3D Printing in the Woodshop?

[Jim Becker]

Wes, Rollers work nicely with metals because metals tend to retain the shape from the last pass of the rollers. I don't think that is going to work with wood because the properties are very different than metals. But you could make a fixture to bang out those laminations without dancing with a bunch of separate clamps by using fixed and sliding (tracked) cauls clamped with pneumatic/hydraulic cylinders pretty easily I would think. I'd use the CNC to design and build the fixture, but don't think I'd try to computerise the clamping. The nature of the adhesive used would indicate how long things need to stay in the fixture and to conserve energy, I'd design it so that once the pistons fully clamped things, some stops/wedges could be used to hold things closed without air/fluid needing to be pumped.

[Rob Young]

I'm totally ignorant about these things, but I can see how it would be useful to have one. Have some basic questions:

How do printed plastic parts compare to injection molded parts in strength and durability?

Are the surfaces smooth or does the layering leave a texture?

Can you just walk away and let it print overnight, or does it require any monitoring?

Can one print models for investment casting that will burn out cleanly?

The parts will not be as strong as injection molded parts. The layer lines are an inherent weakness. There are discussions of tempering and other methods for increasing the layer-to-layer strength. The best thing to do is think about the design and between that and the slicer software, control the orientation of the print elements to take advantage of "across" vs. "with" the layers. Durability is an odd thing because there are lots of filament formulations to choose from including some with carbon fiber. The best carbon fiber filaments have the fibers aligned along the length of the extruded filament and not chopped up tiny and random orientation. And as I said, the material types have different properties, PLA vs. ABS vs. PETG vs. ASA vs. TPU vs. wood-filled vs. blah-blah-blah. The final aspects of print strength have to do with infill patterns, density and the actual printing conditions, temperature, speed and cooling. Plenty of variables to play with!

The layers will appear as layer lines in a finished print. There are ways to control this during the printing process, again a combination of model design, slicing and the printer's physical aspects like nozzle size, temperature, feeds and speeds. Prints can be smoothed in post processing. ABS can be smoothed by exposure to acetone fumes. PLA can be smoothed a bit by annealing and heat treating as well as mechanical methods and the application of BONDO. Same for the other filament materials.

Yes you can walk away but it is pretty common to set up a Raspberry Pi (tiny, cheap, Linux based computer) with a camera and web connection to monitor & oversee the printer(s). Prints can fail because filament runs out, breaks, nozzles clog, power fails, heaters stop or run-away. Most printers are reasonably reliable once properly set up and calibrated but there are always losses. There are a handful of thermal run-aways every year, most just screw up the hot-end. Some cause much more damage. Besides the (remote) danger of fire there are other possible environmental hazards as it is possible that small particles of plastic are being released as well as fumes. Personally I don't like to stay in the same room with a running printer if it is printing with ABS due to the smell (if I can smell it, it certainly can't be good for long-term exposure) but also the noise.

Yes it is possible to make models for investment casting. PLA is commonly used for this. There may be other materials suitable too.

[Wes Grass]

Wes, Rollers work nicely with metals because metals tend to retain the shape from the last pass of the rollers. I don't think that is going to work with wood because the properties are very different than metals. But you could make a fixture to bang out those laminations without dancing with a bunch of separate clamps by using fixed and sliding (tracked) cauls clamped with pneumatic/hydraulic cylinders pretty easily I would think. I'd use the CNC to design and build the fixture, but don't think I'd try to computerise the clamping. The nature of the adhesive used would indicate how long things need to stay in the fixture and to conserve energy, I'd design it so that once the pistons fully clamped things, some stops/wedges could be used to hold things closed without air/fluid needing to be pumped.

Yeah, it wouldn't be fast, letting the adhesive cure out as you go might slow it down a lot. No experience at all with that technology.

And there'd be a fair bit of experimenting to determine how much to 'overbend' so it springs back to the shape you want. Which would no doubt be highly variable. Same with metal, but much more predictable obviously. CNC press brakes are amazing.

But if it would work ....

I've seen a lot of things proposed in my life that I thought would be impossible or impractical, that came to fruition. This, maybe will prove our doubts correct. Like I said, a 'dumb idea'.

Good cylinders don't leak to a level I'd worry about. Other than hydraulic ... as they say, the only hydraulic system that doesn't leak is one that's empty ;-) Its the nature of them, the seals distort under pressure and cycling it turns them into teeny little pumps. It's a necessity, if they didn't seep a bit they'd wear out very quickly.

[Alex Zeller]

I'm not sure about the roller idea but a CNC router could easily make forms of almost any shape to bend wood around. It could even make a mate for each side of the wood. For example if you were gluing up strips to make a curve and you knew the thickness was going to be say 1/2" you could have the CNC make both an inner and outer form with exactly 1/2" difference.

[Jim Becker]

I actually have made multiple cauls for bent laminations for a client on my CNC...the client tells me the desired radii for both the inside and outside ('have to account for material thickness of the layers) and I then draw them and cut them.

[Tony Latham]

I used a 3D printer to crank out my CNC.

[John Lifer]

Didn't know that a Good number of you guys have a Prusa. I started off about 3plus yrs ago with a Tevo tarantula. I still use it occasionally for PLA, but it is like a lot of these, tinkerer machines.
The Prusa MK3 machine is a great machine, the initial $s is more than a lot of folks expect, but it is Well worth it. Put it together and it just prints and prints.... PETG works great, and is strong.
And It can be pretty good structurally. I make a few parts that I could drive my car over. DC fittings, for sure. PLA will last a lot longer if not under stress than folks think. Add a bit more fill and it holds up well.