform |
| Flexibility | True elastomeric behavior with good rebound | Flexible, but print quality depends heavily on tuning | Much stiffer |
| Best fit | Seals, dampers, wearables, production-flex parts | Low-cost simple one-offs and early prototypes | Rigid functional parts and snap-fit components |
Post-processing that helps and post-processing that hurts
After printing, I start with gentle depowdering and only escalate cleaning if the geometry can take it. Flexible powder parts can trap loose material inside channels, around lattice edges, and under overhangs, so the cleanup step is not cosmetic; it is part of the functional result.
What helps is controlled airflow, careful media blasting, and patient inspection of seals, holes, and snap features. What hurts is aggressive blasting on thin webs, over-polishing small contact edges, or handling the part before it has cooled and stabilized. If the part will be dyed or color-matched, I prefer a process that does not round off critical edges first. Compression set is the term I watch closely here: it describes how much an elastomer fails to spring back after being held under load and heat, and it tells me more about real-world sealing behavior than a simple hardness number does.
For production work, I inspect three things every time: whether the part clears powder cleanly, whether the flexible zones rebound as expected, and whether the critical dimensions still sit inside tolerance after cleanup. That is the point where a part stops being a CAD model and becomes something you can trust in a product. If those checks pass, the last step is a quick reality test against the actual job the part has to do.
What I would validate before approving a TPU build
Before I sign off on a flexible SLS part, I run a short checklist. It is usually faster than arguing over CAD assumptions after the first failed batch.
- Confirm that the part actually needs elastomeric behavior, not just a softer feel.
- Check that the thinnest walls, smallest holes, and tightest clearances are all realistic for the chosen printer and powder.
- Make sure the part can be depowdered without damage, especially if it has cavities or fine internal channels.
- Test the part in the same bend cycle, load, temperature, or cleaning environment it will see in service.
- Use a coupon or partial build first if the geometry includes thick solid sections, long flat spans, or critical sealing faces.
That small validation step usually tells me whether the design is ready for a batch or still needs one more round of geometry cleanup. If the test piece keeps its rebound, stays clean, and seals where it should, the full build is usually worth moving forward with.
