# 3D Printing Problems Solutions: An Easy Test and Proven Fix for 14 Failures

- Link: https://additiverealm.com/3d-printing-problems-solutions/
- Published: 2025-10-20T21:24:18+02:00
- Author: TreiD

* [3D Printing Problems Solutions: The 5-Minute Triage Before You Change a Setting](https://additiverealm.com/3d-printing-problems-solutions/#3d-printing-problems-solutions-the-5-minute-triage-before-you-change-a)
 * [How to Diagnose Any Failed Print: Symptom, Likely Cause, One Test, One Change](https://additiverealm.com/3d-printing-problems-solutions/#how-to-diagnose-any-failed-print-symptom-likely-cause-one-test-one-cha)
 * [Failures in the First Few Layers: First Layer, Bed Adhesion and Warping](https://additiverealm.com/3d-printing-problems-solutions/#failures-in-the-first-few-layers-first-layer-bed-adhesion-and-warping)
 * [Too Much or Too Little Plastic: Under Extrusion, Over Extrusion and Stringing](https://additiverealm.com/3d-printing-problems-solutions/#too-much-or-too-little-plastic-under-extrusion-over-extrusion-and-stri)
 * [Parts That Split or Snap: Fixing Layer Adhesion](https://additiverealm.com/3d-printing-problems-solutions/#parts-that-split-or-snap-fixing-layer-adhesion)
 * [Parts That Don’t Fit: Dimensional Accuracy, Shrinkage and Hole Compensation](https://additiverealm.com/3d-printing-problems-solutions/#parts-that-donand8217-t-fit-dimensional-accuracy-shrinkage-and-hole-co)
 * [Prints That Work but Look Rough: Ringing, Z-Seams and Surface Finish](https://additiverealm.com/3d-printing-problems-solutions/#prints-that-work-but-look-rough-ringing-z-seams-and-surface-finish)
 * [Material Baselines and the Maintenance That Stops Repeat Failures](https://additiverealm.com/3d-printing-problems-solutions/#material-baselines-and-the-maintenance-that-stops-repeat-failures)
 * [Frequently Asked Questions](https://additiverealm.com/3d-printing-problems-solutions/#frequently-asked-questions)
 * [Sources](https://additiverealm.com/3d-printing-problems-solutions/#sources)

Most failed prints trace back to three causes, so the fastest 3D printing problems
solutions start with three checks, in this order: the Z-offset, wet filament, then
extruder steps or flow. I check the Z-offset first because one small print tests
it: a single-layer square, nudged in 0.02 mm steps until the lines just merge.

Then dry the spool if the nozzle hisses or pops, and command 100 mm of extrusion
to confirm the extruder pulls exactly 100 mm. Every other failure below is sorted
by symptom, with one test and one first change each.

## 3D Printing Problems Solutions: The 5-Minute Triage Before You Change a Setting

Before changing a slicer value, run a few physical checks. They come before the 
three calibration checks above, because a bad spool, a snag in the filament path
or a greasy plate can spoil any of those results. If one fails, fix it and reprint
before you read further.

Start with the filament. Confirm the material and its temperature range on the spool
label, and check it was stored sealed with desiccant. Measure the diameter in three
spots with calipers; if the readings wander, enter the measured average in the slicer
or replace the spool. Then spin the spool by hand: it should turn freely, with nothing
snagging the path to the extruder.

Next, the hardware. Confirm the nozzle size in the slicer matches the one fitted,
check that the Bowden couplers or the direct-drive tension hold the filament without
slipping, and tension the belts. If the nozzle is partly blocked, [MatterHackers’ troubleshooting guide](https://www.matterhackers.com/articles/3d-printer-troubleshooting-guide)
recommends a nylon cold pull or clearing it with a small needle.

Re-run leveling or the mesh instead of assuming it’s fine, check the Z-offset with
a first-layer test print, and wipe the plate with isopropyl alcohol. Close doors
and keep the printer away from air vents, since moving air cools one side of the
part faster than the other.

