Large Format Printer

I am doing a cost analysis for a large-format 3D printer at work. We want to know the likely cost to build and operate a 10' x 10' x 4' tall printer. (Yes, those are feet.) I have seen many printer systems, like those from Objet, that can make small hand-held parts. Then there are custom systems that can, "print a house," but nothing in between.

Specifically, we are looking into installing a machine that can print patterns for making molds. These patterns would be used in a sand casting system for the production of large steel castings. Right now our patterns are all wood. Whenever we have a new pattern made, the turnaround time is about 6 weeks. Patterns are about $10k-$40k and up. If we could print a polymer (or other suitable material) pattern for near or less than $10k, and in less than 6 weeks, then we would have a winner.

Note, if 10' x 10' is a little too big, we could probably use a 6' x 10' or a similar size. I don't think that the table size will be an issue though. The more likely problem, I'm guessing, will be the cost of the polymer material and finding someone who can make such a machine.

Do you guys know of anyone who can quote on such a system?

Thanks,
Brian

Maybe contact that guy, his machine build area is bigger (15.86 x 15.86 x 12.70 inches) than your needs : [www.grassrootsengineering.com]

Edit : The mendel90 design would probably be easy to extend to that size too. See [hydraraptor.blogspot.com]

Edited 1 time(s). Last edit at 03/15/2012 04:40AM by DeuxVis.

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

Thanks for the reply.

I'm actually pursuing something that is several FEET wide, not inches.

Oopsies. Sorry. Not my usual units

I remember watching a video of a gigantic robot arm moving an extruder around, printing some furniture - chair, table... Might be interesting to you but I suspect your 10k $ limit was largely broken here.

Anybody kept the link to that ?

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

What kind of accuracy/layer height would be required?

---

www.Fablicator.com

Hi Andrew,

3/64" (or 1 mm) layer height would be acceptable. We would want less than 1 mm deviation from the desired tool path. It looks like these numbers are easy to beat in the world of small parts printing. Hopefully this is not too difficult to obtain in large format printing too.

Well there goes my simple idea :p

I was thinking if 1/4in layers were acceptable, you could use a large format laser cutter to produce the "layers" with internal alignment guides for steel bars. It would allow very large and moderately precise parts with just a laser cutter and some software.

See the link for a good example: [www.thingiverse.com]

All the current plastic deposition methods we use have expensive materials, and would have unacceptable warping at the scale you are looking at.

What kind of castings do you make? At 6x10ft I would assume large power plant parts (ship engine components, turbine parts, etc)

---

www.Fablicator.com

You're pretty close, Andrew. 90% of our product consists of liners for the mining industry. These are liners for rock crushers and other processing equipment. We also make large pumps, like those used in municipal water supplies.

That laser sliced Cthulhu is pretty neat. What is the deal with all the Cthulhu stuff lately? I think we would be right back to subtracting material again though. We have a dilapidated CNC router right now, so we could just replace that thing with a modern CNC mill instead.

That's a shame about plastics not lending themselves to larger projects well. I suppose a pattern could be built on some kind of scaffolding to reduce warping. I wonder if simply adhering ABS to a steel or wood backing could solve this problem? Also, the patterns are rarely 4' tall. I just want to make the machine versatile.

Costs are comparable. If we were to make a 0.5m^3 pattern from ABS, that's about 500 kg. ABS is about $40/kg. So that's a $20,000 pattern, which is a little higher than I would like, but not necessarily more expensive than a wood pattern.

What material(s) can the patterns be made out of? Also, what wall thickness are the patterns? Are they one sided, or two sided?

A 1mm deviation is nearly impossible on a 10 ft (or 3048mm) = ±.033%. Not even injection molded parts off the same mold can hold that tolerance (especially when you change temperatures). Although the printer could be made to hold the locational tolerance, plastic is plastic - so it warps and shrinks. There are tricks to accomodate shrinkage, but on that scale it would be quite the trick.

I have some used THK linear slides that are 13 ft long (with a working load rating of 9000 lbs each), very pricey when new but also VERY smooth. The rails alone were almost $1 per mm each (yes almost $4000 each). I have been thinking of a large scale printer design, but the shrink factor would have to be compensated if using plastic. I have 4 plus some that are 8 ft long, so the gantry isn't a problem, the software and hardware aren't the problems, making the material "behave" is.

