Direct answer: Fix TPU stringing by identifying the web pattern before touching retraction. If the spool became stringier after sitting out and extrusion is also rough, poppy, or inconsistent, test spool condition first. If a known-good dry spool leaves long elastic hairs but prints clean walls, test nozzle heat and exposed travel next. Check feed resistance before slicer tuning, and adjust retraction last because too much can replace strings with delayed restarts, thin walls, or a jam.
This page covers one exact failure: TPU stretches into unwanted hairs or webs while the nozzle travels between separate printed features. It does not cover a single raised seam, fuzzy under-extrusion everywhere, support-removal scars, or strands caused by a nozzle dragging through a curled feature.
Use this fast TPU stringing triage
- Confirm the strand crosses open air. Match it to a travel move in the slicer preview. A repeated bump on one vertical line, fuzzy walls everywhere, or a strand pulled from a curled edge belongs to a different fault.
- Print the unchanged baseline twice. Judge the webs, walls, and first line after every travel landing together. If the symptom is not repeatable, inspect spool motion and the feed path before tuning.
- Choose the first check from the evidence. Exposure plus popping or rough flow points to spool condition. Hesitation or changing restarts points to feed drag. Clean walls plus long elastic hairs points to a supported temperature test. Strings limited to one layout point to exposed travel geometry.
- Change one control and repeat. Keep a change only when it reduces unwanted webs in both repeats without adding thin restarts, weak layers, rough extrusion, buckling, or a clog.
- Leave retraction until last. Test it only after the exact TPU baseline, spool condition, feed path, heat, travel, and throughput are stable.
Stop rule: if clicking, buckling, missing restart lines, or unstable extrusion appears, stop the stringing test and correct the feed or hotend fault. A lower string count is not a fix when the printed path is no longer continuous.
Editorial scope: This workflow uses current manufacturer guidance and does not claim hands-on testing. TPU hardness, formulations, extruders, and slicer labels vary. Keep drying, temperature, speed, travel, and retraction changes inside the exact printer and filament maker's supported limits.
Disclosure: GoodPrints may earn from qualifying purchases through the clearly labeled affiliate link on this page, at no added cost to you.
Confirm the defect is TPU stringing
Look for material spanning open air between islands, posts, holes, or separate parts. True stringing follows travel moves. It may appear as fine hairs, elastic threads, or thick irregular webs. Mark where each strand begins and ends, then compare those paths with the slicer's travel preview.
Separate strings from four similar defects
A bump that repeats on one vertical line is more likely a seam issue. Fuzz on every wall can be unstable extrusion rather than travel ooze. Torn material beside supports belongs to the TPU support-scar guide. A nozzle that crosses a lifted edge can pull a strand mechanically; use the nozzle-scraping workflow for that evidence.
Match the web pattern to the first useful check
| What you observe | Leading hypothesis | First clean proof |
|---|---|---|
| Thick irregular webs plus popping, rough walls, or a spool that worsened after exposure | Spool-condition drift | Compare the same slice before and after the maker-approved drying process |
| Stringing changes between runs and extrusion hesitates after travel | Feed drag or flexible-filament compression | Remove spool resistance and sharp path bends without changing the profile |
| Long elastic hairs with clean walls and reliable restarts | Melt staying too mobile during travel | Test one supported temperature step while holding every other variable fixed |
| Webs appear mainly across wide open gaps or one model layout | Exposed travel geometry | Inspect travel preview, then shorten or redirect one crossing |
| Strings fall after a retraction change, but restart gaps or thin sections appear | Over-retraction | Return to the last clean-restart setting; do not count the trade as a fix |
| One raised line or blobs only at restart points | Seam or restart control | Use seam view and the TPU seam-bump branch |
Check 1: restore the exact TPU baseline
Record the printer, nozzle, nozzle size, extruder path, TPU product and hardness, spool exposure history, and selected material preset. Remove forgotten overrides. Save nozzle temperature, cooling, speed, volumetric limit, travel settings, and retraction distance and speed before changing anything.
Do not copy a PLA retraction recipe
Flexible filament can compress, buckle, and tangle where a rigid filament would remain stable. Prusa's current flexible-materials guide recommends proper flexible profiles, slower handling, keeping filament dry, and cautious retraction adjustment. Start from the exact maker's TPU profile rather than a popular PLA number or another printer's result.