Last, a slicer sanity check: set the filament diameter to what you measured, and
leave line width, layer height and flow at the profile defaults until you calibrate.

## How to Diagnose Any Failed Print: Symptom, Likely Cause, One Test, One Change

Each section below follows the same pattern: name the symptom, rank the likely causes,
run one test, make one change, then lock the fix in. Change one thing, re-test, 
and commit only when the improvement repeats. Read the failure before you scrape
it off: note where it broke, which layer it started on and whether it repeats.

Find your symptom in the table, run the one test, then make the first change. The
section in brackets has the full fix.

| Symptom | One test | First change | 
| Round, separate first-layer lines | Single-layer square | Z-offset down 0.02 mm at a time (first layer) | 
| Corners lift after a good first layer | Watch the first layers | Enclosure or draft shield (warping) | 
| Gaps, thin walls | 100 mm extrusion test | E-steps or a cold pull (under extrusion) | 
| Blobs, walls too thick | Profile check, then flow cube | Lower flow multiplier (over extrusion) | 
| Hairlike strings | Retraction tower | Retraction only (stringing) | 
| Snaps along layer lines | Flex a small block | Hotter nozzle, less fan (layer adhesion) | 
| Holes the wrong size | Cube plus hole coupon | Hole compensation by the measured error (fit) | 
| Part offset from one layer up | Belts and grub screws | Tighten, lower acceleration (layer shifts) | 
| Echoes after sharp features | Two acceleration values | Input shaping (surface finish) | 
| Scar where layers start | Check seam placement | Seam to rear (surface finish) | 
| Drooping overhangs, sagging bridges | Overhang test print | More part cooling (surface finish) | 
| Bumps or gaps in the top surface | Check top layer count | More top layers (surface finish) |

## Failures in the First Few Layers: First Layer, Bed Adhesion and Warping

Uneven squish, gaps between lines, lines that don’t touch and elephant’s foot all
point at the first layer. Sort those out before you touch anything else.

![Failed print on a build plate with detached parts and tangled stringy filament](
https://additiverealm.com/wp-content/uploads/2025/10/3d-printing-problems-solutions-
first-layer-1024x935.webp "3D Printing Problems Solutions: An Easy Test and Proven
Fix for 14 Failures 1")

Check the Z-offset first. Print a single-layer square: round, separate lines mean
the nozzle is too high, while ridges or a rough, scraped surface mean it’s too low.
Move the Z-offset in 0.02 mm steps until the square looks right, and only then touch
anything else.

If the lines still don’t stick, slow the first layer down, which MatterHackers lists
as a fix for poor bed adhesion. Keep first-layer extrusion at 100% or less: the 
same guide ties elephant’s foot to first-layer extrusion above 100%, as well as 
to a first layer set too low. Leave the part-cooling fan off at the start, too: 
[Prusa’s cooling guide](https://help.prusa3d.com/article/cooling_127569) says many
filaments print better with no cooling for the first few layers, typically 1–5.

Match the plate to the filament. [Prusa’s PETG guide](https://help.prusa3d.com/article/petg_2059)
recommends a textured, powder-coated sheet for PETG, because it can bond hard enough
to damage smooth PEI; if you must use smooth PEI, apply a glue stick first. Adhesives
such as [Magigoo](https://magigoo.com/) come in grades per material, with separate
ones for PLA, PETG and ABS, for nylon and for polycarbonate. Magigoo’s instructions
for the Original grade are to reapply it for long or difficult prints, and before
every print on flex sheets.

If the first layer looks right but the part still lets go mid-print, check the bed
temperature range on the filament label, and let the plate preheat fully. Only once
the temperature is right does a brim, mouse ears at sharp corners, or a raft earn
its place by adding contact area.

