Continuous print job

What if...You had a lot of time where you were at school, work, etc.
Suppose that even the biggest print jobs don't take the full amount of time that you're at work, your machine finishes and sits idle, It can't print because something's on its print bed already. What if...you had a conveyor belt as a print bed? While rubber is flexible, a plastic or wood sheet isn't. Let's get some modular tread like on this robot:
[www.solarbotics.com]
and put them paralell. Then wire up some motors, and superglue some plastic sheet to the top, in a way that, when flat, they make a single sheet. When going around the ends, they rotate free. So the Reprap finishes its print, spins the print floor, and moves the print off of the print zone. The reprap then continues to print, all we need is code that can have each model wait in line until its turn. I suck at arduino coding, anyone willing to try it?

Could you not source a long continuous sheet of plastic sheeting and use an inject printer as the feed mechanism? have it print a blank sheet every time the reprap sends an m30 command?

Edited 1 time(s). Last edit at 02/22/2010 12:01PM by JohnnyCooper.

I have been thinking about this idea for several weeks. One of the problems that I see is that it may be difficult or complicated to uniformly heat disconnected segments (tracks/treads). Also, the bed would not be one continuous sheet at that point, which may produce undesirable effects on the print (maybe warping of the "bed" in strange ways).

I will detail my solution to this in my contest entry when I register the project.. I am sure many people will have many solutions to this problem.

I have already posted my intention to make a conveyor for my mendel when it's printing.

Some of the things I've found out along the way is that the bed really must be incredibly stiff and immovable because the FDM process puts significant lateral forces on it. The printed parts also require significant force when being removed from the print bed. This will require a conveyor to be very carefully designed.

If you made a bed/belt that didn't require/exert these forces, you would not be able to print your first layer, or the object would peel from the bed after a few layers.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

I was thinking of a continuous conveyor belt made of kapton that would move over a heated bed.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

The problem is you need to hold it down all over so that the warping forces don't lift it. I have tried a vacuum, but atmospheric pressure is not enough to hold the corners of ABS objects down. I think rounding the corners would solve that. Fine for self replication, but not for general use where you might not want rounded corners. I also think Forrest's apron technique might work though.

The vacuum is a bit of a power waster though unless you can get it better sealed than I did.

---

[www.hydraraptor.blogspot.com]

How about a belt tensioner that would force the belting tight against the bed? It wold loosen when moving the belt. Then then, in the new posistion lever against it.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

It doesn't have to be that complicated..... There are many ways to create the necessary tension without getting overly complex. My method will use some appropriately designed rails and flexible but relatively large diameter wire as guide and tensioner. When I get the design up it will be more clear.

Unfortunately the necessary tension to keep a belt flat is infinite. See my explanation to Wes in the comments of this post: [hydraraptor.blogspot.com]

---

[www.hydraraptor.blogspot.com]

So then it is your opinion that it's not possible?

Afterall, infinite tension would be impossible to generate.

I am not so sure... not yet anyway. /grin

[www.dwbowen.com]

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Well you can get it almost flat with slightly less than infinite tension, but the tape would snap. That is why I explored vacuum and magnetic beds. They apply the force vertically rather than horizontally.

---

[www.hydraraptor.blogspot.com]

Ahh, I am not using tape as the tensioned surface. That may make a difference.

Hello all,

If anyone is interested this device [www.dwbowen.com] uses ABS 3.5" plastic discs connected to 4 inch nylon webbing. The conveyor is advanced by a 1 rpm AC gearhead motor. The webbing is stretched using a tensioning device and it seems to stay put during printing... at least enough for my purposes. The total length of the conveyor for this device is roughly 20'. I have also successfully used a 40' version.

Each time the device does a new print it centers itself. Before it begins printing the onion plant it puts down a 2" x 2" raft of ABS. This helps the prints stick to the discs as they get taller. Hope this helps.

cheers!

dwbowen

dwbowen, I'm very interested.

Would you be keen on putting up lots of nice big photos up here:
[objects.reprap.org]

(Maybe after the March 15 show if you're super busy right now)

I'll clean up the Example cruft from the page in a day or so if you don't have a chance to do so.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Yep, good points about the belt tension. I see that now.

Still brainstorming here...

How about something like removable heater beds. Would need to make the electrical connections by pressure contacts. Removed by a robotic arm? Spring ejection system? Explosives?

Or Heater beds on a tank tread like conveyor belt.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

You don't need to remove the heater. I just use a thin sheet of steel covered in Kapton held down by magnets: [hydraraptor.blogspot.com]

It can be lifted off, bent into a slight curve to release the objects and replaced. Just needs some clever mechanics to do that automatically.

---

[www.hydraraptor.blogspot.com]

On the belt / tension, all that would be needed to keep a flat surface is to use strips of trapazoidal, non-flexible material to which you adhere the film, a long 'foot' to press down on both sides of the edges (the foot presses down to apply tension during printing, and releases to allow the belt to be advanced.

Cross section as something like:

==================================== <- flexible print surface
\__/\__/\__/\__/\__/\__/\__/\__/\__/ <- aluminum or steel strips.
<- pulleys at the far reaches ->

Top view like:

__\                   /__
|  \_________________/  |
|   |   |   |   |   |   |
|   |   |   |   |   |   |
|   |   |   |   |   |   | <- reinforced print conveyer
|   |   |   |   |   |   |
|   |___|___|___|___|   |
|__/                 \__|
  /                   \<- clamps down, releases when belt needs to move

Heated bed simply placed under the conveyer; assuming aluminum strips work, heat should transfer fairly well; put the thermistors for feedback on the top under the clamps (?)

