.4mm filament

My question is: What is the smallest feasible filament size for 3d printing? There has to be a size that's practically possible, yet still cost effective. It's gone from 3mm to 1.75mm, but could it go lower? And if it could, like Moore's law predicts, how small could it be produced? Could we one day be printing with .4mm filament through a .4mm nozzle? Consider your mind blown.

Never happen. Time is money.

It takes less time to extrude a kilo of 3mm filament than a kilo of 1.75mm filament.

But if you have a hopper on your extruder full of pellets...you're on to something.

Mind not blown.

i have in the past produced a few hotends for people wanting to use 1mm but this was filament they were making themselves

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Consider the increasingly small size of microchips. The same thing is possible with filament.

Flash forward 20 years. .05mm filament through a .05mm nozzle?

Edited 1 time(s). Last edit at 02/13/2014 01:03AM by ohioplastics.

make a roll for 40$ a kilogram, that is stable in size and tolerant to +/- 5%. I'll buy it from you.

I think that it may be true about miniaturization of electronics to performance and costs, but i do not see the same holding true for large scale items that require the same amount of material. Although there would be incentive for this as it would likely quadruple the amount of prints done in the same amount of time now if motion systems can keep up, using those cheaper electronics....

what are you going to do about the lack of back pressure and die swell to contain the hot zone?

Quote
jamesdanielv
make a roll for 40$ a kilogram, that is stable in size and tolerant to +/- 5%. I'll buy it from you.

I could sell a ketchup popsicle to a man in white gloves. I'll sell you a kilo of fishing line for $40. You've got me on the nozzle swell though. Maybe .6mm filament through a .4mm nozzle.

I think there should be logic behind things, with microchips there is a use for making it smaller, but with smaller filament you will need to extrude faster to keep pressure in the nozzle.

Printing 0.4mm filament through a 0.4mm nozzle will never work unless you develop a special pressure system that works with pressure sensors and extrusion speed.

For me taller nozzle size is more interesting, I like speedy prints, bigger prints, using more material, making more useful, more 'human scale' objects. Going smaller can't be applied to everything. For example, in the Netherlands, houses have become bigger, doors higher, etcetera, people are getting bigger. However, the parts that make up the door, and controllers that run the machines that produce the doors, have become smaller and more efficient, but I see no point efficiency-wise to make filament smaller no.

Fat nozzles are great. Having .7mm layers with a 1mm nozzle is a wonder to behold as it prints. The downside is that you have to print more slowly; that much plastic takes a while to cool, even with fans.

But you have to use 3mm filament. The 1.75mm moves through the hotend so quickly that it doesn't have time to melt properly.

Quote
DonaldJ
Fat nozzles are great. Having .7mm layers with a 1mm nozzle is a wonder to behold as it prints. The downside is that you have to print more slowly; that much plastic takes a while to cool, even with fans.

But you have to use 3mm filament. The 1.75mm moves through the hotend so quickly that it doesn't have time to melt properly.

I can't imagine how this would be better if you have a few pictures or video please post other than faster I don't see any advantage and it has to be way less accurate

Quote
ohioplastics
My question is: What is the smallest feasible filament size for 3d printing? There has to be a size that's practically possible, yet still cost effective. It's gone from 3mm to 1.75mm, but could it go lower? And if it could, like Moore's law predicts, how small could it be produced? Could we one day be printing with .4mm filament through a .4mm nozzle? Consider your mind blown.

How are you going to drive it? Stiffness scales with the 4th power of diameter. 0.4mm filament will buckle in a stiff breeze. Unless you invent a way to break the laws of physics, moore's law doesn't apply here.

Mind not blown.

Edited 1 time(s). Last edit at 02/13/2014 10:49AM by greenman100.

Since the pressure would be zero or near zero, there shouldn't be any breakage (theoretically). I know it's not possible now, but neither was a smartphone 30 years ago.

Quote
ohioplastics
Since the pressure would be zero or near zero, there shouldn't be any breakage (theoretically). I know it's not possible now, but neither was a smartphone 30 years ago.

What? I think you are mixing up force and pressure.

You still have all the viscous drag that you always have. You have less turbulent flow as you are no longer performing a crossectional area reduction, but you still have all the viscous drag.

What is the make or break of a hot end, Mr. Hotend Manufacturer? Why do people bother with PTFE liners and heat breaks?

Viscous drag.

Smartphones didn't have to break laws of physics in order to become possible, we just had to develop ways of making things (transistors and antennas, mostly) smaller. ABS, PLA, and polymers in general have very low buckling resistance. That's why bridges aren't made out of plastic.

Edited 2 time(s). Last edit at 02/13/2014 11:19AM by greenman100.

Quote
greenman100
ABS, PLA, and polymers in general have very low buckling resistance. That's why bridges aren't made out of plastic.

I never said anything about ABS or PLA. Think outside the box.

