✎ This article/section is a stub — some half-sorted notes, not necessarily checked, not necessarily correct. Feel free to ignore, or tell me about it.

3D printing

Theory

Filament printing

Typical filaments

✎ This article/section is a stub — some half-sorted notes, not necessarily checked, not necessarily correct. Feel free to ignore, or tell me about it.

Filaments are thermoplastics, meaning they can be molten repeatedly.

The plastics typically used as filaments are any that have a good balance between forgiving to print, heating restrictions, properties of the result, price, and more.

PLA is probably the easiest balance, ABS and PETG are generally seen as the next two reasonable ones.

There are many more somewhat-specialized ones.

Temperature-wise, you care about three temperatures: melting temperature, the smoke temperature being comfortably higher, and the glass transition temperature (where it starts getting soft). In particularly the last is a fuzzy concept, so these temperatures are never really precise.

Temperature:

PLA is typically printed at 180–210C, and gets soft around 50-60C

note that you can anneal PLA to increase this (to maybe 70-85C(verify), or sometimes little better(verify) -- it depends on the PLA)

but note it's likely to shrink slightly, so not really an option for precise prints

you can't anneal ABS and PETG(verify)

there is also high-temperature PLA - which prints the same way, but the idea is that after heat treating it, will melt well above 100C

ABS is typically printed at 210-240C, and gets soft around 105C

PETG is typically printed at 220-250C, and gets soft around 80-90C(verify)

Smell:

PLA has a sweeter smell, though fairly subtle

ABS has a plasticky smell

PETG smells less

Strength

PLA is a little more brittle

ABS is somewhat flexible, so won't break easily.

PETG is more flexible and stronger than ABS

Life:

ABS, PLA, and PETG will all absorb moisture to differing degrees (PETG more than the other two)

which means they don't do great outdoor

moist unprinted filament makes for less clean printing results, because hydrolysis breaks polymer chains.

Try to store them dry, but note that this will not remove moisture already in there

to dry, gently heat up to a few hours in an oven capable of (steadyish) low temperatuers. You can buy filament driers too.

note that depending on the season, and the climate you live in, this may not be an issue at all

PLA is somewhat brittle

Printing issues

ABS is unlikely to clog the print head.

ABS shrinks a bit as it cools, which can lead to some parts shrinking faster than others and creating cracks

while it's still relatively forgiving, when this is an issue consider a warmer room, enclosing the print area, and keeping on the heated bed

PLA is less likely to stick

heating that surface helps, as do some more DIY solutions.

ABS is likelier to stick to the surface you print on.

PETG sticks well - but that also making it poor support structure, and can printing bed details more interesting

PLA expands as it melts, which can make for a finicky start (a drop of oil seems to help), though once it's going it should be fine

PLA has much less shrinkage as it cools, meaning you can do larger prints with less worry

PETG ingredients vary more per vendor, making planning a little harder

PETG is a little more likely to residue on the printhead and get transferred to your object

PETG needs a heated pad (verify)
PLA does not, but doesn't mind it
ABS (verify)

Environment:

PETG, ABS, and PLA can be reused, though doing this yourself requires more equipment

PLA can be reused
PLA is typically made of sugarcane or cornstarch and can be composted - though needs some heat for that, and won't do it in your garden composter or a landfill

PLA is usually fine just sitting outside, but in extreme

hygroscopic

http://www.protoparadigm.com/news-updates/the-difference-between-abs-and-pla-for-3d-printing/

https://all3dp.com/pla-filament-guide-3d-printer/

Further materials

Nylon

CPE

HIPS (High Impact Polystyrene) -

cheap, lightweight, solid

needs heated bed, and a warmer extruder (230-245 C), and can work better in a warm environment

needs ventilation

in theory it is dissolvable in d-Limonene so can be used as a support material; in practice there are reasons this may be impractical

PVA

PC

BVOH

Flexible

TPE - rubbery, but TPE is a very broad category so can mean various things

TPU - flexible, relatively stretchy

PolyFlex - not stretchy

FiberFlex - a little stretchy

Flexible is harder to print on generic extruders - this varies per design, and per filament (e.g. underextrusion due to the delay, and jamming if there's too much space within the extruder for it to go where it shouldn't)

https://www.youtube.com/watch?v=7wMKp6q9ktE

Note that 'soluble, so useful as a support material' comes with footnotes. In particular, this is more true in 2-nozzle machines. In single nozzle you could do it with a pause, but only when the model does not occlude itself vertically (or can be oriented not to).

Imperfections

Overextrusion -

Threading -

https://www.matterhackers.com/articles/a-guide-to-understanding-the-tolerances-of-your-3d-printer

Filament quality

Filament tolerance is the amount of variation in thickness, i.e. variation between the material that comes out - which at typical tolerances is easily a few percent of volume.

There are differences between filaments, e.g. in layer adhesion,

https://all3dp.com/1/3d-printer-filament-types-3d-printing-3d-filament/

https://rigid.ink/blogs/news/156667655-the-12-best-3d-printer-filament-companies-the-ultimate-review

Tolerances

Most printers can reasonably print down to ~0.2mm,

https://www.matterhackers.com/articles/a-guide-to-understanding-the-tolerances-of-your-3d-printer

Practice

Print settings

Infill

Infill refers to a foam-like structure on the inside, basically to save plastic.

For parts that bear no particular load, down to 15% or so is fine.

100% infill might be stronger in theory, but won't help much when all stresses are on the outside and not so much what you're filling.

