[MakerBot] Concept for Extruder

[baxsie]

Here is an idea for making a new extruder.

Ref: http://img696.yfrog.com/img696/4599/extruderconcept.png

This is really rough, just conceptual at this stage. Here is what I
had in mind:

1) Get the heater (nichrome coils) close as possible to the tip where
we want the melt to happen.

2) Use PTFE tubing to keep the hole slippery, but do not depend on the
PTFE at all for structural support.

3) Get the mounting washer + nuts as far away from the heat as
possible, while keeping the brass tube as short as reasonable so that
the machining of it is not impossible. The brass will start as a 1/4"
or 5/16" rod threaded, turned and counter-bored.

4) To work with the existing Makerbot chassis, and keep the brass
short, the Torlon may need to be longer. I do not think newer designs
would need such a long extruder (that is the Torlon could be even
stubbier if the dinos were shorter).

5) The extruder body may need to be back-bored to accept the top end
of the PTFE tubing sleeve, which ideally would start just below the
idler +pinch wheel.

6) The thermistor would contact just above the nichrome coils.

7) Perhaps active cooling of the washer (small fan blowing on it), to
assure that the filament solid well into the extruder. This may not
matter, since it is unlikely that the filament would melt all the way
back to the pinch wheel. I have read that the we want the melt zone to
be short, cooling the washer while still keeping the tip hot would do
that.

8) There would be thermal insulation below the washer, as usual.

9) I do not think that the torlon would need to be threaded. The PTFE
tubing sleeve makes it so there is no where to leak. In fact, it may
not even need to be fancy high-temp plastic. In some previous
experiments the fan cooled washer kept the top of the stock makerbot
tube at like 50C to 60C.

Please point out any shortcomings you see. If you have some idea of
how to make it better, please comment or mark up the sketch.

Thank you.

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.

[Koen Kooi]

Op 28 apr 2010, om 07:00 heeft baxsie het volgende geschreven:

> Here is an idea for making a new extruder.
>
> Ref: http://img696.yfrog.com/img696/4599/extruderconcept.png

[..]

> Please point out any shortcomings you see. If you have some idea of
> how to make it better, please comment or mark up the sketch.

I came across some comments about extruders at http://makerblock.com/2010/03/whats-so-great-about-a-reprap-anyhow/#comments this morning. The second comment seems to hint that the nichrome needs to cover a much larger area. I don't know if that's true, but it is something to think about :)

regards,

Koen

[TeamTeamUSA]

Nice! nophead and Zaggo have done some very interesting extruder
experiments. If you haven't seen, check 'em out:

nopehead
http://hydraraptor.blogspot.com/2009/11/no-compromise-extruder.html
http://hydraraptor.blogspot.com/2009/11/pinchless-extruder.html

Zaggo
http://pleasantsoftware.com/developer/3d/2009/12/06/back-to-the-extruder/
http://pleasantsoftware.com/developer/3d/2010/03/08/printruder-ii/

Poke around on their blogs for more fun.

Go!

=ml=

[Owen M Collins]

And have you sen Zach's flickr stream? He is working on the MK5 plastruder:

http://www.flickr.com/photos/hoeken/4515934498/

It looks like they are replacing the idler wheel and some other things.

O.

[Koen Kooi]

Op 28 apr 2010, om 15:07 heeft Owen M Collins het volgende geschreven:

> And have you sen Zach's flickr stream? He is working on the MK5 plastruder:
>
> http://www.flickr.com/photos/hoeken/4515934498/
>
> It looks like they are replacing the idler wheel and some other things.

Now I'm confused :)

The wiki ( http://wiki.makerbot.com/plastruder-mk5 ) implies it will be a bowden style extruder, but the picture ( http://www.flickr.com/photos/hoeken/4515934498/ ) seems to have the feeded and hot end integrated. I guess it's time to find a dxf viewer for OSX and inspect makerbot svn to look at the exciting new changes :)

regards,

Koen

[baxsie]

Thanks for your input. Lots of good links.

> The second comment seems to hint that the nichrome needs to cover a much larger area.

I plan on making the length of the active heating section about 1/2"
long. Considering its small diameter, that will probably mean 2 layers
of nichrome to keep the same 6 ohms. Having the same power input over
a small physical volume should give the electronics good control
authority. One thing that bugs me with the current MakerBot is that
you should see the extruder heater light cycling even when extruding.
If there is no cycling, then the temperature must be below the set
point, and since thereis no cycling, you have no idea how much below
it is.