Check 2: prove whether spool condition changed
Move moisture to the front only when the timeline supports it: the same spool printed cleanly before exposure, then became stringier, rougher, poppy, or less consistent. Follow the exact filament maker's drying instructions and protect the recovered spool during the repeat. Do not invent a universal drying temperature for every TPU.
Use a same-spool comparison
Print the unchanged tower test before the approved drying process and again afterward with the same printer, nozzle, profile, layout, and room setup. If thick random webs and extrusion noise fall together, spool condition mattered. If clean walls still carry the same fine hairs, continue to heat and travel. The TPU-after-sitting-out guide owns that narrower timeline.
Check 3: remove feed-path resistance
Watch the spool, inlet, tube path, gears, and nozzle feed during the baseline. Look for a spool that snags, a sharp tube bend, a guide that squeezes soft filament, gear marks, buckling, or delayed extrusion after travel. Feed instability can change both the amount of stringing and the restart, making later retraction conclusions unreliable.
Fix the mechanical variable before the slicer variable
Let the spool turn freely and remove avoidable bends while preserving the supported path. If lines are thin or missing beyond the travel landings, use the under-extrusion workflow. Do not increase global flow to hide a path that intermittently starves.
Check 4: test nozzle heat inside guidance
When a known-good spool feeds consistently but leaves long elastic hairs, run one small nozzle-temperature comparison within the filament maker's range. Keep travel, speed, cooling, retraction, spool, and model fixed. The goal is not the lowest possible temperature; it is the cleanest travel behavior that still gives reliable extrusion, bonding, and restarts.
Read walls and restarts with the strings
A temperature step is useful only if strings improve without introducing weak layers, dull starvation, rough flow, or delayed restart. If the part begins to split along printed planes, move to the TPU layer-cracking diagnostic instead of continuing downward.
Check 5: inspect exposed travel
Open the slicer's travel preview and compare it with the strand map. If strings occur mainly where the nozzle crosses wide open gaps, rotate the model, alter part spacing, or use the slicer's documented path controls to reduce those crossings. Keep the material settings unchanged for this proof.
Do not hide every travel inside the part automatically
A path that avoids open air can add time, cross vulnerable surfaces, or create new scars. Inspect the complete preview for perimeters, supports, seams, and collisions. If the nozzle is dragging through a curled overhang, use the TPU overhang-curl guide.
Check 6: test speed only after feed and heat are stable
Printing speed, travel speed, acceleration, and minimum-layer behavior can all change how long TPU remains able to ooze. Use the printer or filament maker's supported TPU starting point. If the current profile visibly compresses or hesitates during extrusion, test one slower print-speed or volumetric-limit change before touching retraction.
Separate print speed from travel behavior
A faster travel may shorten ooze time, but machine limits, vibration, collision risk, and the exact path still matter. Do not change extrusion speed and travel speed together. Bambu's current TPU printing guide treats drying, speed, and retraction as related TPU setup controls; use the correct printer-specific guidance rather than transplanting values.
Check 7: tune retraction last
Retraction can reduce pressure during travel, but flexible filament does not respond like rigid PLA. Start from the exact supported TPU profile. Make one small change, print the same test, and inspect the first extrusion after every travel move. Stop when a lower string count begins to create gaps, thin restart sections, surface scars, buckling, or a clog.
The winning setting must restart cleanly
Count strings and failed restarts separately. A test with fewer webs but missing wall material is worse, not better. Prusa's flexible guide explicitly warns that lower retraction can reduce clogging or tangling risk and suggests increasing cautiously from a low baseline when needed.
Apply only the fix your evidence supports
- Exposure-linked rough webs: follow the exact TPU drying guidance and protect the spool during the repeat.
- Hesitation or changing flow: remove spool drag, path friction, compression, or buckling before slicer tuning.
- Long elastic hairs with clean extrusion: test one supported nozzle-temperature step.
- Webs across specific wide gaps: shorten or redirect exposed travel after inspecting the full preview.
- Compression at the current throughput: test one supported speed or volumetric-limit reduction.
- Residual fine hairs after every earlier check: adjust TPU-specific retraction cautiously and preserve clean restarts.
