Vik and Simon are right at the coalface getting the Mk II to perform right now. I thought that a special thread to explore the issues they are uncovering would be a good idea.
From a recent blog entry by Vik...
So I made a 0.5mm nozzle with a more conical profile. This extrudes a 0.8mm filament, which with care and attention might make a trail narrower than 1mm. Maybe even not too far away from the Stratasys' 0.6mm minimum reliable extrusion width.
But lo, the pressure generated inside the extrusion mechanism is higher with the smaller nozzle, even when the motor speed is decreased to allow for the reduced throughput.
As Vik's experiments have amply demonstrated, the Mk II can do some fairly catastrophic things when it jams or otherwise misbehaves.
It seems to me like it would be worthwhile exploring using our rotary encoders and hall-effect current detectors to predict and avoid impending jams.
As we up the performance, the inceased nozzle tip pressure will be a nuissance. One possiblility: a different kind of thread on the rod, something with a larger thread depth like a drywall screw, or a setscrew connecting the inner rod and a nut on the outside of the teflon.
Another possiblity I don't think as much of: toss the internal threaded rod and use something that grips the teflon on the outside. There's a pipe adapter here that has a 3/8" female internal thread about 1/4" long, and tapers down to a nozzle with a 1/8 male external thread. It doesn't look as 1/4" is enough thread, but we could drill it deeper, and tap more thread.
via McMaster-Carr http://www.mcmaster.com/
Precision Thrd Hi-Pressure Brass Pipe Fitting 3/8" X1/8 NPT, Fem X Male Adapter, 3500 PSI, 1-1/4" L
Note: Please suggest a metric equivalent - I don't know a website for metric plumbing supplies.
I like the setscrew idea, it looks easiest.
I wrote an email earlier today asking if we might go back to cylindrical rod shaped feedstock rather than filament shaped feedstock, and hashing out the design considerations as I see them. I decided not to send it because I didn't think it would actually contribute to the discussion. (And I persuaded myself filament was better.)
Something I didn't consider - cylindrical rod would require the same pressure at the nozzle to get the same extrustion rate, but that pressure times the larger surface area would mean a much greater force on the brass nozzle.
Earlier unsent email follows:
Also, I've been wondering, why are we using 3mm filament rather than 5 to 10mm cylindrical rod? My central question is: how fast can we make 3mm filament at home? If it takes a big machine and 24 hours to turn 1kg of old parts into a 1kg bobbin of filament, the end users are going to have trouble recycling their scrap.
This is how I understand the design considerations: Filament works better for a mobile extruder head, and requires a lighter weight extruder head. Cylindrical rod requires a heavier extruder head, and a hopper/cartridge feeding mechanism. This may be the sticking point here: I think the Strasys used a cartridge, feeder, and a little bridge to the mobile extruder head, which just looks too complicated and error prone. Probably patented, too. We could go to a flexible chute going up to the hopper, where the chute would be a long series of tapered nesting hollow tube sections, like the chutes used in demolitions to guide debris into a ground level dumpster.
On the plus side, with cylindrical rod, you can use any kind of meltable feedstock - I'm not sure if we can turn beeswax into 3mm filament.
We have to fabricate the filament using a 3mm filament extruder. 5-10mm rod we can fabricate using a plaster or silicone ingot mold, a ladle and a double boiler, and the process is fast. 3mm filament requires a machine we haven't created yet, the process is inherently slower, and the scrap tank and 3mm filament head is inherently unreprap-able.
I'm willing to go ahead with 3mm filament, because I would rather have a simple extruder head and a complicated filament making-machine than a complicated extruder head and a simple silicone mold for the rods. And frankly, right now we really don't care about beeswax.
The last filament experiments I ran gave me a method of being able to produce good 3 mm rods. I simply took the heated extrusion barrel off of the auger pump extruder, locked it in a vise filled it with CAPA powder then heated it up to about 100 degrees Celsius. I then hand thrust a piston into the extruder barrel and collected the extruded rod/filament/whatever... One charge made half a metre of 3 mm rod in about 5 seconds.
That's nice and simple. Do you have a gauge to test the results? (A 3mm hole drilled in a bit of scrap; pass the cooled rod through the hole to check it doesn't jam.)
That would certainly be trivial to do. Simpler still would be to get a bit of bar stock, tap it for a cartridge heater with a thermostat, warm up the bar stock to maybe 55-58 degrees celsius and then run the rod/filament through it to smooth down the high points.
That might work. But we may find that it melts bits that are OK as well. I think a go/no-go gauge would be safer. No-go rods can always be recycled, after all.
I'd been imagining a device with a small heated vat with a nozzle on the bottom that fed a trickle into a pair of grooved wheels, and then through other wheels as needed. Like a rolling mill.
Any other ideas?
I'll have a go at bodging something together - eventually we' need something that produces continuous streams and recycles bulky objects.
I think a water cooling bath would be a good addition, too.
Log in to post a comment.