How to improve consistency of filament diameter

[Ralph Person]

Hi all,

I've been reading about the inconsistent diameter of the filament some users are extruding, and I thought I would start a discussion about what can be done to fix this.

From what I've read, the problem occurs right at the nozzle.  The molten filament is very sensitive to any tugging or environmental factors and therefore deforms enough to mess up the diameter of the filament.

There is a thread which I found here: http://www.soliforum.com/topic/3439/lyman-filament-extruder-v3

... in which some folks discuss a modification to a filastruder.  The video appears to no longer work anymore, but they are discussing a method of using water to cool the filament immediately after it exits the extruder.  A design is available on thingiverse which seems to depict something similar: http://www.thingiverse.com/thing:146105

What do you guys think about this?

Suppose someone could come up with an attachment to use instead of the nozzle which comes with the kit.  The attachment would have a consistent 1.75 or 3mm inner diameter the whole way through, but by use of water, compressed air, or some other means the filament would be cool by the time it exits the attachment.  Eliminating the molten characteristics of the filament at its exit point from the machine should in theory reduce the deformation which has been known to occur.

Thoughts?

[Thomas Braun]

Am 14.10.2013 22:09, schrieb Ralph Person:
> Hi all,
> I've been reading about the inconsistent diameter of the filament some
> users are extruding, and I thought I would start a discussion about what
> can be done to fix this.
> From what I've read, the problem occurs right at the nozzle. The
> molten filament is very sensitive to any tugging or environmental
> factors and therefore deforms enough to mess up the diameter of the
> filament.

[resending email to list]

Hi,

the cited approach sounds reasonable. Although I think one should first
characterize the current filament diameter, diameter as function of
filament length. And then think about ways to get it more consistent.

What are your currently achieved diameters? With that type of pellets
and temperature?

My results so far were with ABS pellets are not so good. With 1.75mm
nozzle the main motor overheats and with 3mm the filament has most of
the time severe poppings.

[Raymond Luo]

Currently I am getting a variance of -10mm from 1.75mm with PLA.

From what I have observed there are three issues.

1. Gravity

PLA has quite a low viscosity when it is melted/molten and while I am getting straight up 1.75mm diameter for the first 3cm that's extruded, anything beyond that seems to be pulled by it's own weight.

2. Nozzle length

The nozzle that is provided has a length of about 3mm? The thought I have is that the extruded filament being pushed out from the nozzle is still hot and thus the 'shaping' done from the nozzle is easily reshaped from the factors such as 1. gravity 2. cooling fan 3. spooling motor etc.

Thus my theory for improvement here is that if the length of the nozzle is longer, the filament passing through it will have time to cool and solidify into it's shape, since the nozzle (if longer) would be further away from the heater band, the difference in distance might just be enough for the filament to solidify but not clog.

Looking at Filastruder's nozzle it seems that it would be perfect if it can be adapted to the Extrusion bot. If anyone here can confirm that the nozzles are of the same size please let me know!

3. Cooling fan

As briefly mentioned in 1. and 2. the cooling fan puts the hot filament in quite a bit of stress and stretch. Industrial extruders use water cooling systems for this, however, since the extrusion bot is vertically extruded a solution is not likely to come 'off the rack'.

[Ralph Person]

Here is an interesting apparatus on thingiverse for your consideration.  http://www.thingiverse.com/thing:43499

I'm picturing a modified version of this mounted with the circular opening centered around the extrusion bot nozzle.  It might also be possible to swap the PC fan in favor of a single opening or port to attach tubing connected to an aquarium air pump.  The air would be directed in such a way that it hits all sides of the molten filament with the same intensity.  Plus a lot of the nicer air pumps have a little dial built right in so you adjust the intensity as needed.