Find all your DIY electronics in the MakerShed. 3D Printing, Kits, Arduino, Raspberry Pi, Books & more!


3D Thursday is a feature about CNC Machining, 3D Printing, 3D Scanning, and 3D design that appears in MAKE every Thursday. 


This miniature, high-precision assembly started with a CAD model and not much more. It cost about $10 to make it at home – with no 3D printer required.

In the past couple of years, the concept of low-cost 3D printing has captured the hearts and minds of millions of geeks. The allure of an upcoming manufacturing revolution has seeped into the mainstream, too: take The Economist, which ran about two dozen articles about this technology within the last year alone. Something must be in the air!

The charm of 3D printing is easy to understand, especially as it coincides with the renaissance of the DIY movement on the Internet. But all this positive buzz also has an interesting downside: it makes it easy to overlook that the most significant barriers to home manufacturing run very deep, and probably won’t be affected just by the arrival of a new generation of tools.

After all, affordable and hobbyist-friendly manufacturing tools that convert polygons into physical objects have been available for more than a decade. Take desktop CNC mills, for example: home- or office-friendly and costing about as much as a 3D printer, they have revolutionized the lives of many jewelers and dentists; they have shaken up quite a few other niche industries, too. But spare for a small community of hobbyists, these self-contained and tidy mills have not brought on-demand manufacturing into our garages or living rooms.


Roland MDX-15 – a desktop-sized, enclosed CNC mill popular with jewelers. This model debuted on the market around 12 years ago.

CNC mills and 3D printers are different in many ways, but they also have a lot in common; and looking at the parallels, it’s reasonable to suspect that the prospects of home manufacturing may have relatively little to do with the choice of a particular tool.

Design for Manufacturability

Anyone can download an open-source 3D renderer such as POV-Ray or Blender, and quickly learn to draw a sphere or a cube in 3D. But after the initial excitement wears off, we have to face the blues: most of us don’t have the skill or perseverance to make the next Avatar any time soon.

The same holds true for industrial design – for a couple of reasons:

  • CAD is genuinely difficult. Gaining proficiency in a CAD application is even harder than mastering a general-purpose 3D tool. It takes hundreds of hours of practice to simply get to the point where you can use a two-dimensional input device (and an equally two-dimensional screen) to accurately sketch any complex organic shapes or intricate mechanical assemblies.
  • There is a lot more to industrial design than meets the eye. Most of us, even if given a hypothetical 3D printer that makes flawless parts out of any metal of our choice, still wouldn’t be able to produce a working nail clipper or a soda can. Industrial designers spend years studying the design, potential uses, and practical trade-offs of anything from spur gears to hundreds of various types of linkages, hinges, joints, or cams. Heck, there are at least four sophisticated design decisions that went into making the lid for a box of Tic Tacs.
  • Mechanical engineering is a real science. Plastics and metals are fairly imperfect and finicky materials; they are not easy to turn into parts that are durable, practical, and aesthetic at the same time. Flat sheets of these materials are almost always disappointingly wobbly and easy to bend. Even items as trivial as phone cases and Lego bricks make use of carefully placed ribs, gussets, and bosses to prevent the parts from deforming or falling apart. The basic engineering principles take time to master and properly apply in your work.
  • Manufacturing processes are not perfect – and won’t be any time soon. Part design is greatly complicated by the need to account for manufacturing tolerances, material shrinkage, minimum feature size, the need to support the part through the process, and so on. Very few advanced designs can be quickly sketched and broadly disseminated without paying attention to these factors, and tailoring them both for the general manufacturing method, and for the specific copy of the machine used to make the part.


A very thin but high-rigidity base platform used in the project pictured earlier on. Note the use of reinforcing ribs.

The high profile of 3D printing means that a vast majority of people who buy low-cost ABS extruders in the heat of the moment won’t be aware how difficult it is to progress from ideas to viable parts. That may hurt the community in the long haul.

Of course, universal availability of design skills is not strictly a necessity: it may be possible to settle for a model where the select few experts publish their designs for free, and millions of other users simply click “print”. But this brings us to another issue…

Toward engineering-grade parts


A complete chassis for Omnibot mkII, made out of a high-strength engineering plastic, silicone rubber, and an assortment of metal parts.

The existing hobbyist-friendly additive prototyping methods tend to produce parts from a very narrow choice of materials, all of which exhibit fairly poor mechanical characteristics; there are no signs that this will change in the coming years. With CNC mills, the situation is much better – but some of the essential materials remain difficult or expensive to process (for example, most rubbers don’t machine particularly well).

In popular view, 3D printers are a tool that will enable us to directly make almost anything; this way of thinking is exemplified by the commercial arms race to deliver FDM machines that print in color. But this pursuit may be misguided: as it is, both 3D printing and CNC machining tends to be more useful for producing tooling patterns – that is, shapes that serve as an input to another, more specialized manufacturing process carried later on.

In the industrial world, CNC-machined patterns are used for thermoforming, metal stamping, injection molding, and several varieties of casting. Not all of these can be safely and cheaply attempted at home – but some are surprisingly easy to work with. For example, resin casting combines ease of use with extreme fidelity and a broad range of properties attainable for the final parts. Without any sophisticated equipment, you can make squishy rubbers in any color you please – and five minutes later, switch to a composite material reinforced with carbon fibers or glass.


A relatively simple, single-part mold for resin casting.

Of course, these manufacturing workflows can be mastered by any determined hobbyist. Nevertheless, they add another level of complexity that may be unexpected and insurmountable to many; the obsession with direct manufacturability does very little to help.

Where do we really stand?

I am excited about 3D printing, but also uneasy with our way of thinking about the future of home manufacturing. For the driven hobbyists, the printer is just another tool that allows them to bring their designs to life. It shares many of its problems with the approaches that existed before – and adds its own serious challenges to the mix. Perhaps the best we can do is to learn from the manufacturing industry, rather than proclaiming its untimely death.

In fact, the preoccupation with reinventing the manufacturing process may be causing us to root for the wrong solution to begin with: the now-popular ABS extruders may not be able to achieve a reasonable precision and produce consistent and predictable results simply due to the limitations of the material: its extremely gooey consistency and an ill-defined melting point makes it difficult to control its deposition. The high temperature gradients created in the process don’t work in favor of the technology, too.

Consequently, we are probably not giving enough attention to some of the alternatives that seem to be able to deliver. For example, the (still insanely expensive) wax deposition printers from Solidscape achieve amazing levels of detail simply by working with a more suitable substance, and by combining additive and subtractive steps. But subtractive processes are not sexy, and the output from these printers is a fragile material that works only as a casting mold. With no popular appeal, the odds of the prices coming down are pretty slim.

Stereolithography is another interesting choice with promising results: the existing high-accuracy $30,000 printers seem to be going toe-to-toe with low-cost devices such as the upcoming Form 1. But the messy and wasteful operating principle limits their pop appeal, too.

One day, a silver bullet solution may materialize; if it does, it will be probably nothing like any of the existing technologies we are experimenting with. Until then, it pays to focus on the process, not on this week’s most-hyped tool.


An assortment of CNC-machined, resin-cast parts.

If you’re involved in a particularly revolutionary or awesome project and would like to write about it for 3D Thursday, or you have a related product that you’d like us to review or write about, please contact Eric Weinhoffer at Thanks for reading!

Michal Zalewski

Security engineer by day, a maker of tiny plastic gears by night.

blog comments powered by Disqus

Featured Products from the MakerShed