How fast can an $18,000 commercial system extrude....

I found this interest, as they extruded a cup with the goal of finding the speed breaking point:

[www.youtube.com]

For normal speed, also watch:
[www.youtube.com]
Also watch: [www.youtube.com]

The comments say it is a "Dimension BST 768" machine, these have a suggested retail price of USD $18,900 (and probably sell cheaper). I see references to ABS as the material.


Some interesting things in the comments:
==========================================
-- The printer prints in layers of .01" of an inch. A hair is roughly .002 to .004" inches in diameter. As I stated in the video the minimum recommended floor and wall thickness of parts sent to the printer should be .04"
-- I just finish a 25 hour print of a guy. The average tolerance is +/- .005"
-- I shot this original video with a cheap webcam that I placed inside the 158 degree F. print chamber.
-- This cup is 1 inch tall and is made up of 100 .01" layers.

Edited 1 time(s). Last edit at 12/31/2007 10:27AM by RoundSparrow.

Also found out this machine uses 1KG ABS cartridges that cost USD $250 each; this suggests to me that they are trying to make money on the selling of the print material also (much like Inkjet printer market).

Edited 1 time(s). Last edit at 01/01/2008 05:28PM by RoundSparrow.

Yes these videos have been my aiming point for what the sort of spec I want to achieve from my machine.

I am currently at 0.02" filament.
I have also found that 0.04" is about the minimum wall thickness that will self support.
I am on 0.016" layers.

Not sure how I compare on speed (currenly at pi cubic mm/s) but probably in the same order of magnitude for an order of magnitude less cost.

So someway to go after 11 months work, but looks achievable in the next year, probably with a move from HDPE to ABS. I am planning to make a heated table to see if I can get away without the chamber. They obviously don't include it for fun so we are bound to get some reduction in accuracy if we stick to room temperature unless we can find a thermoplastic which shrinks less than ABS.

---

[www.hydraraptor.blogspot.com]

I found one reference that said a Stratasys prints at 1 cu in/hr (not sure which model). That's 16 cc/hr or 4.6 cubic mm/s. Assuming .01" extrusion (.25mm), the head would be moving at 3.5 in/s (88mm/s).

If that's right, then it'd take 2.6 days to make the Darwin RP parts. Granted, it's at 4 times the resolution of what you're doing but it seems to be only slightly faster volume-wise. I wonder what would happen if you increased your head speed by a factor of 4.

Just how far can you take the "stretch the filament and increase the resolution" concept before running into problems? What would happen if you, say, extruded at 2 mm diameter but used a head speed 64x the free-standing extrusion rate?

64mm/s with 1mm filament would get me 0.25mm, i.e. the same resolution. However, my machine struggles to get to 64mm/s, but it is pushing a few kg around so a Darwin with the same electronics as my machine (anti-resonant micro stepping 48V chopper) probably could. I can get there with accel and decel ramps but the extruder does not react that fast so I would get too much plastic at the corners. I am at 24V full coil at the moment, 48V half coil might get me some advantage.

I think the only problem with stretching is the tendency of the filament to try to cut corners but tomorrow's experiments should confirm whether or not it increases warping.

To achieve their resolution probably just requires a 0.25mm nozzle and a lot more pressure. To get the build speed (less important to me) requires a light weight head and fast motors. Servos would have no trouble. Steppers with high spec electronics should also be able to get there as well.

If I can make the parts for a Darwin I will see how fast that can go with my electronics. That is how I plan to progress to my goal.

Edited 1 time(s). Last edit at 01/01/2008 07:12AM by nophead.

---

[www.hydraraptor.blogspot.com]

RoundSparrow Wrote:
-------------------------------------------------------
> Also found out this machine uses 1KG ABS
> cartridges that cost USD $250 each; this suggests
> to me that they are trying to make money on the
> selling of the print material also (much like
> Inkjet printer market).

That only half of it, the "Empty" spools still have well over a hundred feet of material on them! Not to mention they don't even have a recycling program for the empty cartridges. You are just supposed to toss them out (all that holds them together is 2-3 screws). The only thing that defines them as empty is a tiny obscure SMT IC. I can't find the one I had lying around right now but it really only had 2 pins: data and ground (even though it was a 6 or 8 pin package). It would sap power from the data line to accomplish the reads or writes to the internal memory. Its been driving me and my old teacher crazy!

Sounds like those empty spools could be a good source of filament. Is it 3mm?

The spools use 2mm filament. I wouldn't reccomend using the left over material for a reprap or repstrap, but perhaps one could easily "repurpose" the empty spools. It would be a trivial task to convert one to hold your own materials.