Scalable 3D Printer

Question

Is there a definitive scalable 3D printer?

I've seen examples of Chinese companies printing entire houses, and I'm curious as to printers / filaments that are intended (or at least able) be scaled up for (very) large print jobs.

Since most hobby printers can take hundreds of hours for something that can still be held in our hands, so I'm curious if there are any designs for printers that are meant to extrude material efficiently with a easily scalable printing area.

Open sourced / free is preferable; though I'm interested in any designs that exist, commercial included.

Answer 1

One big challenge with scaling anything up (or down), is that not all properties or characteristics scale linearly.

Consider a trivial case: a small cube. If you double the size, you've quadrupled the surface area and octupled the weight.

If you take a desktop-sized 3d printer design, and just double the size, it will weigh 8 times as much. But all the bolts holding it together are only 4 times as strong. So weight-bearing bolts are effectively under twice the strain (per unit of cross-sectional area); the exact factor depends of lots of things (orientation, leverage, etc).

If you scale up by 10x, they'll potentially be under 10 times greater stress, and probably snap.

Many, many parts will have similar issues: Drive belts that are twice as wide and twice as thick will still have 1/2 the (relative) strength -- and stiffness for the motors to overcome.

The stepper motors have to move 8 times the weight, and you'll want them to move far, far faster. But steppers lose torque when you spin them faster.

My main printer is pretty much a MendelMax 2, but the Y axis is about 6 times bigger (X and Z are normal). On MM2 the whole build platform moves along Y -- on my printer it weighs far more than on a regular-size MM2 -- it's 6 times the mass just because of size, besides that it has to be much stiffer to avoid sagging over that distance. I went to a larger motor, but it could still barely move the axis. I eventually got it to move at pretty normal speed, but to print long objects it really should be 6 times faster or so.

That would require an amazing motor -- and really big electronics to drive it.

Another challenge is the printing material -- 20mm diameter PLA fiber would be really impressive, but a little hard to find, and a spool would be hard to lift. The power needed to melt it fast enough would be impressive, too. Bridge and house printers I've seen in reports, use pumped concrete through about a 100mm wide nozzle (I have no idea what the nozzle is made of, to stand up to the abrasion). See Watch this giant 3D printer build a house for some cool pictures and video.

Scaling up is a fascinating project -- but it's harder than it looks.

Answer 2

If I understand your question correctly, it sounds like you're looking somewhere within the RepRap realm. The RepRap community is mostly responsible for the boom in consumer 3D printing in the past 10 years, and that's most likely because it's open source. RepRap designs are mostly dynamic (and most parts can be 3D printed), so you could theoretically build a larger frame for your machine and use a slicing engine that allows you to set the build volume. I believe Slic3r allows you to customize the build space, I'm not sure though.

Answer 3

As far as I know, right now, all the "3D printers" that can print houses, bridges, etc. are experimental models (and sometimes, vey elaborate art projects) - they just don't exist except as one-off creations designed as a proof-of-concept showing this can be done.

Maybe some of those project published their plans and code but they are not designed for mass production or general use, even if you manage to build another copy of one of them it will probably be able to print just the same project the original printed.

You specifically asked about speed, material and build area:

Speed: the speed really depends on the size of the printer's nozzle, the typical desktop printer has a 0.4mm nozzle, if you replace it with a 1mm nozzle the printer will be approximately twice as fast (1.25 nozzle area vs 3.14 nozzle area, assuming you have an hotend designed for larger nozzles that can melt the plastic fast enough).

Now, the "house printer" does not use a standard desktop hotend (see material below) and buildings don't have fine details so you can make the head much larger.

However, as far as I know the early prototypes are still much slower then conventional construction methods.

Material: houses are typically not made of plastic, the printers I know of have welding equipment instead of an hotend and extrude steel

There are probably other methods but I bet all of them use typical construction materials such as steel and not thermoplastic.

Size: and last but not least, it's obviously not practical to build a printer with as house sized platform, the house printers are actually relatively small robots that travel across the construction project, leaving material behind them and then climb over the previous layer to print the next part.