Project Tesseract

Good ol' plastic extrusion
User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Project Tesseract

Post by dragonator » Sat Sep 05, 2015 11:48 am

Named after a 4 dimensional cube, and somehow no one has thought of naming a printer this way yet. This is going to be the first entry of the development of my new FDM printer, Tesseract.

The idea has been there for a while now. How to make a printer that makes printing a lot of stuff interesting. For me there are 2 reasons to want this. One, I get an immense amount of requests for making and selling my props, and it simply isn't possible cost-wise for me to do so. The other is the immense amount of time some of my props take to print, most notably the Laser Rifle. The longest my printers can run on end is 6 hours, and the average is 1.5 hours. I would have to constantly restart my printers. This means that I effectively have 30 hours of printing a week, and those are 30 miserable hours.

There are 3 ways of solving this problem:

1. Have a really, really big printer. This way you can have 2 or 3 full days of printing in one printbed. This works in principle, but there are a few problems with it. For starters, the printer required needs to be huge, and not everyone has a lot of space. Secondly, reliability. Problems arise once every x hours, and if you have a printer that is working on 100 parts at a time for lets say 50 hours, the chance of something going wrong and ruining the whole batch is reasonable. Last, the printing quality by definition decreases every time you make a printer bigger. I do not posses the space for this kind of printer, nor do I like the dependability of the process.

2. make a printer that simultaneously prints the same part a few times. I know a company that actually does this with a dual extrusion 2x5 printer. It has a massive gantry with 10 pairs of nozzles, in a 2 by 5 grid. This way of printing is incredibly fast and has none of the reliability issues, but there is one flaw obvious. You can only print the same part multiple times. This would only solve on of my problems, and still take quite a bit of space.

3. Make a printer that can unload itself. This has been tried a few times before. you can try to slide the part of, put a blade under the part, or break it off, but the most efficient and effective way is to peel it of. The most known example is a conveyor: http://www.makerbot.com/blog/2010/09/13 ... d-platform

The Makerbot ABP worked reasonably well. The concept was to have a conveyor of mylar or kapton, and convey the part off the bed. At the bending edge, the printed part was peeled of the bed, thus requiring little force. But it had one downfall. It was not durable. The bed would warp after a few dozen uses, and remained semi flexible. It worked well the first few times, but quickly degraded.

User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Project Tesseract

Post by dragonator » Sat Sep 05, 2015 12:05 pm

Fast forward to now and why this topic is here. I think I may have a solution. Polymers by definition degrade under stress over time. It is what they do, so any plastic conveyor would degrade in under a hundrerd or so prints. Metal on the other hand lasts a lot longer. Easily in the 10.000+ range. Making a conveyor out of metal would be extremely hard, but a full loop of steel is not required.

Picture a printbed made of 0.3mm - 0.5mm (stainless) spring steel. Flexible enough to have a 50mm - 100mm bending radius. This is hung in a frame that has a coarse radius at both ends of the bed. This is to keep the steel from bending in the direction perpendicular to the bends (it will become more obvious later). When the part is done, this sheet is moved towards one of the bent sides. When the part hits the bend, it cannot flex with the bed, and starts to peel off. When the bed has moved completely over the bend, it moves back to its original position and the printer starts a new print again. Maybe some assistance with blades and scrapers is needed, but that will become clear once I have more experience to work from. More schematics will come soon'ish to accompany the words here.

I ran a test on my UP! with a sheet of 0.3mm spring stainless steel. The part stuck quite well and peeled of without effort. I am going to design a printbed to test on my UP! and see how well the part peels when the bed has massive adhesion, and also test how thick I can make the bed before forces become too big.
2015-09-05 13.24.32.jpg
2015-09-05 13.24.32.jpg (154.05 KiB) Viewed 10977 times
When these tests are successful, I will design a printer around this bed, aimed to be as reliable as possible. No tiny stamp size stepper motor drives, CNC drivers. Stiff frame, heated chamber, good printhead with bowden extruder, maybe bed leveling, all the good stuff. This printer is going to be roughly shaped like a tesseract if I can make it so, simply because I can, and if it works, I will also try to design an automated filament changer so I am also not limited to the size of my spools. I aim to design a printer that can print for weeks on end.

More posts when stuff happens.

ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Project Tesseract

Post by ezrec » Tue Sep 08, 2015 9:08 pm

Very interesting.

Have you considered printing on a roll of PVC? Roll from one end to the other. ABS sticks pretty well to it if you 'juice it' with a light spray of acetone.

User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Project Tesseract

Post by dragonator » Wed Sep 16, 2015 4:06 pm

I have not yet considered it. I want to stay away from plastic based belts because over time they warp, but it is mot definitely a possibility to print on PVC. I have no experience with printing on PVC itself, so I have to take your word for it that it is a decent printing material.

The current idea is the bed below. It is a sheet of stainless spring steel of 0.3-0.6mm thick. The is kept bent at both ends to prevent warping in the other direction (it cannot bend in 2 directions, only 1). On this surface a print is made. When it is done, the sheet moves on the guides and the whole part is pulled over the bend, peeling it off.

I want to mount this bed on either an ultimaker based frame (I am currently trying to get my hand on an unfinished incomplete ultimaker) or mount it on a delta printer, since I do not want the bed to be moving in anything other than a Z-axis.
TEBEX1 Bed WIP 2.JPG
TEBEX1 Bed WIP 2.JPG (275.01 KiB) Viewed 10888 times

ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Project Tesseract

Post by ezrec » Fri Sep 18, 2015 5:53 pm

I very much like the spring steel bed idea, now that I see it visualized.

