We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.

Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]

Russ Mitchell on the 3D printing revolution – and its current limitations

“Russ in Texas” (actual name: Russ Mitchell) commented most interestingly on this posting here about 3D printing, the point being that he had, or soon would have, personal experience of actually doing this stuff. I urged him to write about any such experience, and here (with apologies to him for the delay in doing this posting) is the email he recently sent:

Here’s my experience:

3d Printing is mature and ready to go NOW – if you need something in plastic, resin, or maybe ceramics.   If you need functional metal parts, the revolution is not here yet.

Background: decent-enough 3d modeling skills with graphics/animation software like Blender/3dsMax.
Tools Used:   TinkerCAD (godsend!), 3dsMax.
Formats needed by Pros: STL, DWG.

So, modeling arrowheads, etcetera, based on historical artifacts was not very hard.   In some ways, this was preferable to scanning because of distortions called by corrosion (holes in artifact), rust bumps, bits missing, etcetera.   TinkerCAD online proved to be REALLY fast for slapping together the rough models for figures based on intuitively jacking together various shapes (and then distorting them) – those who have difficulty visualizing in 3 dimensions might have trouble seeing how a pyramid, rotated, stretched, and then narrowed, gives you a scalene triangle, but it’s there and very doable.

The providers:   Sculpteo and Shapeways.   Their setup: entirely painless.   Their materials?   Affordable enough.   Some of the arrowheads can be duplicated for a couple of bucks a pop in resin or plastic, up to 10-12 bucks…. COMPLETELY affordable.


Write it off.   3d printing in metal is still OBSCENELY expensive (a 70-dollar arrowhead, made in 20-hrc stainless that can’t hold an edge??), and what I wound up having to do was take models to a guy I know with a laser/waterjet rig… who then recommended old-school forge dyes and stamping.

So that’s where we are now.   It’s coming, and for the right material, it’s here now:   stupid-easy modeling programs like tinkerCAD will get somebody 90% of the way to a useable model for simpler stuff, (almost) no skills required.   But the material’s the clincher.

The more I hear about this stuff, the more revolutionary (in a good way) it strikes me as being. And we are now only at the beginning of the story.

Tweet about this on TwitterShare on FacebookShare on LinkedInShare on TumblrShare on RedditShare on Google+Share on VK

9 comments to Russ Mitchell on the 3D printing revolution – and its current limitations

  • Russ Mitchell

    For those with deeper pockets, private suppliers such as GPI have come back with 15-5 stainless (a notable upgrade from the 420 Shapeways uses), at a steep premium — something on the order of six times the cost.

    Bounding-box issues for these are still severe, as well. The order of the day is still small parts: 100x45x25mm for 420 ss with Shapeways, for instance.

  • llamas

    The most exciting trends I hear about are for approaches that allow the use of plastics and other easily-3D printed materials, in combination with other filament materials, to replace metals entirely.

    Ideas like matrix printing, where plastics are co-printed with filament materials like carbon-fibres or metals. Think of pre-stressed concrete, of fibreglass, just on a smaller scale and a thousand times more-accurate. If you can print plastic material in a continuous filament, the next logical step is to print other filaments that are embedded in the printed plastic, to give it the properties it doesn’t have on its own. There are very few ways to do this sort of combination in conventionally-formed plastic parts – it’s hard to get a stiffener inside the part – but it’s easy with matrix printing.

    This doesn’t help you with the surface harness that metals bring, or the ability to produce strong high-precision features like edges and teeth. But it can get you very close the the structural properties of the metal – tensile strength and modulus – the point where many things that are now made of metal could be made of a printed composite.

    Fascinating, it are.



  • newrouter

    for the gun folks

    Get ready. It’s now possible to print weapons at home.

    An amateur gunsmith, operating under the handle of “HaveBlue” (incidentally, “Have Blue” is the codename that was used for the prototype stealth fighter that became the Lockheed F-117), announced recently in online forums that he had successfully printed a serviceable .22 caliber pistol.

    Despite predictions of disaster, the pistol worked. It successfully fired 200 rounds in testing.

    HaveBlue then decided to push the limits of what was possible and use his printer to make an AR-15 rifle.


  • Bruce

    NONE of the printing materials is suitable for making serious, highly stressed components.

    HOWEVER, you can make objects of considerable complexity. If made from the correct material, these objects can be and are, everyday, used as PATTERNS in the investment-casting process.

    Once you “tree” up the patterns, hie thee to a foundry that actually understands the performance requirements and stand back.

    A bit of heat-treatment and finish-machining (if required) and away you go.

    Anyone “printing” actual barrels is, well…………..nuts.

    Working with pressure vessels that run at 25,000 PSI and beyond, but that are made from “interesting” materials, seems to be an “interesting” concept.

    Many “real” maufacturers have been using variations on the CAM / investment casting, cold forging theme for a decade or more. The difference is that they employ the services of actual materials scientists and engineers to avoid nasty accidents and even nastier litigation

  • Milling machines can handle metals. Lots of small ones out there and the G codes to run them are very easy to develop. I was doing it over 10 years ago with home made electronics and software accurate to a light wavelength in a 50 meter machine (if the machine could step finely enough and hold tolerances). I could do a 7 axis machine with one parallel port. Adding 7 more required one more port. The calculations were so fast that a 100 MHz 486 machine could easily do 10,000 steps a second. If the hardware could keep up.

    Written in interpreted Forth. Some fun.

  • Printing a gun that can only fire two hundred times, instead of tens of thousands, becomes feasible when you can print the throwaway version for a few dollars… simply have ten of them printed up and ready to use.

    Materials and methods are advancing rapidly… there is no stopping the technology now. This is a disruptive change in ways we havent even thought of yet.

  • The ability to make your own stuff is a major challenge to the global supply chain. It puts a ceiling on how much the middlemen can charge. When the difference between mass produced and distributed vs printed at point of use is exceeded by middlemen costs, people will just print at point of use. Primary materials such as printer feedstocks won’t be affected but the vast majority of our manufacturing is under notice. They have a new constraint to their business models.

    Not being permitted such printers is 21st century serfdom.

  • Steven Groeneveld

    This discussion (unless I missed a similar comment) is missing something. It is not always necessary to print in the desired material of the end product. Some 20 years ago (when the then called Stereolithography was in its infancy) and the material weak and often dimensionally unstable I was involved in rapid prototyping shaped blocks for flat pattern pressing metal parts by making a mould of the part in a 3D printer, and casting in a harder material (hard epoxy for example).