Dualstrusion Print Quality Issues

185 views
Skip to first unread message

Hammerhead

unread,
Nov 18, 2012, 8:25:57 PM11/18/12
to make...@googlegroups.com

Just wanted to see what I'm doing wrong, or if this is just the nature of the technology.

When doing a dual extrusion print with Sailfish, I'm finding that some print surfaces come out just awful.  See the attached photo 4064.  is this a problem with my model?  If so, why doesn't this characteristic reveal itself when printed alone without dualstrusion?

The second problem I'm having is how to deal with eliminating 'dribble' from the second extruder when printing two colors.  Looking at both photos, 4064 was created with dualstrusion and 4061 is two different files printed separately but with the 2nd color beginning higher in the Z axis to clear the part.  As you can see in the first (dualstrusion) print, because both print heads heat, the color oozes or dribbles, attaching to the part, and causing a sloppy print.  In the second print which is created from two separate files (individually started manually), though the part is clean on the side, as the extruder passes over the part from the Z0 anchor point on the plate, it leaves residue on the part; again making for a sloppy looking print.

So, what am I doing wrong with the Dualstrusion programming, and what can be done to prevent the 2nd color from fouling the part prematurely?

Apologies in advance if this is starting to sound too much like a bad script for a blue movie.


Dan Newman

unread,
Nov 18, 2012, 8:53:57 PM11/18/12
to make...@googlegroups.com

On 18 Nov 2012 , at 5:25 PM, Hammerhead wrote:

>
>
> <https://lh6.googleusercontent.com/-6CAkOjzdcB8/UKmKevSShSI/AAAAAAAAAQI/xl7keFnSUY8/s1600/DSCN4061.JPG>
>
> <https://lh6.googleusercontent.com/-v3E43eSyqfc/UKmKR6rJ0oI/AAAAAAAAAQA/XJwSa-Uxnnk/s1600/DSCN4064.JPG>
> Just wanted to see what I'm doing wrong, or if this is just the nature of
> the technology.

In the second picture it looks like a large bridge over nothing? You need to have a
more solid fill layers to do that well: otherwise, with acceleration, there's not a
lot for the plastic being extruded to be drawn out against. (To some extent it gets
tacked down and then stretched.) You need to have some internal fill: maybe 10%?

> When doing a dual extrusion print with Sailfish, I'm finding that some
> print surfaces come out just awful. See the attached photo 4064. is this
> a problem with my model? If so, why doesn't this characteristic reveal
> itself when printed alone without dualstrusion?

Would take an examination of the gcode to see what's different. My guess is
that you got slightly different print speeds and acceleration profiles and
that made all the difference in trying to get that large span to work. You
need to have some internal fill to help provide something to draw the plastic
out against, or have 5+ solid surface layers. Also, it could be an issue
with the extruder temperature. (When running two extruders and keeping them
plus the HBP heated and having 5 stepper drivers enabled, the rep 1 seems to
behave a little differently -- an issue with the PSU not quite having enough
oomph most likely. Not that much differently, but I've seen the voltage drop
down a bit at times when I put it on a scope.)

> The second problem I'm having is how to deal with eliminating 'dribble'
> from the second extruder when printing two colors.

There's really nothing that works well to eliminate that: the extruder's
melt chamber is under pressure and the only effective way it has to relieve
that pressure is to push plastic out the end of the nozzle -- the dribble. Even if
you slowly keep on retracting the filament, that doesn't help: the solid
filament just separates from the molten plastic in the melt chamber -- it
only pulls/reduces the pressure by a small ammount. (We've tried this and
found it doesn't help.)

> So, what am I doing wrong with the Dualstrusion programming, and what can
> be done to prevent the 2nd color from fouling the part prematurely?

You can manually wipe the nozzle as the print starts, but that's about it.

Hammerhead

unread,
Nov 18, 2012, 9:02:38 PM11/18/12
to make...@googlegroups.com
Dan, apologies for the confusion.  As forum software goes, Google Groups is lousy compared to something like vBulletin, so I wasn't able to more accurately place and document my photos, nor edit after posting to make it clear.

The picture of the surface that looks like shredded wheat is the same identical part and parameters as the other.  When printed on its own, it's fine.  It comes out perfect with 10% infill and 2 walls when on its own.  But when I program for dualstrusion with those parameters along with the white layer on top, this is what comes out.