Warping comes from a thermal gradient: the upper layers cool and shrink while the
bed holds the bottom warm, so the edges pull up and the corners lift. Watch the 
first layers: if corners lift after a good first layer, you have warping, not a 
Z-offset problem. The first change is to stop the air moving, with an enclosure 
or at least a slicer draft shield and a closed door.

For ABS, [Prusa’s ABS guide](https://help.prusa3d.com/article/abs_2058) gives a 
bed range of 80–110 °C, with larger objects needing the top of it, and calls for
an enclosure or a draft shield. Prusa also notes that ASA warps less than ABS and
that PETG has very little thermal expansion, so a material switch can be the fix
for large, flat parts. In CAD, fillets on sharp corners remove the points where 
lifting starts.

![3d printing problems solutions featured](https://additiverealm.com/wp-content/
uploads/2025/10/3d-printing-problems-solutions-featured-947x1024.webp "3D Printing
Problems Solutions: An Easy Test and Proven Fix for 14 Failures 2")

## Too Much or Too Little Plastic: Under Extrusion, Over Extrusion and Stringing

Gaps, thin walls and missing layers mean the nozzle isn’t getting enough plastic.
MatterHackers lists three usual causes: the wrong filament size in the printer settings,
an extrusion temperature that’s too low, and a clogged or uncalibrated nozzle or
extruder.

To test for under extrusion, heat the hotend to printing temperature and mark the
filament 120 mm above the extruder entry. Command 100 mm of extrusion, then measure
from the entry to the mark: you should have 20 mm left. More than 20 mm means the
extruder is slipping or miscalibrated. Exactly 20 mm with gaps still in the print
points at the nozzle, the temperature or the speed.

A short pull means calibrate E-steps (rotation distance on Klipper) with our [extruder calibration guide](https://additiverealm.com/extruder-calibration/),
then fine-tune flow. A full pull with gaps points to a partial clog, so do a cold
pull or swap the nozzle before raising temperature or cutting speed.

**Clogs and heat creep.** If extrusion stops mid-print or the extruder starts clicking,
heat the hotend to printing temperature and push filament through by hand. If it
won’t flow under firm pressure, the first fix is a cold pull. If it clears but jams
again partway through prints, suspect heat creep: check that the heatsink fan spins
whenever the hotend is hot, and that the PTFE tube is fully seated with no gap.

![3d printing problems solutions filament skipping](https://additiverealm.com/wp-
content/uploads/2025/10/3d-printing-problems-solutions-filament-skipping-936x1024.
webp "3D Printing Problems Solutions: An Easy Test and Proven Fix for 14 Failures
3")

![3d printing problems solutions - underextrusion](https://additiverealm.com/wp-
content/uploads/2025/10/3d-printing-problems-solutions-underxtrusion-1024x934.webp"
3D Printing Problems Solutions: An Easy Test and Proven Fix for 14 Failures 4")

### Over Extrusion: Blobs, Zits and Walls Too Thick

Blobs, zits and walls that measure too thick mean too much plastic. Often the cause
is a wrong number in the profile, so the first test needs no printing: open the 
profile and check flow, filament diameter and line width before anything else.

If the profile is clean, print a single-wall flow cube, measure the wall at 3–4 
points with calipers and average the readings. The new multiplier is current multiplier
× (target wall width ÷ measured wall width): at 1.00 with a 0.45 mm target and a
0.47 mm average, that gives 0.96. Closed printers that don’t expose E-steps leave
this slicer flow calibration as the whole job.

Pressure advance (linear advance in Marlin) tames blobs at corners and line ends,
and elephant’s foot compensation handles the squished bottom. Save the calibrated
flow in each filament profile, so the fix survives the next spool change.

### Stringing: Retraction, Temperature and Travel

Hairlike strands between features are stringing. The big levers are retraction, 
temperature and travel, and wet filament makes all three worse. If you hear hissing
or popping from the nozzle, dry the spool before you touch retraction.