Some questions:

* Are you going to make the spring steel a continuous belt, or a series of plates in guides?
* How are you going to drive the spring steel?
* If you plan to print in ABS, how do you plan to get good contact with the bed heater?

I would suggest a series of laser printer fuser rollers under the spring steel - could solve both your bed drive and heating issues at the same time.

I have a Delta now (cheap Geeetech thing, and only PLA at the moment), so maybe I can try some smaller experiments for you.

User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Project Tesseract

Post by dragonator » Sat Sep 19, 2015 2:50 pm

Impulse purchases are the best purchases (and also the worst).

I got the Ultimaker I was talking about. I not only bought it for this project, my old Mendel was also not really useful anymore. It is really something brilliant and exactly what I needed. The guy I bought it from made the frame himself on a CNC mill. The frame is slightly modified to give more room for electronics underneath. It has a RAMPS 1.4 and an LCD + SD instead of the Ultimaker controller, which means I have another motor controller (for the bed motor). It still needs a bed, but because it is a RAMPS, I can add my old Mendel's bed without any modifications.
2015-09-19 16.13.59.jpg
The old mendel that is very tired
2015-09-19 16.13.59.jpg (212.31 KiB) Viewed 10859 times
2015-09-19 16.13.47.jpg
New, shiny, to-be-finished Ultimaker
2015-09-19 16.13.47.jpg (250.32 KiB) Viewed 10859 times
The current order of business for this printer is:
  • -get it to work with all Ultimaker parts and my old aluminium printbed
    -Use up all my 3mm filament that I had left (Ultimaker is a 3mm printer)
    -Make my old 1.75mm extruder and hotend also fit on the side of the carriage
    -Slowly start printing with 1.75mm

    -Get and E3D printhead, and replace the air cooled body with a tiny water cooled block
    -Enclose the Ultimaker buildspace so it has a heated room (while protecting the hardware like motors)
    -mount the tesseract concept buildbed
    -renaming the Ultimaker to: "Ultimatic"
It may take a while, but this is going to happen. First I am going to get it in a printing condition though.

As for answers:
*I am making one solid sheet of steel my printbed. I can't really make it continuous due to the huge radius the steel needs.
*this one I have not fully figured out yet.
*I want to print in a heated enclosure. At 80C, a heated bed is no longer necessary. I doubt the ultimaker frame can handle this, but I will make sure the Tesseract frame can handle it.

A delta is actually even better for this concept because it has none of the moving axes on the printbed, but I wanted an Ultimaker for stability. When I get to testing, I am sure I can send you a bed.

ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Project Tesseract

Post by ezrec » Tue Jan 26, 2016 3:55 am

Here's another idea for an infinite bed printer:

* Printer is a core-xy style, but the core-xy gantry is movable on Z, bed is fixed - to the floor.
* Floor has a ring of bed material (glass, BuildTak, whatever)
* Fixed pole on printer to pivot in center of ring
* Single drive motor on opposide side of printer from the pole.
* 'Bulldozer' type blade on front on printer
* Empty area in frame on back of printer for parts to exit

Operation:

* Printer prints first layer on the floor build material
** Core-XY gantry moves up a layer, prints next layer, etc...
* Printer advances around ring the part 'exits' the back of the printer.
* Printer prints next part...
* By the time the printer is back at the first part position, the bulldozer blade will scrape off the cooled part.

User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Project Tesseract

Post by dragonator » Tue Jan 26, 2016 10:43 am

I for the life of me cannot figure out why this projects was put on hold. Funny you should post here now, I just picked the project up 2 days ago. Expect Tesseract related posts within a week. The first unmotorized test bed is assembled and will be tested this week.

The basic idea seems like it will work. Scrapers and 'bulldozer' like systems have been tried before. I cannot provide links now, but I have seen a few printers that unload this way. One with a blade that scrapes along the bed, and several mendel style printers that have a pusher on the printhead that pushes the parts off. It is funny to see a part being removed and flying a meter away.

Bulldozers have been used to automate 3D printers for quite a while now, but they do seem to have trouble with bigger parts and fragile parts. If the bed adhesion is too high, the part will require more force than it can handle to remove and the part might break. The frame itself also needs to be tough to handle the scraper's force. For smaller, taller parts it works just fine though, which is the reason some people make them.

The ring itself seems like a difficult part, and I do not know how much cooling the parts need. Is there any added benefit to a ring opposed to a bulldozer pusher on a normal printbed?

ezrec
Posts: 112
Joined: Sat Aug 15, 2015 1:31 pm

Re: Project Tesseract

Post by ezrec » Tue Jan 26, 2016 2:42 pm

It would be more of an art project than a "real idea".

Of course, you could have multiple machines doing separate tasks in the ring:

* First machine (FFM) prints some layers, with open areas for embedded fasteners, wire tracks, and electronics.
* Second machine (adhesive extruder) lays down some epoxy for pick & place parts.
* Third machine (Pick & Place) marches behind it, with pick & place for adding non-FFM components
* Fourth machine (conductive trace extruder) lays down some conductive material to connect electronics

.. and the first machine is back to add more plastic and continue the process.

Fully automated additive assembly.

User avatar
dragonator
Site Admin
Posts: 408
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Project Tesseract

Post by dragonator » Tue Jan 26, 2016 10:38 pm

That puts it into a different (and a lot more creative) range of machines. That allows for much more than just a 3D printer.

Imagine 3D printing and assembling robots, that get of the assembly line working, to give just one idea. You could even go back and forth between stations if the prints are a bit more complicated. Even better if something can FLY off the assembly printer.

Not really my type of projects, but I would be glad to help in the creative process. I like the idea. It would be fun to see in real life.

Post Reply