Magnum Opus is born

Hello everyone, I would like to build a large printer. The size I have in mind is approx. 8 to 10 feet long by 4 to 6 feet wide by 6 feet tall. If possible I would like to have numerous heads on this machine. Probably one every foot. I have no problem building the machine. What I will need help with for sure is controlling it. I am new to the forum so any help would be greatly appreciated. I have just ordered a Prusa i3 as my first unit. The unit I will be building right away will be a single head version of the big guy, and build the big guy bit by bit as I progress more and am sure that is the best way to set it up. I am not sure about the head number but for sure it has to be multiple. Maybe one that does the edges and a bunch that fill in? One head will be way to slow at this size. So who ever knows about that I need your help. Tomorrow I am checking out some ibeams. I wanted to use DryLIn style bearings, its like a slide shuttle that moves on a rail. The shuttle has a very tight fit riding on the contoured rail. I was thinking about using proximity sensors instead of limit switches. Drive system I am not sure weather to run belts or a direct gear drive on a toothed rail. So that's where I'm at. I called regarding the ibeams and they will be getting back to me later today or tomorrow.

The i3 has a bed moving in the Y axis. That isn't something you'll want to scale up to 8'x4' or even larger because the moving mass will be so large it will be difficult/expensive to control at any reasonable print speed. You want to look for an architecture in which the bed doesn't move at all or only moves in the Z axis. Look into CoreXY, QuadRap or even at delta machines.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Agreed, the x and y will be on the top beam. Z up to the moment is unknown, I am trying to avoid moving the table if possible as you say for weight purposes.

I've sketched up some ideas for multiple printers on one bed. Especially how they would handle the "collision-zone".
It would require three digital signals more between each neighbor printer and some extra coding.
-Olaf

Well this is excellent. A bunch of people here coming together making something happen. That is good that you have the understanding of this because I am lost at the moment on that subject, coming up fast but still far behind. Hopefully you can continue to work on this while this thing comes together. one suggestion for collision zone would be slow one guy down until the other moves out of the way. maybe it can not be allowed to enter that side by side zone until the other is gone. I think that is what you meant with the flags. This thing is 8 feet long so print zone can be side by side so each one has a 2x4 section if it was a 4 head unit or 1x4 for a 8 head. It can also be 2x2 with 4 heads on each side of the machine or more but heads of both sides would be easier for me to make. I would like each one to be as small and light weight as possible. They can each start at the beginning of their section and move to the end at the same time so they should never end up together. The other way could be length wise. So 8 feet long by 1 wide, 4 head side by side length wise. You can have one take off first, give it some time to move away then the next and so on. A 16 head would be fast at 8 feet by 3 inch. So what ever happened with that did corpse build it?

Edited 4 time(s). Last edit at 12/01/2015 04:18AM by MagnumOpus.

A question about the print area size, what determines this. Is it the physical size of the machine because it cant move any further or is it something in the firmware or other program?

The only reason to make a big printer is to print big parts. If you're going to divide the printer up into smaller areas you're better off making multiple smaller printers that can function independently. That way, if one is down with a problem, the others can continue to function.

When you talk about making a print head cover 8' x 3" area it sounds like you're trying to rasterize printing, similar to the way an inkjet printer works, in this case, depositing plastic as the head passes by instead of making the head follow the shape of the object. I don't know of any off-the-shelf firmware that is written to do that. The extruders we use for FDM printing are designed to deposit plastic in a continuous line, not discrete dots.

Print size is limited by the bed size and/or the limits of motion of the mechanism, whichever is smaller. There may be some limits in the firmware, but I've never heard of anyone running into them. You'll need to ask firmware developers about size limits baked into firmware. For a large machine you will want to skip the Arduino/RAMPS stuff and go to a 32 bit controller like Duet or SmoothieBoard.

It sounds like you're not looking for FDM printing at all, but something else. You're going to need to delevop the machine, the extruder(s), and the firmware.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

That idea was suppose to print big parts just working in different areas on the same z axis. These parts would not be joined later, it would happen right away. I have been watching different styles of this machine run in my head for a while now and I see many weakness points. I think if a large structure is built separate like that even though they are joined it will be weak. This needs to be able to over lap significantly. Specially if your running a carbon fiber or fiberglass through it as well which I would like it to do. Heat differentials, delamination and warping. This head setup is going to be very tricky. Once I can see a head setup run properly then someone who understands firmware can have a look at that point. Currently I am thinking about somehow doing the edges with some fine extruders and have another set that comes in after and fills in the centers. I will be able to develop the unit, what I know I am totally missing is the firmware. Olaf sounds like he's onto that part. You reminded me of some printers we have at work. They fire 80,000 dots per second of electrically charged ink, a magnetic field manipulates the stream and it prints something off, letters pictures what ever you want. The print head does not move it deflects the ink stream. If you could rig that guy up to shoot a material for the 3d printing world it would interesting to see what it could do. This is suppose to be FDM, to lay out fiberglass or carbonfiber, that is the end goal.

Edited 2 time(s). Last edit at 12/01/2015 07:35AM by MagnumOpus.

How many years do you expect the development of this beast to take? What is your budget like?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Years!?!? My friend I do not have that kind of time or patience to mess around, this is happening now. Months at the most I hope. I run into worse things than this at work all the time. There I have a team of electricians that assist me when I have questions. I need them to go into the PLCs and mess around with the program. Anything can be done, question is how. This is the same. I will make the machine but I will need the people here since they are way more experienced than myself to help me with the programing. Sounds like firmware at this point. I will be heavily relying on the firmware guys for sure. Budget I do not have, what ever it takes, I do not have endless funds which would be nice but I will spend what ever it takes to make this rig. I will do my best to keep the cost down as much as possible but will never sacrifice quality.

Things are starting to get ugly. This is not looking good for simplicity. I am only visualizing 2 methods at this moment that will work for the big unit

Option 1 - 2 sets of print extruders. First a fine set that will come out first that will handle the edge detail and internal carbon fiber structures if there are any. Then the second set comes out and fills in the core with a much larger diameter nozzle. This will have a stationary table and the extruders traveling along the 8 foot length.

Option 2 - This will involve rotation, Either the table or the extruders. This one can have a high number of extruders and I think will rely heavily on material feeding speed. They will all have to be operating a different speeds to compensate for the table motion. The other option is for the extruders to rotate but that means everything will be rotating. I think rotating the table is to much again because of the weight. Therefore the top must rotate. Which throws a big giant wrench into the mix.


So that's the scoop for the moment.

Edited 1 time(s). Last edit at 12/01/2015 10:04AM by MagnumOpus.

This sounds like an idea instead of specifications.
A short deadline with no design is rife for failure.

You are correct it is an idea at this moment. There is no deadline only hopes of a completion that is not years and confidence it should only take months. Failures I am 333% positive there will be many, many failures. I am expecting murphy's law to be completely enforced, but I will win in the end. Man vs Machine, I am ready for battle.