![3D printer nozzle with a thin green filament string dangling from the tip, a typical
sign of stringing and oozing](https://additiverealm.com/wp-content/uploads/2026/
10/additiverealm-534fcc26-576x1024.jpg "3D Printing Problems Solutions: An Easy 
Test and Proven Fix for 14 Failures 5")

The test is a retraction tower, and our [Retraction 101 setup](https://additiverealm.com/retraction-101/)
walks through it. [Simplify3D’s stringing guide](https://www.simplify3d.com/resources/print-quality-troubleshooting/stringing-or-oozing/)
puts most direct-drive extruders at 0.5–2.0 mm of retraction, says some Bowden setups
need as much as 15 mm, and finds retraction speed usually works best between 20 
and 100 mm/s. Run the tower inside that window and change retraction first, nothing
else.

If strings survive, lower the nozzle temperature: [NinjaTek’s troubleshooting guide](https://ninjatek.com/support/troubleshooting-guide/)
suggests 5 °C steps, and Simplify3D 5–10 degrees. Then raise travel speed so the
nozzle spends less time over open air, enable combing or avoid crossing perimeters,
and add wipe and coast. A worn or crusted nozzle oozes too, so clean or replace 
it if the strings never fully stop.

## Parts That Split or Snap: Fixing Layer Adhesion

If layers split, or a part snaps cleanly along a layer line when you bend it, the
layers never fused. A quick test is to print a small block and flex it.

The first change is heat. Raise the nozzle in 5 °C steps, staying inside the spool’s
range, because each new layer has to remelt the top of the one below it to bond.
If that’s not enough, reduce or turn off the fan for ABS and nylon; for PETG, Prusa
runs the fan off for the first few layers and at half power after, or fully off 
for the toughest parts. Wet filament also weakens layers, so dry it before you chase
temperature any further.

If the outer wall peels away from the inner ones, NinjaTek’s fix is to slow the 
outer wall to about the bottom-layer speed, then raise flow 5% at a time.

If the layers are fused but the part is still too weak, add walls before anything
else: [Prusa’s layers and perimeters guide](https://help.prusa3d.com/article/layers-and-perimeters_1748)
says a model’s strength comes mostly from its perimeters, not its infill. For infill,
gyroid is a good default, since [Prusa’s infill patterns guide](https://help.prusa3d.com/article/infill-patterns_177130)
describes it as equally strong in all directions and fairly quick to print. A part
is always weakest across its layer lines, so orient it so the load runs along them.
For high-temperature materials, an enclosure keeps the layer below warm enough to
bond.

Annealing is the last lever, and it only suits some materials. In [Prusa’s annealing tests](https://blog.prusa3d.com/how-to-improve-your-3d-prints-with-annealing_31088/),
PLA was strongest after annealing at 90 °C and above, PETG improved above 110 °C,
and ABS and ASA were classed as unsuitable. Expect the part to change size: the 
samples shrank most along their longest axis and grew in Z.

![3d printing problems solutions - layer adhesion issues](https://additiverealm.
com/wp-content/uploads/2025/10/3d-printing-problems-solutions-nozzle-clog-880x1024.
webp "3D Printing Problems Solutions: An Easy Test and Proven Fix for 14 Failures
6")

## Parts That Don’t Fit: Dimensional Accuracy, Shrinkage and Hole Compensation

The test is a calibration cube plus a coupon with the hole sizes you actually use,
measured in X, Y and Z with calipers; our [3D printer resolution guide](https://additiverealm.com/3d-printer-resolution-guide/)
explains why those numbers matter. If an axis is off by the same percentage at every
size, so the error grows with the part, check steps per mm (rotation distance on
Klipper) and that the pulley and belt pitch match the firmware.

A slack belt causes backlash instead, which shows up as a roughly fixed offset or
rounded, uneven corners, so tension the belts when you see that. If outer walls 
come out oversize by a fixed amount regardless of part size, check flow and line
width.