There is a certain amount of heat needed to melt the filament when it
it moving, and less when it is stopped. If the electronics had good
control authority then I think the difference in cycle time should be
visible between extruding (more time on) and Idle (less time on).

Making a thin wall in the brass will reduce the overall heat needed in
the system, and give a steeper thermal gradient.

On this page:

http://hydraraptor.blogspot.com/2009/03/rheology.html

This image:

http://2.bp.blogspot.com/_QYCIPYPZ-pc/SbGSkXNpWdI/AAAAAAAACqs/HBjl3SRdM4c/s800/transition.jpg

Shows what I was thinking, but much more elegantly.

The big thing I am trying to do is to make sure the filament stays
solid as long as possible, and the viscous plug is as short as
possible, and that all of this transition region is in a PTFE
environment.

To keep the viscous plug as short as possible, you need a strong
thermal gradient.

This design is the MOATG (Mother of All Thermal Gradients) with the
massive heat sink and the super thin wall,

http://pleasantsoftware.com/developer/3d/2009/12/06/back-to-the-extruder/

My theory (and limited experience) in the original MakerBot extruder
is that the entire brass barrel gets into in the viscous fluid stage,
which means that the thermal gradient and plug has to move up into the
PTFE. The problem here is the leakage between the PFTE and the brass,
and a large region of plastic to push through the brass. Plus the PTFE
tends to deform and become unstable at these temperatures. Below full
operating temperature, the plug is in the brass and I think the plug
sticks to the brass causing the filament to strip.

My hope is to get a decent friction fit between the outside of the
PTFE sleeve tubing and the inside of the brass bore, that is where my
design would leak.

(Secondary goals: able to build it on a $335 harbor freight mini-
lathe, use most of the maker-bot bits and pieces, especially the
electronics, and the extruder.)

I will need to get the worm gear filament drive going, the original
MakerBot gear deforms the filament so much that it will not fit into
the PTFE tubing am thinking of.

OK, this is all thought experiments at this time. Once the tubing and
drills and brass show up then it can turn into actual parts.

Thanks again for your suggestions and input and links.

[RyanP]

I don't know much about bowden, but it seems that the problems people
are having with the MK4 call for a much simpler approach. Zach's new
design reflects nophead's "no compromise work", but i could be wrong.
There is no reason a bowden upgrade would not be on the roadmap, but
the firmware and hardware both have a long ways to go and a lot of
testings.

After about 10 rebuilds of my plastruder I'm getting close to perfect
for my purposes. I most of Makergear's pla extruder directions and a
PTFE washer under the M6 nut.

Ryan

On Apr 28, 9:45 am, Koen Kooi <k...@beagleboard.org> wrote:
> Op 28 apr 2010, om 15:07 heeft Owen M Collins het volgende geschreven:
>
> > And have you sen Zach's flickr stream? He is working on the MK5 plastruder:
>
> >http://www.flickr.com/photos/hoeken/4515934498/
>
> > It looks like they are replacing the idler wheel and some other things.
>
> Now I'm confused :)
>

> The wiki (http://wiki.makerbot.com/plastruder-mk5) implies it will be a bowden style extruder, but the picture (http://www.flickr.com/photos/hoeken/4515934498/) seems to have the feeded and hot end integrated. I guess it's time to find a dxf viewer for OSX and inspect makerbot svn to look at the exciting new changes :)

>
> regards,
>
> Koen
>
>
>
>
>
>
>
> > O.
>
> > On Apr 28, 2010, at 3:33 AM, TeamTeamUSA wrote:
>
> >> Nice! nophead and Zaggo have done some very interesting extruder
> >> experiments. If you haven't seen, check 'em out:
>
> >> nopehead
> >>http://hydraraptor.blogspot.com/2009/11/no-compromise-extruder.html
> >>http://hydraraptor.blogspot.com/2009/11/pinchless-extruder.html
>
> >> Zaggo

> >>http://pleasantsoftware.com/developer/3d/2009/12/06/back-to-the-extru...

