Before we dig into this one, a bit of a history lesson is in order. In 2010, MakerBot released the Automated Build Platform for the MakerBot Cupcake. This build platform was like nothing seen before or since. It’s a combination build platform and a conveyor belt for a 3D printer, allowing the Cupcake to become a completely automated production machine. Start a print, let the machine run, and when the print is finished it’s rolled off the bed into a bin, allowing a second print to start. If you’re using 3D printers for production in a manufacturing context – like Makerbot was – this is a phenomenal invention.
The Automated Build Platform was released under an Open Source license, then quickly patented by Makerbot. Since 2010, the idea of an automated build platform has been dead. No one is working on a similar device, lest they draw the ire of a few MakerBot lawyers.
This year’s Midwest RepRap Festival saw a device that’s an even better idea than MakerBot’s Automated Build platform. Yes, it’s a continuous factory of 3D printed parts, but there’s an even better reason for you to build one of these things: this printer has an infinite build volume.
This printer – it doesn’t have a name; this is just a one-off project – is the work of [Bill Steele] of Polar3D. The core of the build is just a hacked up MakerBot Replicator, but with one important difference. This printer has an Automated Build Platform tilted away from the nozzle at a 45-degree angle. What’s the benefit of this setup? Continuous printing and an infinite build volume.
Despite being downright bizarre, the mechanics for this printer are actually pretty simple. The bed is a standard MakerBot heated bed, rotated 90 degrees in the axis you would expect, then rotated 45 degrees in the axis you wouldn’t. A conveyor belt made of Kapton-coated paper is strung between two rollers and connected to a motor.
To produce a print, this printer starts at the very back and the very top of this conveyor belt. The first layer is added, the conveyor belt rolls forward a bit, and the second layer is added on top. The effect for each print is that the layer lines are 45 degrees from what you would expect.
When the print is finished, the belt just rolls forward until the part falls into a bin. Of course, since there’s nothing stopping this printer from producing a meter-long part on this build platform. [Bill] has already produced a 3D printed chain using this printer that was four feet long. Each segment of the chain just fell off the end of the printer when it was done.
There’s still some work to do with this idea. There isn’t a way to tension the belt on this printer, and [Bill] is looking for a material that’s better than Kapton coated paper. Still, this is the most innovative printer you can find at the Midwest RepRap Festival, and it’s not encumbered by the MakerBot patent on the automated build platform. You can check out a video of this printer below.
54 thoughts on “MRRF 17: The Infinite Build Volume Printer”
Glorious outside the box thinking. Even better that it doesn’t have a name.
More outside the box: tilt the whole device by 45 degrees, so the belt is level and heavier objects can’t peel off under their own weight.
I think that would defeat the purpose somewhat. If all you wanted was to print an infinite number of things, then yes, your suggestion would work, but if you want to print something larger than the printer, then you need the layers to overlap and I can’t see how that is going to work when tilting the whole printer.
Why wouldn’t it work? The mechanical configuration doesn’t change – just the orientation of the gravity vector.
dont need to tilt it at all, the motor driven belt and the rounded ( or made sharper) edge of the conveyor belt should be enough. If you ever messed with one of those roll fed sticker label printers, you would understand.
because this way there is always interleaved material in a meter long part, it is an essential property of the way the layer angle relates to the print bed.
it significantly reduces the effort involved in making parts larger than the bed, the alternative would be to have a slicer that took extended parts into account and printed in a special pattern to produce interleaved layers, this might easily make some parts hard to print, where the tilted bed configuration should work even on thin walled hollow parts.
@kakureru: I believe Dax is talking about tilting the printer set up so that the print bed is level, as the printed object may prematurely peel-off when it becomes heavy.
I don’t think you understand what I’m talking about, so I made a picture:
i think i misunderstood your original post, if what you mean is that one would still have the nozzle at an odd angle to the bed, then yes that should work.
that said i think it would need to be some massive parts for it to become an issue in the original configuration to begin with.
Oh bloody hell, just copy & paste the URL of the link
i gotcha, my bad.
nice concept drawing, anything that has been realized?
Well that was fast, and looks awesome [Dax]! Hope someone does it, if it hasn’t been done already.
even the belt platform can be shorter – i guess you just need a couple of inches, until the printed out filament hardens and is cooled down enough. as it rolls backwards, the printed out piece will peel off, so ideally just a supporting platform is needed that doesn’t need to provide adhesion anymore. but it needs to be low friction, so i could imagine a set of loosely placed rollers, and you’re set. it can be even more portable this way.
So I’ve been thinking about this ever since it was posted (I love the whole infinite y idea), and in my thought experiments, I found a glitch in your suggestion.