Run one controlled two-tower test
Use a short test with at least two separate features and enough height to show whether the behavior is stable. Print two unchanged baselines. Photograph the strands, walls, and first extrusion after each travel landing. Then change one variable and print two more copies.
Define success before tuning
Accept a change only when repeated tests reduce unwanted TPU webs without restart gaps, thin walls, weak layers, rough extrusion, or a new feed problem. Save the project and label spool condition, temperature, travel choice, speed, and retraction for every sample.
Use limits: when fewer TPU strings still is not a passing result
- Loaded, snap-fit, impact, or fatigue part: a cooler nozzle or more retraction is not a valid fix if the first line after travel becomes thin, a layer opens, or the part loses its intended flex. Keep the exact TPU inside its supported process window, inspect every restart, and repeat the real load or cycle proof before releasing the part.
- Seal, sliding fit, cable guide, fan guard, sensor opening, or moving mechanism: elastic hairs can foul a face, passage, contact, or moving clearance. Define a residue, fit, and function limit. Pulling off visible webs does not prove the hidden passage or assembly is clean.
- Flexible foot, grip, bumper, hinge, or wearable feature: a clean two-tower coupon does not prove the real feature kept its dimensions, softness, recovery, or seam strength. Print the feature in its intended orientation and cycle it under the representative deflection and environment.
- Repeated clicking, buckling, missing restart lines, a growing hotend blob, material leaking above the heater block, smoke, or an electrical smell: stop the print and follow the printer or hotend maker's safe inspection procedure. More travel speed or retraction cannot correct a feed-path or hotend fault.
- Food-contact, medical, pressure, lifting, electrical-safety, or other consequence-heavy use: a clean stringing coupon is not proof of sanitation, pressure integrity, dielectric safety, load capacity, or compliance. Use an appropriate validated process, material, inspection plan, and safety factor.
Release sequence: first repeat the controlled two-tower test, then print the actual critical feature in its intended orientation. Let it cool, inspect every post-travel restart and hidden passage, verify the real mate, and run the representative load, flex, or exposure test. Repeat the job before treating the setting as stable.
If the supported TPU baseline prints clean continuous walls but the real model still forces many long open travels, stop escalating retraction. Reorient or split the model, shorten exposed travel, define a controlled cleanup step, or choose a process that better fits the geometry. The goal is a repeatable functional part, not a perfect tower purchased with weak restarts or unstable feeding.
Avoid these common failed fixes
- Changing heat, travel, speed, and retraction together: the cleaner result will not identify the working control.
- Maxing retraction immediately: this can trade strings for buckling, tangling, delayed flow, or a clog.
- Drying without a timeline or control: drying is evidence only when the same-spool repeat changes the symptom.
- Using one universal TPU number: hardness, formulation, extruder path, nozzle, and profile all matter.
- Judging only cleanup effort: easy-to-remove hairs can still accompany weak restarts or unstable walls.
- Ignoring model layout: a geometry-specific travel problem should not rewrite a stable global profile.
Choose the next troubleshooting branch
If the TPU became stringy specifically after exposure, use the TPU spool-condition branch. For one raised restart line, use the TPU seam guide. For a rough dragged roof rather than open-air webs, use the TPU rough-top guide. For separated first-layer roads, use the TPU first-layer-gap diagnostic.
Common questions
What should I change first for TPU stringing?
First classify the webs and restore the exact TPU baseline. If the spool worsened after exposure and extrusion is rough or poppy, test spool condition. If the spool is known-good and walls are clean, test nozzle heat and exposed travel before retraction.
Can wet TPU cause stringing?
Yes, especially when stringing arrives with popping, roughness, inconsistent gloss, or a clear exposure timeline. Prove it with the same slice and same spool before and after the filament maker's approved drying process.
Should I lower TPU temperature?
Only as one controlled test inside the filament maker's supported range. Keep the change only if webs fall while extrusion, interlayer bonding, surface quality, and restarts remain acceptable.
Should I increase TPU retraction?
Retraction belongs near the end. Make a small TPU-specific change from the supported baseline and reject any result that adds restart gaps, thin walls, buckling, tangling, or a clog.
Why does TPU string on one model but not another?
One model may force longer exposed travels, more separate islands, different minimum-layer behavior, or crossings near curled features. Compare the travel preview before changing a global material profile.