> For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en.- Hide quoted text -
>
> - Show quoted text -

[Joe Allen]

You really do need to check out MakerGear. Rick already has a hybrid
insulator (PTFE sleeve inside PEEK barrel) and his heater core is very
compact and does the job extremely well. I commend your efforts to
improve the extruder, as it is definitely the weakest link in the
MakerBot design, but it seems like a lot of what you are talking about
has already been addressed.

At least for the issues I've had with my machine, the real flaws are
with the feed mechanism more than the hot end. I'm working on
building one of Zaggo's printruders to test that design out, as it
seems to simplify things quite a bit.

Also, some fellas I met recently asked the simple question: why don't
we use rubber wheels to feed the filament? With the right material,
I'm sure it could have plenty of grip, and you could really crank down
the tension without worrying about deforming the filament like all the
geared mechanisms do. No more cracked acrylic, no more stripping
filament and flossing of gears, etc.

On Apr 27, 10:00 pm, baxsie <br...@crystalfontz.com> wrote:

[baxsie]

> You really do need to check out MakerGear.

I have a bunch of MakerGear stuff on order. I definitely plan on
checking out how they do it. Their heater looks awesome, but I think
it would be hard to improve on having the nichrome being right on the
end of a single, thin brass machining. Time and experimentation will
tell.

> hybrid insulator (PTFE sleeve inside PEEK barrel)

Yes, but I think the melt/freeze interface on the MakerGear deign can
still be in the relatively sticky brass. My idea is that the near-full
length PTFE sleeve will ensure that the melt/freeze interface is
always inside the Teflon.

> the real flaws are with the feed mechanism more than the hot end.

I disagree . . . I think the problems in the hot end (specifically
freezing in the brass) cause the stripping of the drive. I may be
wrong--though that would be a first ;) I think the drive pushes like
crazy. I do not like the way it deforms the heck out of the filament
(hence the work on the half-round worm gear setup), but I think it
pushes plenty hard.

Reading through this article gives me hope that I am on the right
track:
http://hydraraptor.blogspot.com/2009/03/rheology.html
I'll have to build it to find out if if really works.

> why don't we use rubber wheels to feed the filament?

I made a small tire by grinding down high temperature automatic
transmission fluid hose, and slipped it over the stock MakerBot drive
gear. It gripped fine, but not hard enough to make a good extrusion.
The MakerBot gear pushes really hard. We need to solve why all that
force does not result in filament flowing through the nozzle.

It might be once the friction/sticking in the hot end is fixed that a
rubber drive may work fine.

[Joe Allen]

Well, I did say that the mechanical side was the culprit for most of
*my* issues. And I guess I should have clarified that I was talking
about issues for which there is no clear current solution. I
originally built my bot with the MakerGear PEEK insulator and heater
core, so I've not had any issues with the hot end. The improvement
over having the nichrome wrapped on a single-piece heater tube is that
the heater core can simply be unthreaded if it things need to be
rebuilt, instead of having to unwrap and rewrap and tape the nichrome
every time.
On the other hand, I do get very tired of reassembling the acrylic
parts, and unmounting the motor to floss the gears, and futzing around
trying to get the idler wheel tension just right, etc. I agree that
the force of the motor is not in question, but I find it to be a
tempermental setup to maintain and rebuild when things do go wrong.
Certainly issues with the hot end can lead to feed problems (and vice
versa), which is why troubleshooting can sometimes be so tricky.
There are several chickens and eggs when it comes to the current
extruder, so figuring out where the true problem lies isn't always
simple. Try new things and see what works. Any simplification of the
design I think is an improvement, and that's what this group is all
about: figuring out better and better ways to do things so the machine
evolves into a better and better one.

On the idea of a full-length Teflon tube, though, wouldn't that get
too hot? People already have problems with bulging insulators, so I
hate to think what a mess it would make of things if the sleeve got
soft in the heater barrel.

[baxsie]

> the mechanical side was the culprit

Here is my big assumption: If the filament never sticks in the
extruder, then the drive gear should never strip.

My idea is that the Teflon sleeve is "backed up" by the brass that
surrounds it in the hot region. The pressure in the "melted" zone is
be radially outward. There will be a region of force downward at the
viscous plug. If the sleeve gets soft enough that it mushes and slips
farther down into the nozzle, then it would be a failure.

I have ordered PTFE, FEP and PFA samples from
http://www.fluorostore.com/servlet/StoreFront
(Wow, I ordered samples 5 feet each, now the store says 50 foot
minimum! I must have made them angry.)