Material shrinkage also scales with the part: Prusa puts ABS shrinkage at usually
around 1–2%, so a large ABS part can measure small on a well-calibrated printer.
Scale the model for that material instead of changing steps per mm.

The hole coupon catches the most common misfit: holes that come out the wrong size
while the cube measures true. The first change is hole compensation in the slicer,
set from the coupon: measure the hole and compensate by the difference.

Two PLA studies show why one fixed number works. In an [Atlantis Press study of hole shrinkage in PLA](https://www.atlantis-press.com/article/25883323.pdf),
10 mm holes printed 0.20–0.50 mm undersize, layer thickness moved the error most
of the four settings tested, and 0.36 mm of compensation brought one to 9.96 mm.
In an [ASEE study of 3D-printed PLA accuracy](https://peer.asee.org/preliminary-investigation-of-dimensional-accuracy-of-3d-printed-pla-a-project-based-learning-experience-wip.pdf)
on a MakerBot and a Prusa, holes came out 0.5–0.9 mm oversize, mostly as a fixed
offset, and modelling them 0.9 mm smaller cut the error to 0.007–0.128 mm.

The two studies erred in opposite directions, which is why the coupon comes before
the setting. Holes shrink as the plastic cools, and a [study of FDM hole shrinkage compensation](https://link.springer.com/article/10.1007/s00170-017-1018-2)
corrects it through the part’s interior structure instead of plain scaling.

A clearance parameter on every hole means one edit updates the whole model instead
of a slicer tweak per print; the method is covered in [parametric design in Fusion 360](https://additiverealm.com/parametric-design-in-fusion-360/).

### Layer Shifts: Belts, Pulleys and Collisions

If the whole part steps sideways at one layer and carries on offset, that’s a layer
shift, not a calibration error. Check belt tension, the grub screws on the motor
pulleys, and whether the nozzle hit curled-up plastic at that height. A pulley that
slips now and then leaves walls that wander back and forth instead of one clean 
step. The first fix is to tighten the loose belt or pulley; if nothing is loose,
lower acceleration.

![Two translucent printed blocks side by side: the left with even layers, the right
with wavy, shifted walls from a pulley problem](https://additiverealm.com/wp-content/
uploads/2026/10/3d-printing-problems-solutions-7-1024x619.jpg "3D Printing Problems
Solutions: An Easy Test and Proven Fix for 14 Failures 7")

“troubles in pulley-town” by TalpaDK is licensed under CC BY-SA 2.0. To view a copy
of this license, visit https://creativecommons.org/licenses/by-sa/2.0/.

## Prints That Work but Look Rough: Ringing, Z-Seams and Surface Finish

Once the printer is reliable, chase polish. Ringing, or ghosting, shows up as faint
echoes of a sharp feature repeating along the wall just after it, and it gets worse
as speed and acceleration rise. The test is to print the same part at two acceleration
values: if the echoes fade at the lower one, it’s ringing. The first change is [input shaping](https://additiverealm.com/marlin-2-1-2-input-shaping)
if your firmware has it; otherwise lower acceleration and jerk.

On Klipper, input shaping and pressure advance are configuration settings in printer.
cfg. On Marlin 2.1.2 and later, input shaping and linear advance must be enabled
in Configuration_adv.h and the firmware recompiled and reflashed, unless your printer’s
vendor build already ships with them turned on. Our [Klipper, OctoPrint and Marlin comparison](https://additiverealm.com/klipper-vs-octoprint-vs-marlin/)
covers where each one lives.

A Z-seam is the scar where every layer starts and ends. Set it to rear, or to the
sharpest corner so the geometry hides it, and use wipe and coast to shrink the blob.
Newer slicers also offer scarf seams, which taper the start and end of each loop.
If the seam still bulges, pressure advance is the next change, not temperature.