Where are you located? I'm in the US.

I'm in Arizona, in the Phoenix Metro Area.

We use two patterns for each mold - one cope pattern, and one drag pattern. There is no wall thickness associated with out patterns at present. The patterns are solid wood. Each pattern looks exactly like the top half or the bottom half of a casting. The pattern probably cannot be hollow because it needs to support a ton or more of sand. The patterns themselves are probably about 1 ft thick in areas.

And you're exactly right. The mechanics of a machine that can do this are not really a problem. The machine ends up being very much like a large CNC mill.

It's possible that plastic is not the best material for this. I don't know if the shrinkage of ABS or a similar plastic will be a problem. I will need to defer to others in the plant for that. We use wood now, and obviously that shrinks and warps too, though perhaps more with moisture than with heat.

Brian

You might be better off cutting it out of foam. It can then remain in the mold, as it evaporates during casting (lost foam casting) see
[en.wikipedia.org]

Another idea : Why not buy a shopbot (I was told it's a pretty cheap machine for its abilities) and mount an extruder on it ?

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

You bet! That's how we would handle this project. We may not use Shopbot. There are quite a few CNC options.

On another note. We did some volume calculations. A typical volume for a pair of patterns (2 patterns per mold) would be 7500 cu in.

We would probably print shelled patterns with a 1/4" wall thickness or something like that. The shelled pattern would have supports under it, probably in a cross hatch pattern, and probably with a 1/4" wall thickness again. That volume would be 2000 cu in.

Based on the cost of a spool of ABS, the cost of such a pattern(s) would be much less than than the cost to have a wood pattern made for us. The question is how much it would cost us to hire someone to operate the machine, pay his health insurance, upkeep on the machine, etc.

at 5in^3/hr (more than triple the typical value) it will take over 16 days to print the object.

With existing software there is no way to restart the print from it's point of failure if it fails (kinked filament, power loss, comm error, running out of filament, clogged nozzle, etc) You will need to modify it appropriately and likely have some form of power backup.

I will reiterate my doubts about the suitability of any plastic, too. At that size warping stresses will be very significant without a heated chamber which is still under patent (use at your own risk)

If you do go through all the trouble to get it working, it will be very impressive though (and the worlds largest FDM printer)

---

www.Fablicator.com

simply not practical.

Why not just print the molds directly?

BAM!

[www.youtube.com]

That's amazing! Printing hundreds of molds would take too long though.

Welcome.
This topic has been several months. If you are interested in large format is still in touch with me.
I can help you in such great prints, I designed this printer.
Exactly such a system printer.
The first stage finished first built the smallest printer in the system print dimensions 480mm x 460mm x900mm
Printer plans x1000mm print dimensions 1000mm x 2200mm.
next machine is 2500mm x 900mm x1700mm
I know that is not enough for your needs, but plans can change.
Also, I'm going to print forms.

Look at Voxeljet (http://www.voxeljet.de/systems/produkte/3d-drucker/) They aren't cheap, but they are big...

I was asked about a large (at least 3 ft. x 3 ft. x 4 ft.) 3D printer earlier today and have been researching. The problem I see with the Voxeljet is the material strength. It is less than 1/10th that of FDM ABS (4.3 MPa v. 52 MPa). That's just not acceptable for most applications. Sure its 4 m x 2m x 1m build size is impressive but I fail to see how the models printed could be used for much of anything. If you printed a chair with their plastic it would fail if a 200lb (90kg) person sat in it and put any kind of stress on it. I guess it would be good for making positives to then be used as casting models, but that's about it. Might be a good option for the OP, but makes me sad like bull for other applications.

I am looking for a large format 3D printer, for a build-up of 2,000 x 1,000 x 800 mm.

The main products to produce are plastic parts for the automotive industry, hence the pieces should be able to resist any usual movement that would relate with a car driving down the motorway.

I am in touch already with Voxeljet, but any other suggestions are welcome.

Thanks,

Geni
geni@straudico.com

Hi, I am looking for a printer about 1000 x 800 x 500 mm But using FDM ABS plastic