So like what used to be your Y axis has a point where it touches the print bed, let’s call that Y=0, even though it isn’t. When we tilt the motion system back, Y=0 is at the bottom of a hill. If the motors turn off, the extruder slides down, crashing into the print bed.
I am pretty sure the belt idea first came from reprap project, also a sand printer has this type of format.
Should have been an animated GIF of that chain being printed. :)
There should be a relatively easy way to tension the belt. Piezoelectric actuators would be super-precise.
A GIF!? Remember the last time HaD had a GIF on an article? All the comments were people complaining that a 1MB GIF was killing their bandwidth and that their homebrew TTL CPUs running CP/M were being bogged down rendering it.
There’s no reason to use GIFs. The last ones were 9~15MiB if I’m not mistaken. If HaD wants animated banners, then go for webm. GIFs should not be used anywhere.
By the way, I find animated banners distracting, so I’d rather not have them. If it’s inside the post (after a “Continue reading”), then I don’t mind.
These days GIF doesn’t generally refer to the original file format; increasingly they’re actually webm’s or mp4 under the hood.
If you write “9-15”, it makes no difference if it is MB (MByte) or “MiB”, only a few % difference. And with today’s bandwidths it makes no difference at all.
You could probably just print a four foot chain on a regular build platform, just like a snake game. Coil it up so it all fits on the platform, rather than hits the end? Can you really print chains with these type of printers like that?
You do not need actuators for this. Springs and a tension screw would be sufficient. The strain is much less than a chainsaw chain.
This method looks promising for printed food designs.
To be clear- It has infinite build volume because you have unlimited travel in one axis (until the unsupported mass of print peels the rest of your print off the bed). Good for printing long items. Pretty cool.
One can have a landing tray like in a paper printer, and the angle doesn’t necessarily need to be 45 degrees.
One can have a letterbox-shaped opening and a short belt just long enough to span the print-head, and tilt the device until the bed is level again, and you basically have a 3D printer that looks like a paper printer, but with a taller opening, which can push out arbitrarily sized objects.
Ah, Yes, That would work well. Would the angle (print head relative to conveyor style bed) cause any issues for the slicer software used?
Does infinite build volume require infinite filament? Not to mention infinite heat and electric power.
No infinite energy or heat needed. Just use RG8 for your heating wire, grounding the shield, and you should have a neat 50Ω heater filament.
They’re talking about total power used in an infinite print, not instantaneous power.
Love the idea behind the chain demo. You could use this machine to print yourself a new roll of filament!
Longcat is the new Yoda.
Someone needs to print a loooooooooooooooooooooooooooooooooooooooooooooooooooongcat sculpture on one of these.
print yourself a new roll of filament!
Hell, use this to turn one roll of filament into multiple new rolls of filament. Magically double your raw materials.
Then use one roll of filament to print another printer after which you can print filament twice as fast..
Now, that’s a hack!
So THAT’S what that was.
https://3dprint.com/10811/3d-printer-broom/ Patent issues for continuous production? Here’s a solution. I have patented this after the Makebot debacle, specifically so that I can avoid enforcing the patent. Easy build, I need to get off my ass and put the schematics on my wiki, in the meantime if anyone wants it, poke me.
So, when can we open-source the design on this, so that Makerbot can patent it as well? It’s obviously an improved design, and MB has no problem obtaining patents for prior art, so this should be easy. ^_^
First to file in the USA now! When does the patent office open on Monday?
That is so ingenious. One of those “I wish I though of THAT!
so why not add some more printers down the line for faster prints or other filament types?
Like Autodesk did with Project Escher?
Now thats hack for mass production of 3d printer plastic parts…
years and years we’ve been talking about perfecting the way to level the printbed and when all problems seem to be solved this happens… it makes me sad and happy at the same time.
But seriously, great idea. Although I slightly worried if there are alignment problems of this additional “axis”.
I saw an article about an industrial size version of this several years ago. The company was planning to create I beams. So the patents were probably established at least 5 years ago.
Can anybody explain to me how makerbot can patent something they seem to have previously already released to the public? And how no other company has gone “your own release constitutes prior art suckers” and released their own product? I looked up the automatic build platform patent and it seems to be just a generic description of what they had already released to the public by then.
You put a carriage on wheels. It has a telescoping arm that can also change vertical angle. The main base can also telescope up (and even have sections added as it goes, like an crane.) The base can move around the object and continually change the length of the arm and rotate around so that, in theory, you could make something as wide as imaginable (within the limits of tensile strength of the arm, of course) and as tall (within limits of the strength of the base.) You could, in theory, 3D-print another World Trade Center.
So is there a continuation?
So is there a continuation?
There are several projects that this design evolved into, including the latest, the Creality CR-30 3DPrintMill.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)