To reduce visible layer lines, drop the layer height, or use variable layer height
so only the curves get thin layers. Prusa keeps layer height below 80% of the nozzle
diameter, about 0.32 mm on a 0.4 mm nozzle, and generally not under 0.10 mm. For
finer detail, change the nozzle: [Prusa’s mixed-nozzle example](https://help.prusa3d.com/article/experimental-printing-with-different-nozzle-diameters_821176)
prints perimeters with a 0.25 mm nozzle for a detailed surface and infill with a
0.6 mm one.

**Overhangs and bridges** that droop or sag mean the plastic stays soft too long
before it sets. The test is an overhang test print, which shows the angle where 
the surface starts to curl; MatterHackers advises keeping overhangs at 45° or less.
The first change is more part cooling, which PrusaSlicer sets separately for bridges
and overhangs, then a slower bridge speed if bridges still sag. On small parts, 
a minimum layer time slows each layer so it can cool before the next one lands.

**Pillowing and gaps in top layers**, meaning bumps, holes or a quilted look on 
a flat top surface, show the top skin has too little support underneath it. The 
test needs no printing: check the top layer count in your profile. The first change
is to add top layers, or raise the top thickness so the slicer adds them for you
at any layer height.

## Material Baselines and the Maintenance That Stops Repeat Failures

Keep per-material starting points and adjust from there, one variable at a time.
These are Prusa’s recommended values; where your brand’s label differs, the label
wins:

 * PLA: 215 °C nozzle on the first layer and 210 °C after, 60 °C bed ([Prusa’s PLA guide](https://help.prusa3d.com/article/pla_2062)).
   The Atlantis Press paper gives 190–230 °C as the usual PLA range.
 * PETG: 230 °C on the first layer and 240 °C after, 85 °C bed then 90 °C, fan off
   for the first few layers and half power after.
 * ABS: 255 °C nozzle, 100 °C bed (80–110 °C by part size), an enclosure or a draft
   shield, and about 1–2% shrinkage. ASA warps less.
 * Nylon (PA): 285 °C nozzle, 110 °C bed and an enclosure for pure PA, which warps;
   carbon-fiber PA warps far less ([Prusa’s polyamide guide](https://help.prusa3d.com/article/polyamide-nylon_167188)).
 * TPU and other flexibles: 230–245 °C nozzle, 60–75 °C bed, around 20 mm/s and 
   no more than 30–40 mm/s, with retraction reduced or off ([Prusa’s flexible materials guide](https://help.prusa3d.com/article/flexible-materials_2057)).

Dry filament before you chase settings, without overheating it: [Prusa’s drying guide](https://help.prusa3d.com/article/drying-filament_332086)
lists PLA at 45 °C for 6 hours, PETG at 55 °C for 6 hours, TPU at 60 °C for 4–6 
hours and ASA at 80 °C for 4 hours. For nylon, Prusa says at least 4 hours below
90 °C. Go hotter and the filament softens and sticks together on the spool.

Carbon-fiber blends chew through brass nozzles. A worn nozzle shows up as under 
extrusion and stringing long before it looks worn, which sends people chasing the
wrong settings. Our [hardened steel nozzle guide](https://additiverealm.com/hardened-steel-nozzle-carbon-fiber-filament/)
compares four materials for that job.

Clean the nozzle and bed regularly, revalidate the mesh, store build plates flat
and keep filament sealed with desiccant. Belts, rail lubrication and nozzle and 
PTFE replacement go by print hours, and our [3D printer maintenance schedule](https://additiverealm.com/3d-printer-maintenance-schedule-print-hours/)
lists the jobs at 50, 200 and 500+ hours. Keep the firmware current, and if temperature
fixes stop behaving, check the hotend reading against a separate thermometer.

The bench kit is short: digital calipers, isopropyl alcohol, a brass brush and nozzle
cleaning needles. The test prints are a temperature tower, a retraction test, a 
flow cube, an XYZ calibration cube and overhang and bridge tests; download them 
from Printables or Thingiverse and keep using the same files, so results compare
across spools.