Hopefully one of these will be stiff enough at the ABS melt
temperature to stay in place in the brass tube.

I have a MakerGear ceramic heater (and nichrome and thermistors, etc)
on order . . . I'll have to take a look at it and see if it is good. I
already have some other ideas to mount the thermistor.

[ddurant]

I'm very much looking forward to seeing how this thread (and others)
turn out but this sounds like it's getting sorta close to the point
where teflon becomes evil..

http://www2.dupont.com/Teflon/en_US/keyword/fumes.html?src=search_us_teflon_toxic_gas
says "DuPont non-stick cookware may produce fumes at significantly
higher temperatures, greater than 500°F (260°C)."

I don't know how the PTFE barriers are made and what their
characteristics are but you might want to make sure that they're at
least good (and safe!) for the temperatures we're melting stuff at,
with a bit of margin...

[Lawrence Harris]

I am going at some point to add a current shunt to the motor. This is going on the assumption that the current flow will be an indicator of how hard the motor is running. If the current rises to 'stall' current then either the temperature needs to be raised or the project stopped until the blockage is cleared. There are probably some startup issues here as the current profile would need to be characterized and it may be too small to profile accurately. A rotary encoder on the motor or idler wheel might be just as good or better.

My basic thought is that from a cold start you should be able to turn on the motor and monitor the current, then enable the heater. When the motor starts turning the current will drop and you can save the thermister value as the 'set' point. If it works out you would not have to calibrate for the exact thermister value or adjust the temperature for different materials. If the current start to rise you could raise the set point and if the current reaches the stall point you can assume there is a problem and shut down before anything breaks (hopefully).

Lawrence

[Rick Pollack]

Brent - By all means make it and test it but from my vantage point having tested many plastruder designs including using PTFE tubing inside a brass barrel...lining the brass nearly all the way down with PTFE tubing is unnecessary and just adds complexity to the extruder.

There is really is no such thing as "sticky brass" - the transition from filament to a melted, extrudable material is a very manageable process. The problems with the MK4 have to do with there being  too many opportunities to mess it up - nichrome can be wrapped incorrectly, insulator can be put on wrong, washer can siphon off to much heat, pinch wheel can be loose or misaligned, thermistor table can be wrong, wrong print temp,  etc. For someone new to it - especially if more than one element is out of whack - it can be very challenging to get the extruder working consistently. But get all of the elements aligned and it will extrude, consistently.

Based on what I've read in this thread, most of what you are trying to solve with the PTFE tubing is really a non-issue. I'm not saying not to try it but if you like nophead's post on rheology then before investing a lot of time and money (torlon is really expensive) in this design I'd suggest making a glasstruder - it is basically the laboratory companion to the rheology piece.

Rick

[beak90]

I think the PTFE shouldn't be a problem. The extruder shouldn't ever
get hotter than 240C under controlled circumstances. If you do get
hotter than that then the ABS will likely start emitting fumes as
well.

The only problem with this design is the mounting system. The washer
dissipates a huge amount of heat. This could cause clogs when the
plastic reaches its glass transition temperature. The glass transition
temp is the temp at which the plastic gets soft but doesn't melt yet.
It could also cause problems like the barrel not heating up enough
because of too much dissipated heat.

On Apr 28, 4:03 pm, ddurant <dduran...@gmail.com> wrote:
> I'm very much looking forward to seeing how this thread (and others)
> turn out but this sounds like it's getting sorta close to the point
> where teflon becomes evil..
>

> http://www2.dupont.com/Teflon/en_US/keyword/fumes.html?src=search_us_...

[beak90]

I just had an idea of a new extruder. Its basically the same as the
current Makerbot extruder, but it has a different mounting system and
allows for use of any type of insulator imaginable. The mounting
system I came up with makes it so that there is no heat loss as there
is with a metal washer.

Here's the blog post: http://makerbot448.wordpress.com/2010/04/29/extruder-mounting-system/
And here's the picture by itself: http://www.flickr.com/photos/49325300@N07/4561367969/

[ddurant]

> I think the PTFE shouldn't be a problem. The extruder shouldn't ever
> get hotter than 240C under controlled circumstances. If you do get
> hotter than that then the ABS will likely start emitting fumes as
> well.