I am not following you BeagleFury. How do you hold the middle of something down by tension at the edges? As I stated earlier it needs infinite tension to hold something flat.

---

[www.hydraraptor.blogspot.com]

bjb's idea is exactly what had popped into my mind - just like tombstones for CNC machining. I was thinking of a removable heater bed with 4 'legs' that raise it off of the X-Y bed, with the legs also carrying heater current. Robotic 'forklift' would pull a bed off and place a new one on... The idea of chaining multiple ones together like a tank tread also occured to me, but I'm wondering if a circular carousel rather than a 'loop' might be better.

nophead Wrote:
-------------------------------------------------------
> I am not following you BeagleFury. How do you hold
> the middle of something down by tension at the
> edges? As I stated earlier it needs infinite
> tension to hold something flat.

I guess it depends -- is there an epoxy, cement, or other adhesive compound that will attach the kapton film to aluminum? Pretty much the same as bjb's idea, I suppose.

I whipped up a quick render in blender to illustrate the idea in rough form..

I see what you mean now. Kapton tape comes with high temp silicone adhesive that sticks to AL.

---

[www.hydraraptor.blogspot.com]

If those treads were made of springy steel I wonder if there would be a way to deform them to get the part off. Perhaps squeezing a section of track on the sides or perhaps going around the return axle. Maybe even some sort of scraper or spatula? What if the plate was cooled to a much lower temperature?

It seems to me with of these setups (except for a robotic arm) the bed would probably be stationary and the print head move on X,Y and Z.

I could see in a factory/assembly line setting that it may be advantageous to have the part stuck to the print plate for awhile for transportation and protection but for home use it would be an unwanted feature.

---

B^2 : [replibot.blogspot.com]

~~ We Are The Factory ~~

So why don't we make the belt out of magnetic sheet that is used to make refrigerator magnets and have something ferromagnetic on the bed?

I really think the heated bed is going to be too expensive(cost and power wise) for the Gada prize, so problems with heating shouldn't be an issue. Or maybe we can get the benefits of the heated bed with sufficiently elastic material on the bed(bed contracts with the plastic). Though this might not work if the forces on the bed are enough to deform it during printing.

Another option which might not work is to use electroadhesion instead of a vacuum or magnets. Electroadhesion can hold 0.8-2.3 psi(close to what nophead was running with a vacuum bed and a small vacuum pump) and takes about 20 microwatts per newton. Though this might not be enough to hold the conveyor belt down.

[www.sri.com]

Nophead, about how much force(vertically and horizontally) is exerted on the bed by printing? Have you measured it?

bjb - cooling the plate is probably right on the money. ABS has a coefficient of thermal expansion around 4 times that of stainless steel and around 3 times that of aluminum.

Quote

Nophead, about how much force(vertically and horizontally) is exerted on the bed by printing? Have you measured it?

No I don't have any measurements. With a cold bed it is massive. For example if you make a 20mm block of ABS and it wants to warp up in the corners. The force to hold it flat is that same as that required to bend 20mm of ABS.

When heated above 100C the force is much less, but still more than 15psi at sharp corners, as I have observed it lifting film directly over one of my vacuum channels. If you spread it though, with something like 0.3mm steel plate, then the average force must be less than that.

I don't know what the horizontal force is. With a vacuum it wrinkles the film and breaks the seal when it cools. With a steel plate and Kapton it stays attached so it must not be able to contract. The horizontal pull of the glue on the Kapton tape will be a lot more than a vacuum on thin film provides. In the later case it is only 15psi diluted by the coefficient of friction that stops the film sliding horizontally.

---

[www.hydraraptor.blogspot.com]

Following on form another thread discussing heated beds.

If you were making heated tiles there is no reason that you could'nt make a conveyor bed using segments that are in fact heated tiles.

You use your sticky kapton film as the surface of the bed and as the conveyor binder. ie it is a continuous belt stuck to the segments that are made of heated tiles.

You only want to heat the ones that are being printed on so may be sprung contacts might be enough.

Your conveyor rolers would be prismatic witht he face of each prism being the size of the base of the segment. When tensioned should give plenty of grip.

To eject parts run them to the end of the conveyor, If they are placed over an inter segment join the segment going round the end roler will attempt to detach the part. If you combine this with an off loading ramp/plate you may be able to get the part all the way off.

This method though may fail for parts that are smaller than the segment width.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

What about a cartridge loader similar to that used on Laser cutters / milling machines?

I figure you need a stack of pre-prepared (masking tape?) beds with a lip on the two Y axis edges so they will stack without the surfaces touching. The bottom bed in the stack slides in to rails in the machine, locked in place with a solenoid. A pusher (a bit like a desk draw) sits behind the stack to load the next bed.

The last one (now printed on) is just pushed out on to your bench top.

It may not be totally continuous but you could probably print ten beds worth at a time (with a big bench).

The bed could potentially be held in place by a spring clamp on either side which is lifted as the pusher starts to move, then released once it has retracted.

It would be pretty easy to build one for a Mendel.

Si

Why not just have different sheets of acrylic the size of the bed and have it take off one acrylic sheet when it's filled up and put a fresh one on.

I think it would be easier on one where the x and y movements are all done with the extruder. so the loading mechanism can be adjacent to the bed.

The approach I'll be taking is to have two print beds on a lazy susan, the table will rotate 180 degrees between prints.