When the next print fails, I’d enter these 3D printing problems solutions at the
symptom instead of reading them top to bottom. My rule is short: run one test, make
one change, and write down what worked. Change two things at once and you can’t 
tell which one fixed it.

## Frequently Asked Questions

**What should I check first when a print fails?** Run the 5-minute triage: filament,
hardware, bed, slicer and drafts. Then check the Z-offset, dry the spool and run
a 100 mm extrusion test, in that order.

**How do I tell under extrusion from over extrusion?** Under extrusion shows gaps,
thin walls or missing layers; over extrusion shows blobs, zits and walls that measure
too thick. A single-wall flow cube measured with calipers settles it.

**Why does ABS warp, and how do I stop it?** ABS shrinks as it cools, around 1–2%
by Prusa’s figure, so the cooling upper layers pull the corners up. Keep the bed
around 100 °C, print in an enclosure or with a draft shield, keep the fan off for
the first layers, and add a brim.

**How often should I clean or replace my nozzle?** Clean it whenever extrusion or
surface quality drops, and replace it when clogs keep coming back or the orifice
is worn; the maintenance schedule above sets the intervals by print hours. Carbon-
fiber and other abrasive filaments wear brass nozzles much faster.

**Do I need different settings for each brand of filament?** Yes. Start from the
material baselines above, print a temperature tower and a retraction tower for each
new brand, and save a calibrated profile for every filament you use.

## Sources

 * [MatterHackers: 3D Printer Troubleshooting Guide](https://www.matterhackers.com/articles/3d-printer-troubleshooting-guide)
 * [Prusa Knowledge Base: Cooling](https://help.prusa3d.com/article/cooling_127569)
 * [Prusa Knowledge Base: PETG](https://help.prusa3d.com/article/petg_2059)
 * [Magigoo: 3D printing bed adhesives by filament](https://magigoo.com/)
 * [Prusa Knowledge Base: ABS](https://help.prusa3d.com/article/abs_2058)
 * [Simplify3D: Print Quality Troubleshooting, Stringing or Oozing](https://www.simplify3d.com/resources/print-quality-troubleshooting/stringing-or-oozing/)
 * [NinjaTek: Troubleshooting Guide](https://ninjatek.com/support/troubleshooting-guide/)
 * [Prusa Knowledge Base: Layers and perimeters](https://help.prusa3d.com/article/layers-and-perimeters_1748)
 * [Prusa Knowledge Base: Infill patterns](https://help.prusa3d.com/article/infill-patterns_177130)
 * [Prusa Blog: How to improve your 3D prints with annealing](https://blog.prusa3d.com/how-to-improve-your-3d-prints-with-annealing_31088/)
 * [Atlantis Press (ICMIA 2017): Research on the Shrinkage of Model with Hole in PLA Material Based on the FDM 3D Printing](https://www.atlantis-press.com/article/25883323.pdf)
 * [ASEE 2024 Annual Conference: Preliminary Investigation of Dimensional Accuracy of 3D-Printed PLA](https://peer.asee.org/preliminary-investigation-of-dimensional-accuracy-of-3d-printed-pla-a-project-based-learning-experience-wip.pdf)
 * [Springer, International Journal of Advanced Manufacturing Technology: Shrinkage compensation of holes via shrinkage of interior structure in FDM process](https://link.springer.com/article/10.1007/s00170-017-1018-2)
 * [Prusa Knowledge Base: Experimental printing with different nozzle diameters](https://help.prusa3d.com/article/experimental-printing-with-different-nozzle-diameters_821176)
 * [Prusa Knowledge Base: PLA](https://help.prusa3d.com/article/pla_2062)
 * [Prusa Knowledge Base: Polyamide (Nylon)](https://help.prusa3d.com/article/polyamide-nylon_167188)
 * [Prusa Knowledge Base: Flexible materials](https://help.prusa3d.com/article/flexible-materials_2057)
 * [Prusa Knowledge Base: Drying filament](https://help.prusa3d.com/article/drying-filament_332086)