I agree but the Dupont article was about cookware - I have no idea
what the toxic fuming temperature of these barriers are and this
design (which is far better than anything I could have ever come up
with, BTW) moves it closer to the heat source.

If people know all this and still declare it good, I'm happy. Just
wanted to point it out and make sure somebody, somewhere has thought
about it.

> For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en.- Hide quoted text -
>
> - Show quoted text -

[baxsie]

Thanks for all the good input. I love being able to push the "thought-
level" design of this thing out there and benefit from the collective
experience of the group.

Can we put images in here?

<img src="http://img34.yfrog.com/img34/9150/extruderconcept2.png" />

Here is the link in case that does not work:

http://img34.yfrog.com/img34/9150/extruderconcept2.png

Again, this is just a concept at this stage. Comments and suggestions
are welcome.

My biggest worry I have at this point is that the ABS will flow back
out around the (soft because it is hot) PTFE sleeve between the sleeve
and the brass body. My idea is to make the hole in the brass a snug
fit around the PTFE and kind of force the PTFE down into the tube. I
hope that will seal.

Another worry is that the PTFE I ordered is not that much bigger than
the filament, and I think the deforming and burrs caused by the drive
gear may not slide through the PTFE sleeve. Hopefully the worm gear
drive will not mash the filament too much.

The comments about the MakerGear ceramic heaters makes me want to coat
the nichrome wrap with a thin layer of ceramic, but the small volumes
of ceramic mix that MakerGear sells ( http://www.makergear.com/products/castable-ceramic
) were out of stock. Now it is back in stock. Maybe I better place an
order ;)

Rick Pollack: I love your pile of extruders*. What design had the best
luck? Did you ever find one that consistently works over a lot of
conditions? The glasstruder looks interesting, but part of my goal is
to use the MakerBot electronics, so I feel I need to be very careful
of the heat budget. I think the heat input into the glasstruder must
be large, considering that the washer is near the tip and that there
is a large volume of melt in the glass. Does it get hot enough with
12v and the 6 ohm NiChrome?

* The fact that you have so many extruders reinforces to me that we
need to keep working on this until a good, manufacturable, repeatable
design comes out. I assume that most of the ones in your drawer are
not in use because they have _some_ failing.

beak90: I think that design could work. I think the ABS will be fluid
in the whole brass section (this is what I saw in the original
MakerBot design). How will you keep the fluid ABS contained from the
brass up? Frankly, the MakerBot design works fine _IF_ you can get a
seal between the brass and the PTFE. That is pretty darn hard once the
PTFE gets pretty darn soft :(

The washer dissipating heat is desired in my proposed design. That
cooling of the washer gives control over where the melt zone is. There
is some question if the brass tube can be manufactured thin enough to
make it thermally resistive enough that the 12v/6ohm heater element
has enough power left over to give good control authority. I would
like to see no more than ~75% on time during extrusion and perhaps
~25% to ~50% on time if stopped. I do not have the brains to try to
model that, but the prototype should show us what is going on.

Lawrence Harris: If you control the current of an ideal DC motor, you
create a constant torque, which will end up being a constant pressure/
force. The stickiness and negative feedback that nophead refers to
here ( http://hydraraptor.blogspot.com/2009/03/rheology.html ) begs
for a constant speed (= constant voltage in a ideal DC motor). A huge
benefit of constant torque would be that the torque could be set below
the strip force. A drawback is that the extrusion speed would vary
based on the amount of friction in the system. A "smart" hybrid
control approach might be good: constant speed to some limiting
current (below the stripping force) then stall the motor at constant
torque until the force is relieved. Anytime a stall is encountered in
a build you would risk a void.

One of the drill bits came today:

http://yfrog.com/0msmalldrillbit04mmj
http://bit.ly/c3L6s5
http://bit.ly/bPZG8S

but I am still waiting on brass stock, tubing and Torlon stock. Lately
I am really liking the idea of replacing the Torlon with wood :)

OK, all great ideas. Please keep any feedback or experience coming. We
can beat this thing.

[Rick Pollack]

If you'd like to discuss MakerGear specific stuff, please post on the MakerGear Google Group.

But, yes, a "good, manufacturable, repeatable" and much simpler plastruder is already available...

