The ability to 3D print functional plastic parts at home has been an incredible boon to inventors and tinkerers, but still fails to achieve the coveted direct design-to-print ease that has been promised for years. One reason for this failure is that most designs have features (namely overhangs and bridges) that are not ideal for layered FDM buildup. There are often ways to design out these features, or to design in support, but at some point these design changes will become contrived, and eventually will adversely affect the functionality of the part.

To help in certain cases, and generally to open up the design space to better and more-creative designs, Diabase has developed 4th-axis, or rotary printing, capabilities for the H-Series machines. That means you can now design and print on a cylindrical “bed” rather than your typical flat one. To go into why that might be useful, let’s look at one use case: the “boot” or “bellows” for a rotating joint (these boots are typically difficult to manufacture, expensive, and often require customization for specific applications).

Now try to imagine how you would print that part, from a flexible material, on a flat bed. While definitely possible, it would require either raising the angle between the ribs to some printable overhang (affecting part function), or using support material from both the bed surface and from the part surface. Again, both methods are possible, but neither is ideal. For this particular part, a better printing method is to build out radially from a cylindrical substrate.

Radial printing offers a couple of advantages here:

  1. Support material, if needed at all, is minimal and only from the bed
  2. All the stresses from material cooling are pulling inward, toward the same point.

This eliminates warping where we don’t want it, and makes for stronger parts in the direction we want them to be strong. (similarly, bed adhesion is not an issue, because the part is pulling itself toward the bed). This printing method easily produces a functional, failure-resistant part:

Looking forward, this build strategy has broad potential: you can print directly on any substrate that will fit through the rotary axis (such as a golf-club shaft, arrow, or ski pole); you can use two more axes (U and Y) to “iron” parts after they have printed by turning the part and keeping the nozzle perpendicular to it at all times (this would make all the print lines radial, rather than “islands” as they would be immediately after printing); or you could add the 5th axis and print in spherical coordinates rather than just cylindrical.  These more advanced techniques are currently software limited – no slicer can generate g-code for a 5-axis printer, but the mechanical capability is built into the H-Series. Here are a couple other examples of rotary prints:

Finally, software is an important topic here because it is a typical choke point in the toolchain. There are two “linking” software tools that we use to go in the CAD-SLICER-MACHINE gaps. The first takes the design and “unwraps” it. We then use a normal slicer (S3D) to generate g-code for the unwrapped part. Then we re-wrap it, and scale the travel distance in the Y (A) direction onto 360 degrees of rotation. This toolchain will be unified as time goes on, but it works well today, and anyone can try it out with an H-Series machine with rotary axis. Below you can see a comparison of the CAD model (left) and the “unwrapped” model (right).

Rotaryflat