Gears, Rotation, and Internal Complexity
Following the goals stated in my AMPEL September update, ability to print models capable of motion is an important skill to wield in this journey through additive manufacturing. Some basics were shown by example in that article, but the last few weeks have been even more interesting as I delved into the mechanics and design. Updates on a couple of my platforms also presented new capabilities.
In this article I’ll discuss the implications which come with internal complexity of 3D Printing as well as showcase some of the models which prompted the discussion. To wrap it up, we’ll see a sneak peak of a project which I’ve discussed previously but am only now starting to really design for.
Internal Complexity
As I meet my goals of 3D printing smaller objects I am also meeting my goal of designing more complex objects. Compared to early models, which were little more than interesting geometric structures, current models are designed with these features in mind:
Print-in-place: this means that as soon as the print is finished and cooled it is ready for use. A “no assembly required” mentality to design means that we can save time in post-processing just by flexing mental capacity during design. Though many products throughout the 3D printing world require assembly due to size, material, or a number of other factors, practicing this design thinking for myself is an important capability to wield.
Motion: this is pretty self-explanatory. Since my goal is not to be printing simple models (a la Baby Yoda) but to build a design library of functional equipment, motion and connection to/utilization of power sources must also be considered in design. Thus far, rotational motion as a part of print-in-place models has proven difficult but the practice has yielded success.
As I incorporate the two features listed above, the internal structure increases in complexity. This has a positive correlation to the size of print file and time to print. A FDM printer (which I currently use) is essentially a piece of robotic equipment which you program to move in planes across 3 dimensions. Every time a new “corner” is added, necessitating the print head to change direction, the slicer creates an extra line of Gcode to direct the machinery’s motion. Stack enough of these and your print file becomes ginormous.
This isn’t to push people to shy away from designing complex models, but to paint a pictures of the trade off between print-in-place models and ones which require post-processing assembly. Let’s look at some examples
Implementation
In this section we’ll look at some recent projects. As director of the Additive Manufacturing Prototyping and Experimentation Laboratory, part of my job is making things which, honestly, I may not really have use for right now. Learning, practicing, manifesting, and documenting these experiments and prototypes remain directives for application of energy.
Solid Geared Bearing
See a demonstration of this project’s rotation in this video.
This project was undertaken with the intent of incorporating rotational motion with internally complex print-in-place design. This piece is a significant step in this regard as it Was the first success. As soon as the model was finished and cooled it could be pulled off the print bed and rotated. The model took almost 8 hours to print, which is surprising for a model of this size. This can be attributed to the complexity, as you can clearly see from number of directional variances in the final design.
As displayed in the video above, the rotation is quite effective. Conservation of momentum, not so much. In the design of future bearings I’ll try to design cavities in which metal pieces can be inserted. The current use of ABS material, no matter the density, is incapable of using rotation to conserve momentum at full potential. This was attempted in the next project.
Reuleaux Triangle Bearing
See a demonstration of the object’s rotation in this video.
This project failed in many ways but demonstrated success in others. First, the process of print-in-place models was successfully integrated into this. Like the Solid Geared Bearing, it was fully operational as soon as the print was finished and cooled. It also demonstrated success in unique gear and connectivity design. On close examination, we see that the outer ring of gears are actually triangles rather than perfect circles. Designing this system requires engineering diameters, depth of gear-set, and position in relation to other orbiting bodies.
As noted previously, you can suppose that this project also took an extensive amount of time to print.
The failure realized in this project was the attempt to insert extra metal components in post-processing in order to maximize energy conservation. I misjudged the size of bearing inserts. As we can see in the video above, the gear still works relatively well in transference of energy but is not effective at the conservation of momentum.
The Maze Box
To see a demonstration of this project, watch this video.
This project was really done on a whim but yielded some excellent results. My primary CAD software, Autodesk's Fusion 360, released an update which lets the user emboss a pattern on a cylinder. An SVG file of the maze was transcribed with this feature. The outer shell is smooth in the interior except for a single notch to "navigate" the maze.
A notable experiment on this project was a change in slicer settings. Normal protocol with a 0.4mm nozzle is to print each layer at a height of 0.2mm. For this project I tested printing at a layer height of 0.15mm to see if it would provide better fusion. This was successful but at a cost. With a decreased layer height, some partially fused material was pulled across open spaces or pushed off to the side of the model. (I also experienced this issue in printing the Hexagonal Fractal, but for other reasons which have since been troubleshooted). This didn’t ruin the model but just required that I scrape some extra filament off the sides.
This piece is a great Christmas present for if you want to annoy the bejesus out of a friend or family member. Put a $1 bill inside and watch them struggle figure out the maze for such a paltry reward.
Linkage System
See a demonstration of this prototype’s capabilities in this video.
This project’s prototype was inspired after watching a video about NASA developing “armor“ capable of being 3D printed. Of course, their development is in much finer detail and has been practically applied using both stereo-lithography and select laser sintering rather than fused deposition modeling as I have.
This linkage prototype worked splendidly. The next step in my experimentation with this design will be to run an application with tulle paper. This will necessitate a design which is thicker on the base. The GCode for printing should be modified to pause at a specified layer at which point I’ll lay down the perforated tulle paper. Once the tulle is stretched tightly and securely across the partially completed print, the print will resume and finish the design. The finished print will be a malleable and interconnected type of chain mail.
This project has the potential to run in many different directions. Incorporating a self-sealing latch, waterproofing, electrical components, rigidity under stress… these are all things which can take shape in a 3D printable format. Design thinking should be applied to enable a print-in-place model.
Conclusions
A reality of this time in technological history is that CAD software, CAM software, and additive manufacturing technology are evolving faster than I could hope to keep up. Building compoundable principles into my designs is more important than creating specific or really marketable pieces at this stage in the game. With this understanding, I’ve also come to the realization that I’ve nearly maxed out my capabilities with current hardware. Prototyping and Experimenting in a wider variety of materials and printing methods must be a priority as the library continues to grow.
