Monday, July 28, 2008



"Thou shalt raise up the foundations of many generations, and thou shalt be called, the repairer of the breach, the restorer of paths to dwell in." Isaiah 58:12

Well. It seemed appropriate...

Inspired by Nophead's outstanding experimental results, I've added code to the Java host software (in the repository here) optionally to put foundation layers under a build. This works particularly well with ABS (above - not yet up to Nop's quality, but servicable). I suspect it should also work with PLA, though I haven't tried that yet.

The host starts by putting down several widely-spaced hatch patterns on top of each other, then a narrower pattern, but still with horizontal gaps, then finally a hatch at the normal infill horizontal hatch gap. This last layer, though, is not the standard height-gap above the previous one, but less. This means that it is forced into the gaps in the previous layer, giving a very strong bond. The first layer of the part is then plotted where it should have been had the gap not been reduced, giving a bigger gap than normal between the part being made and the foundations. Consequently it doesn't stick that firmly to the foundations and can be separated easily with a penknife blade.

Nophead has found that this process is sensitive to ambient temperature (hot days give more stick), so we may have to allow for that. He suggests simply lowering the extruder temperature by the same amount that the ambient temperature exceeds some norm (say 20oC)

Foundations allow us to do two things:

  1. Build on an absorbent or rough base to give good keying. This will prevent parts separating from the base as they cool and tend to curl up. The part above (the direct-drive extruder motor mount - see that link for what it does) came out flat as a pancake, even though ABS is prone to curl.
  2. Experiment with layer separation strategies (like the changes in layer height that I mentioned). Once that is fully understood (or even before...) we can use the polymers as their own support for overhangs.

That's great. skeinforge is still a bit exotic to me, even though I usually prefer modular CLI tools over GUI.

I'll try it as soon as my extruder works again. Don't know when that will be though...

Nice to see a different color for a change (neither black nor white can officially be called colors...).
One more thing Adrian, are you using balsa wood for ABS as well? I knew it worked for some plastics (either PCL or PLA if I recall correctly)...

If so, I think I'll order some.
Yup - that was printed on balsa. I set the initial layer so the head was only 0.1mm above the balsa. That forced material into the grain, giving a good bond.
Is it still flat after a day off the raft?

I wouldn't describe anything I make more than 20mm across as "flat as a pancake". There is always some visble warping.

I also find Balsa is not strong enough for ABS but here you show that it is and BruceW is using anti-stat foam successfully I think.

It leads me to believe that I get more warping than the standard RepRap setup for some reason.

Please can you point me to the file and describe the build parameters so I can try a comparative test.
BTW, where did you get the black ABS from?
If the separation isn't destructive, perhaps a whole-range, re-usable foundation would use less material in the long run?

Along those lines: understanding separation would also make the platform replicable, naturally.
This comment has been removed by the author.
LOL! I'd have thought that Isaiah 40:4 would have been more appropriate! I guess that's just me and HDPE, though. :-p
Yeah, nophead like myself is dying to print those black chess pieces :P
"Every valley shall be exalted, and every mountain and hill shall be made low: and the crooked shall be made straight, and the rough places plain."

Thanks Forrest. Now. Enough with the Bible quotes, already.

The black ABS came from:

It was £18 a kilo, so not the cheapest.

Like Henry F. they said I could have any colour I wanted as long as it was black...

I've just measured my pancake. It does have a slight curl: 0.6mm in a length of 50mm. And Nophead is right: it developed that quite a while after I separated it off. It was flat when I first made it to within 0.2mm.

Conditions: my thermistor reckons that the barrel temp is 200 C, and I'm laying down the material with an XY head speed of 420 mm/min (thanks to Zach's stirling efforts at making all the code work in real units internally on his way to the G-code stuff). I'm using the geared extruder with the PWM mark set to 215/255; I don't know what that is in ml/min; as soon as I get some more pins on a bigger Arduino I'll implement Nop's extruder feed control - the hardware (slotted disc + optoswitch) is all in place.
PS. The file is

There are two objects in there; you want the flat one called

I've been meaning to try putting it in an oven just below the melting point to relax it before removing it from the raft. The problem is that my base materials melt at the same temperature so would probably warp themselves. Now you have got it to stick to balsa that should be possible.
"Thanks Forrest. Now. Enough with the Bible quotes, already."

If you stress relieve in an oven, you will releave stresses. In releaving the stress, the thing will probably warp in even more ways.
I assume the warping results from laying a hot layer on top of a cold one that has already shrunk - as the new hot one cools it contracts and so the combined beam is warped up at the ends.

The solution is to keep all the layers hot and contract the completed part as a whole = even contraction = hopefully no warping.
I made one and got about 1mm curl so possibly I can improve on that by reducing the speed.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]