Rick

[beak90]

The plastic shouldn't flow up the tube. The plastic doesn't melt
immediately. This creates a plug at the top. The whole heater area can
be the same temperature as long as it isn't too long. I would imagine
that if you make the melt section too short, you would have to raise
the temperature to get the plastic to melt soon enough.

> > here (http://hydraraptor.blogspot.com/2009/03/rheology.html) begs

> > for a constant speed (= constant voltage in a ideal DC motor). A huge
> > benefit of constant torque would be that the torque could be set below
> > the strip force. A drawback is that the extrusion speed would vary
> > based on the amount of friction  in the system. A "smart" hybrid
> > control approach might be good: constant speed to some limiting
> > current (below the stripping force) then stall the motor at constant
> > torque until the force is relieved. Anytime a stall is encountered in
> > a build you would risk a void.
>
> > One of the drill bits came today:
>
> >http://yfrog.com/0msmalldrillbit04mmj
> >http://bit.ly/c3L6s5
> >http://bit.ly/bPZG8S
>
> > but I am still waiting on brass stock, tubing and Torlon stock. Lately
> > I am really liking the idea of replacing the Torlon with wood :)
>
> > OK, all great ideas. Please keep any feedback or experience coming. We
> > can beat this thing.
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "MakerBot Operators" group.
> > To post to this group, send email to make...@googlegroups.com.
> > To unsubscribe from this group, send email to

> > makerbot+u...@googlegroups.com<makerbot%2Bunsu...@googlegroups.com>
> > .

> > For more options, visit this group at
> >http://groups.google.com/group/makerbot?hl=en.
>
> --
> You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
> To post to this group, send email to make...@googlegroups.com.
> To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.

> For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en.

[baxsie]

> The plastic shouldn't flow up the tube.

Absolutely. But this is exactly what it did do on one of my makerbot
runs. Maybe I just got unlucky.

I love nophead's and pleasantsoftware's work. I am a by surprised how
close this concept is to pleasant's design:

http://pleasantsoftware.com/developer/3d/2009/12/06/back-to-the-extruder/

I am still struggling conceptually with how to machine the outward
inner taper that nophead and pleasant use. I really like the idea of
the concentrated heat and single piece machining, but I do not see how
to do that and have the expanding inner hole. Lots of thinking to do.

I have wondered: What if the taper was 90 degrees? Imagine a small
heated can with the filament coming in the top center and the
extrusion coming out the bottom. The thermal gradient would be radial
(cool in the center where the filament comes in, and hot on the
outside of the can where it is heated by nichrome wrapped around it.
Well I guess I will see how this design turns out (get it, lathe?) and
then decide how to proceed.

I am still wondering how or if the PTFE sleeve will work out. Rick:
What happened with your model that had the PTFE sleeve? Did it fall
down, and if so, where and how?

The 3/4" x 12" Torlon stock showed up today. I think it was around
$150 shipped. HFS! Anybody want to buy an inch or two? :) I am now
hoping that the top of the extruder can be cool enough that a low-tech
material (oak?, pine? balsa?) can be used. Also got more tiny-as-heck
drill bits from Grainger. Still waiting on the PTFE tubing and brass
rod stock.

This weekend I plan on (re)building the worm gear drive wheel. I need
to get that going so I can see what the outer diameter of the "I have
worm tracks down the side of me" filament that is past the drive/pinch
stage. I have thought or making the rim of the pinch wheel concave to
match the diameter of the filament. Is that a common mod?

Extrude body material: nophead used stainless for his tube. I plan on
following pleasant's design of brass. Have you heard of anyone using
aluminum for the extruder tube?

[Cathal Garvey]

A while back, myself and Clothbot independently designed "printable" barrels and had them printed by shapeways in steel. Both designs were pretty pie in the sky and sci-fi seeming! It was cool stuff. But, they didn't print right.

One of the most fascinating ideas Clothbot implemented in his design was a star-shaped set of ridges at the end of the barrel, which was intended to "cut" into the core of the filament as it arrived at the nozzle-end of the plastruder to maximise heat delivery at the point of extrusion. I loved this idea; at the moment, we have an edge-to-center mode of heat delivery, which is hard to beat for reliability but not ideal for rapid, homogenous heat transfer.