Also, also, 100% messes with tolerances due to possible bulging - even just going to 90% helps there.

But usually, adding more perimeters, with just medium infill, is a more reasonable compromise to strength, while still saving plastic.

Infill patterns matter somewhat to directional strength

https://www.youtube.com/watch?v=upELI0HmzHc

Support

Overhang refers to a layer extending beyond the one below it.

Some amount of angle is possible. Most prints can deal with 30-45 easily enough, but -- depending a little on layer thickness, material, speed, cooling, and other things -- you should not count on 45-60, and will rarely get more than.

Lower printing speed helps allow a little more overhang, but may negatively affect layer bonding, and thereby precision and strength.

support structures hold up higher-angle or just horizontal overhang.

They are essentially small-infill volumes under (and between) bits of real model, and with ideally small areas of attachment, to make them easier to remove later.

They can be automatically generated by slicers, so that it can be tuned to the print and printer.

They use very thin walls and are attached slightly(verify) so should snap off fairly easily. They may be structured in a zig-zag way so that once you have a start you can remove a larger volume.

There's always interesting bits, so they are sometimes more than a little extra work to remove, means more care to make surfaces smooth, and wastes some plastic, so people try to avoid support strucrures when possible, e.g. with different orientations, two-part prints, etc.

That said, support structures make many otherwise-unprintable shapes possible, so are a good solution.

Skirt, brim, and raft

Rafts are basically a thin layer below your you can remove later

reasons to add this include:

designs that don't stick so well

heated beds leading to the first layers tends to spread a bit more (elephant foot style), so useful if the dimensions of this side need to be precise

Skirt - those lines printed before and around the model, not touching it.

mostly to ensure that any variations from the extruder in the first seconds of use do not end up in the real print

also helps tell really early whether bed adhesion is decent

Brim - a skirt that does touch the part. This may help with bed adhesion, may avoid some warping, particularly when first layers are small/thin features, and also means well defined corners there won't round off as much (but you'll need to trim).

Ironing

Runs the hotend over the top without extruding.

This tends to make a smoother layer - though only applies to the very top layer(verify)

Setup

Bed level

You need the nozzle to be almost at the bed to print the first layer.

Fancier printers may detect this automatically, but a lot of them:

have a limit switch at almost the edge (to make it hard to crash the head)

let you manually shift the bed up

The typical test is to shift a piece of paper until the head starts grabbing it - but doesn't hold it completely. This means there is on the order of~0.05mm left under there.

Note that you want this initial distance, to be both there and well controlled. Too little means the first printed layer squishes, too much and it probably won't adhere well

You should do this at at least the four corners, to get it decent. Ideally twice, ideally with a test at the middle, but you can't get it fully ideal because most beds surfaces are not that perfectly flat. Nor need to be, usually.

Pragmatics

Bed adhesion / first layers

The first layer not sticking at all may mean a print never gets started.

A bed only partially sticking makes it likely that non-sticking parts will warp.

Depends a bunch on bed material.

Relatively rough beds may stick a bunch, to the point you may like masking tape to make it easier to get off.

Glass, which give smoother edges, may need some adhesive. There are many solutions to this, from gluesticks to hairspray. (In most cases you'll also want some isopropyl or similar for cleanup)

Note that ABS is harder, may not stick as well and warps a little more easily, so generally has some rules of its own.

This (and small-footprint prints) are some reasons people use rafts (see above)

Changing filament

Your printer may have feature to do this.

In practical terms, you heat the nozzle to the inserted filament's melting temperature, and you pull it out.

Note that you may get less residue in the nozzle if you push some out through the nozzle before pulling.

After installing, it helps to run a little filament to clean up what's left.

Some printers allow you to do this while printing. This is not a mechanical features, but a practical one.

At the very least you want it to pause, but preferably it should also extrude a little extra to not gap.

Printing for strength

Speed - there is both a too fast and too slow (verify)

Software

Modeling

Simple, but basic

Tinkercad - simple, mostly combining shapes, and a few tools

3D Slash

MatterControl -

Sketchup - decent balance between simple and capable, particularly ly for something free

FreeCAD - parametric

ObShape -

Fusion360 (free for personal use)

Blender -

123Design -

Artistic

Meshmixer

Scultrix

Zbrush

https://3dprinting.com/software/

Slicing

Modeling gives you a shape. You still need to figure out exactly what to send to the printer, though, because the limitations and options of the specific printering technology, and of the specific printer, varies.

In theory this could be integrated into modelling programs, but it's more flexible to have a specialized tool for this. This also means more diverse 3D programs can be used, because all it takes is exporting a generic file format to a slicer.

Cura

Slic3r

Additions and alternatives

Glass bed

Pads may stick too well, and vary in height over time due e.g. to adhesive underneath.

Glass is flat, which can be nicer for the bottom layer of your print aesthetically, and also make it easier to pop things off.

(it does change how bed adhesion works, of course)

Octopi

Image for raspberry Pi

Manages prints, has a mobile interface, and allows webcam monitoring.

Heated bed

Hacking

G-code notes

Startup

Print

Finish

Pushing the boundaries

Slicers compiling for a specific printer and material is the easiest way to get predictable results.

But what if you want less predictable?

Or more parametric?

See e.g.

Paul O'Dowd's experiments

https://projectsilkworm.com/examples/

http://www.gierad.com/projects/furbrication/

CNC notes

On overall size

On accuracy

On the drilling tool

On kits

Conversions

3D printing and CNC notes