Additive and Subtractive

[Generic Default]

Additive manufacturing (3d printing) has a bunch of advantages over subtractive, but no type of additive process can give the same tolerances and surface finishes that machining can.
So while most of us here focus on thermoplastic 3d printers, this is what cutting edge machine tool manufacturers are doing;


[youtube]http://www.youtube.com/watch?v=s9IdZ2pI5dA[/youtube]


What are your thoughts on this? It wouldn't be that hard to put a tool changer with a rotary tool and a hotend onto a hobby level 3d printer, right? We would need more rigid machine construction and new slicing programs, but that's about it.

[teoman]

Dremel has an extension. So you could have the motor up top and the cutting tool down below.

If you had magnetic arms, you could make a mechanism to change the effector plate automatically.

[Polygonhell]

You'd probably want to avoid having to reorient the part by hand and that means 5 axis machining, which complicates the mechanics, but the real issue becomes software, I know of no reasonably priced software that will produce 5 axis tool paths. At one point I looked into doing this and it's very far from a simple problem, MUCH more complicated than the 3 Axis case, or the trivial 4 Axis cases.

[Generic Default]

5 axis control really isn't that hard, you just need to do some trig to get the tool positions and stuff. I was thinking that printing a few layers at a time, then contour milling the edges would be a relatively easy way to do it.

The software has always been a bottleneck for open source stuff like reprap. There are very few programmers who have the skills and are willing to put in work for free. But look at Cura, Slicer, Kisslicer, and MatterControl.

They are all just as good as the proprietary software that industrial printers use, and they have improved so much recently.

[Polygonhell]

5 axis control really isn't that hard, you just need to do some trig to get the tool positions and stuff. I was thinking that printing a few layers at a time, then contour milling the edges would be a relatively easy way to do it.

The software has always been a bottleneck for open source stuff like reprap. There are very few programmers who have the skills and are willing to put in work for free. But look at Cura, Slicer, Kisslicer, and MatterControl.

They are all just as good as the proprietary software that industrial printers use, and they have improved so much recently.

The control part is easy in any number of dimensions what's hard is path planning, you need to take into account potential collisions and the shape of the end mill, I've written a toy slicer and some trivial milling software, IMO though the operations are similar slicers are simpler than the subtractive equivalents even in 2.5D.
In general in higher dimensions, many of the geometric operations are not trivial (fixed offset from a surface for a given path), and there are numeric stability gotchas, along with dealing with degeneracies.

Even a lot of relatively expensive milling packages only deal with the trivial on axis 4D Operations.

I'm not saying it can't be done, but in my experience software in the OSS world tends to evolve to meet the direct needs of the primary contributors and not very much further.

[Flateric]

I have come to the very same conclusions regarding the 5-axis milling machine delima. Control and construction of a 5-axis is within reach of a semi-skilled diy fabricator of cnc machines for home use. However software for creation of these toolpaths is very expensive, hard to learn/configure and very expensive.....I know I said that twice, but it's such a big point here I thought it worth mentioning 2 times!



But ya, maybe oneday well have an option. After all, there was a time before mach2/3 products arrived that even controlling any cnc machine was outragously priced. And if you had stated to anyone that really decent FREE cam design software would arrive they would have locked you up.

[Flateric]

The video shows a really impressive machine however.

But am I the only one who thought, gee, this cool hybrid machine isn't really doing anything a good cnc lathe/mill machine couldn't do with a billet piece of metal as a start point. Using only subtractive strategies?

Also I could be wrong, but I bet a number of cnc mills and lathes with specialized tools could be purchased for less than this one hybrid machine.

[RegB]

The video shows a really impressive machine however.

But am I the only one who thought, gee, this cool hybrid machine isn't really doing anything a good cnc lathe/mill machine couldn't do with a billet piece of metal as a start point. Using only subtractive strategies?

Also I could be wrong, but I bet a number of cnc mills and lathes with specialized tools could be purchased for less than this one hybrid machine.

Agreed, probably faster as well as cheaper - - and I would bet there are some thermal contraction problems to account for.

I guess all I have to do is figure out how to rig up my little Home Depot Lincoln MIG welder to my Rostock... not today.

[Mac The Knife]

Not practical if you're using"normal" materials, But I recall seeing somewhere that they are making that part out of titanium.

[Brian]

Would the cheapest to create method that gives the most dramatic results come from simply making a 3D printer with a big fat nozzle on it and a precise cutting tip that could drop down and retract?

The process would be to deposit a heavy layer of plastic very rapidly and then use the cutting tip to do a perimeter pass to create finely detailed edges and sharp corners and that kind of thing? It seems like it would speed up the process a lot, and improve resolution at the same time.

Maybe you add a lepton imager to the tip so that the cutter would wait until the layer cooled the right amount before trying to trim it to shape?

[Generic Default]

For most materials like steel and aluminum, regular billet machining is faster and cheaper. Plus the machines cost less (still a lot though). But if you're machining something like inconel or titanium or cobalt or any other expensive alloy, the cost of buying a giant billet and removing 90% of it may make it more practical to use a hybrid machine like the one in the video.

Laser sintering may work too, but keep in mind that the porous structure of sintered parts can be undesirable.

Keep in mind that just a few years ago before reprap caught on, all 3d printers cost tens of thousands or hundreds of thousands of dollars and were completely proprietary. Now the cheaper ones are down to a few hundred dollars for FFF type printers, and photoresin printers aren't far behind. I really believe that the cost of these systems (and the software) can be reduced by powers of ten if there is enough demand and enough innovation.