On the picture that looks like the bridge to nowhere is fine, but that little line of white filament left on the top of the part, but you've already addressed the fact that it's just inherent to the technology, so I just have to decide if I want my splooge on the vertical face of a part in a Z axis, or on the surface in a planar axis.  ;-)

Thanks again!


On Sunday, November 18, 2012 8:54:04 PM UTC-5, dnewman wrote:

On 18 Nov 2012 , at 5:25 PM, Hammerhead wrote:

 
In the second picture it looks like a large bridge over nothing?
Would take an examination of the gcode to see what's different.  My guess is

wjsteele

unread,
Nov 19, 2012, 6:05:40 AM11/19/12
to make...@googlegroups.com
You could print a sacrificial column outside the print with the second color (white) so that the printer doesn't have time to build up the chamber pressure to avoid the string.

The other problem can be eliminated if you add an additional shell layer (or two) to the top and bottom.

Bill

Hammerhead

unread,
Nov 19, 2012, 10:14:08 AM11/19/12
to make...@googlegroups.com
Yep, tried that.  Turns out the result is actually worse, because there's still a 'drag line' across the face of the part as the extruder moves to position, but it's so thin that it's hard to remove.  This way, at least the filament is thick enough that it can be more-or-less easily removed with a knife.

Looks like it just goes with the territory, I suppose.

Thanks!

Jetty

unread,
Nov 19, 2012, 11:09:35 AM11/19/12
to MakerBot Operators
We gave Dual Extrusion a fairly good try in Sailfish to see if it
could be improved over the stock
MBI firmware.

As Dan said, dribble is the main issue, and any amount of retraction,
doesn't stop this unless you completely
remove the filament, so you're always going to get melt from the
unused extruder deposited in the
currently active extruder if the part is big enough. The reason is
the melt just runs down inside the
extruder and oozes out over time.

So it's never gonna be perfect on any firmware unless there are some
mechanical changes
(valve like / internal screw or rapid cooling).

There are some issues we found that you need to be aware of when dual
extruding
that will make your life easier.

1. Unused extruder temperature. If the unused extruder is inactive
for a while,
eventually the area around the pinch wheel will melt an indent, and
filament will
cease to extrude when needed. This is typically an issue if you have
a long
print and the 2nd color isn't used till near the end (i.e. it has time
to melt).

2. Oozing. Noodle will always ooze, wiping doesn't really solve it
unless
you wipe constantly which isn't practical.

3. Calibrate extruders. If your extruders are over filling, you have
double the poor printing
effect from 2 extruders moving around on the top sufrace.

4. Level your extruders. This is critical with ABS if you want a
clean surface.
Even more critical with PLA, because it's so hard. It's quite easy
for the
2nd extruder to knock the part off the surface.

Regarding your shredded wheat issue. It looks similar to the problem
that occurs when the
pinch wheel is sleep, i.e. grub screw not tight, filament dust around
the wheel, stepper
too hot, acceleration or feed rate too high for the pressure that's in
the melt chamber.

Another possibility is that if the extruders weren't level, with
smaller infills, the 2nd extruder
could be catching the melt as it's being laid down and repositioning
it.

Regarding leveling, it's often 1 piece of paper thickness that gets
it's just right.

PropellerScience

unread,
Nov 19, 2012, 1:58:48 PM11/19/12
to MakerBot Operators
I have my left extruder nozzle screwed in higher than the right
nozzle. That's the fix for dual extruders from Makerbot. You get one
nozzle out of the way so it doesn't interfere with the working nozzle.
After you do that you can rename that extruder "spare parts".

dob71

unread,
Nov 19, 2012, 9:14:43 PM11/19/12
to make...@googlegroups.com
Here is how I'm currently doing the dual extrusion prints:
This is done using RepRap X2, but might work for Replicator too (with some profile adjustments to work around X2 firmware features I use, like for example relative temperature control).
 
What's also interesting is that due to the thermal expansion/contraction the idle hotend stays high enough during the single extruder prints (as long as I do not over-extrude or have other issues like lifting of the corners etc.).

Hammerhead

unread,
Nov 20, 2012, 12:39:22 PM11/20/12
to make...@googlegroups.com
Ok, guys.  Looks like it's a firmware issue.  I just tried with a different model, and this time after the first few passes in the Y axis, there wasn't enough filament being extruded, and what did come out just balled up on the bottom of the nozzle.  I guess the value of buying a dual extruder model just became zip for me.
Reply all
Reply to author
Forward
0 new messages