There are very few materials i can think of that have sufficent stability to be used with a diameter of 0.4mm.
I'm still not sure what the idea behind reducing the diameter to this degree is. Thinking outside the box is all nice and well, but you should have at least an inkling of an idea of what would be the use, otherwise it's not even a mental exercise let alone any kind of mind blow.
At this point i see more problems popping up with smaller diameters than advantages.
Problems you get:

• reduced number of possible materials
• prints take longer
• the stress during transport of the filament is larger -> higher chance of failure/buckling
• higher precision in manufacturing is necessary -> higher costs

Advantages

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

the filament also has a lot more surface area to absorb water out of the air, or possibly even other stuff from the air that could degrade the quality of the plastic.

But consider this; why didn't dot-matrix printers follow moores law, with a print head with finer and finer pins? it's because laser and inkjet printers were vastly superior. it's the same for SLS; as soon as those patents expire, everyone that sells 3d printers is going to race to put out a SLS printer. I guarantee it.

So printing Thicker is likely where FDM will stay relevant. You can print larger, stronger items quickly, while sacrificing surface finish and feature size.

I think the advantage would be near unlimited printing speed. The ability to produce a part faster than an injection mold.

Quote
ohioplastics
I think the advantage would be near unlimited printing speed. The ability to produce a part faster than an injection mold.

there is no way would even be 100th the speed of the production injection molding machine and as far as quality it would be lucky to be 100th of that to

Quote
cnc dick
I can't imagine how this would be better if you have a few pictures or video please post other than faster I don't see any advantage and it has to be way less accurate

.5mm nozzle, .35mm layer & 2 perimeters compared with 1mm nozzle, .7mm layer & 1 perimeter. I don't recall, exactly, the speed (it's been a while) but it was the same for both. ABS. But I remember the .5mm nozzle took almost exactly four times longer to print.

Currently the limit for reprap like machines isn't the extrusion speed but the speed of the rest of the machine. The mechanics just don't hold up to the accelerations needed. Things start to vibrate and even rip apart long before the extrusion speed you need to get anywhere near mold injection is reached. Smaller filament makes things slower not faster.
The advantage of 1.75mm filament to 3mm is the lower force needed to push the filament to achieve the same force /²surface. This does not simply scale linear though.To get any additional advantage you would need to go for (much) smaller nozzles, which again would increase print time and limit the acceleration of the rest of the machine.
As was pointed out, the way forward to more speed and finer structures isn't fdm but sls or similar techniques. One option i could think of having an interesting future are piezoelectric line printheads that use UV curable fluids or some kind of two component ink.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Quote
Srek
Currently the limit for reprap like machines isn't the extrusion speed but the speed of the rest of the machine. The mechanics just don't hold up to the accelerations needed.

This is only true with Mendel variant printers which is why I am always surprised any wastes their time with them and why I am always shocked to see another variant with a bed that moves in Y. With a Delta printer or Etch-e-sketch (Tantillus, Ultimaker) style printer you can move the axis far faster than you can extrude. The extruder is by far the limiting factor with those other style machines.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Quote
Sublime

Quote
Srek
Currently the limit for reprap like machines isn't the extrusion speed but the speed of the rest of the machine. The mechanics just don't hold up to the accelerations needed.

This is only true with Mendel variant printers which is why I am always surprised any wastes their time with them and why I am always shocked to see another variant with a bed that moves in Y. With a Delta printer or Etch-e-sketch (Tantillus, Ultimaker) style printer you can move the axis far faster than you can extrude. The extruder is by far the limiting factor with those other style machines.

I'm currently using an ordbot with a bowden merlin nozzle. This makes the mass of the printhead minimal. Also i reduced the mass of the bed to a minimum. From the results i get when i near the maximum extrusion speed the merlin nozzle can take it does not look as if a further reduction would make much of a difference. I might be wrong and my next design will at least have x carriage that should not mass more than the one on an Ultimaker.
I would still be much surprised if i can push that anywhere near the maximum extrusion speed without significant loss in print quality.
I'm not much in favor of the Ultimaker approach, simply because i want multiple nozzles to act independently, which is simply not possible with an ultimaker or delta.
Either way, things are constantly moving on, it's interesting times

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

Moore's Law is probably one of the most misapplied laws outside of it's original field. Moore's Law works when applied to computing because you can miniaturise information processing without losing anything. i.e. a small transistor can process a bit equally as well as a large transistor, but having a small transistor has a whole host of benefits.

What works for bits doesn't work for atoms. You still need a car or chair to be the size it is, because humans can't be shrunk (if anything, they are getting bigger...).

In fact the logic of manufacturing is to go the opposite way, to handle materials in bulk, where liquids and powders are easier to handle than discrete items. That's why we use concrete instead of bricks. Using smaller bricks is the opposite of what you want.

Injection moulding is effectively extrusion into a closed volume, and is fast because a lot of material is extruded at once. Making filament thinner just doesn't help at all.

FDM is quite clever because it uses the feedstock to also provide compression for the extrusion, saving an augur or other mechanism. If you have a filament too thin to push, you would need to find another way to create pressure in the nozzle.

---

What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license