A machinable alternative to an exotic printed solution might be to cut slots into the end of the brass barrel and cut channels running up through the threads, so that very fine wire or even nichrome could be run across the end of the barrel before fitting the nozzle. Once hot, a very fine length (or several lengths criscrossing) of fine wire might slice larger filaments or more difficult plastics (PLA?) into more manageable bits right before impacting the nozzle, while perhaps helping to deliver heat into the core of the plastic as well.

It could be nonsense, and it could ruin a perfectly good barrel to even try it. But if this thread is in the spirit of "Make it better or destroy it trying", perhaps it's worth a shot?

--
letters.cunningprojects.com
twitter.com/onetruecathal
twitter.com/labsfromfabs

[Cathal Garvey]

I should add by the way; our stuff didn't print right from shapeways because they were designed very optimistically and didn't (at least in my case) conform to the tolerance guidelines provided for the stainless steel material. So it was in no way a failing or deficit on Shapeways' part!

[Andrew Plumb]

My latest thought-variation on my Shapeways experiment is to fabricate a removable barrel insert with holes, groves, edges, etc. I haven't pulled the trigger on this one yet, but one definite advantage to doing an insert is you should be able to machine its form relatively easily using a drill press or CNC mill.

Andrew (aka ClothBot).

--

"The future is already here. It's just not very evenly distributed" -- William Gibson

Me: http://clothbot.com/wiki/

[baxsie]

Cathal / Andrew: I found Clothbot's extruder here:
http://www.flickr.com/photos/aplumb/4442493498/in/set-72157623517888473/
is there a better link to a write-up/drawings of its features/build
blog/results? Do you have a link to Cathal's Shapeways design?

In general, what do people who actually print parts on a regular basis
use for an extruder? Do they work, or is there still the occasional
drive stripping, clogging, etc? I assume that MakerGear::Rick's brass/
ceramic/peek design must work. Does it ever strip the filament drive?
Are their other simpler designs that give "never strip the drive, I
forgot there was even an extruder on this thing" rock solid
performance over time?

Using shapeways opens up some interesting design possibilities.

Since I now have this big chunk of very expensive Torlon with no
planned use this second, I am thinking of a mini-tank system. I need
to take some time to sketch the concept up. Sources for small high-
temp o-rings ? This tank system would have several (at least 3) pieces
held together, so I already do not like the construction of it.

I do think the "right" design will be very simple in design,
manufacture, assembly and use. We just have to keep poking around in
our collective minds until the bits and pieces fall together.

[Brent Crosby]

Another extruder concept. This idea uses a small tank (maybe 1/2 inch diameter).

It is an expansion on the reverse taper in the transition region championed by nophead. The taper here is 90 degrees :)

The idea is that the transition region is immersed in in the liquid or soon to be liquid ABS.

There are mechanical problems with this already, it would need some way of holding itself together, and the PTFE sleeve is not anchored. :(

This design could be offset, so if the solid filament plunged far into the tank, it would not hit the exit hole. The tank could be horizontal, for instance.

OK, just an idea. fire away with comments, suggestions, improvements, personal attacks, etc.

I am also trying to attach an image using Rick's instructions from the MakerGear group.

Here is the external image: http://img210.yfrog.com/img210/989/extruderconcept3.png

[Eberhard Rensch]

Hi,
I really like your first concept (of course, I also like your photos of building the second design!).

However, I used your sketch as base for a new extruder. See the following blog post for more information on the design and some photos of building it:

http://pleasantsoftware.com/developer/3d/2010/05/15/complextruder/

Cheers,
Eberhard (MakerBot #127)
Blog: http://pleasantsoftware.com/developer/3d/

[Brent Crosby]

Eberhard:

"Complextruder" I love it!

The second one:

http://groups.google.com/group/makerbot/browse_thread/thread/eb1a5c66b7aa00d2/0c231c3032a76915

is complete now. It appears successful and now that the extruder is
not an issue it has unearthed several other problems (platform not
quite level, need a spool to feed in the filament that does not bind,
etc)

I will try to get a "what I learned" post and photos up for it in the
other thread (I have no idea why Google Groups made it a separate
thread :( )

[Koen Kooi]

Op 16 mei 2010, om 03:08 heeft Brent Crosby het volgende geschreven:

> now that the extruder is
> not an issue it has unearthed several other problems (platform not
> quite level, need a spool to feed in the filament that does not bind,
> etc)

The Makerbot Industries people seems to be working on a spool: http://wiki.makerbot.com/spindle-mk1

regards,

Koen