My experiences with 3D printers the last few months has validated to me how critical they can be to the design process. In fact I've realized I'm going to need more printers relatively quickly; I'm queueing up a number of projects. The current crop of printers in the pro-consumer range all appear to have fundamental limitations and are unreliable for various reasons (design, mechanics, software, environmental factors, etc.). I've decided it should be possible to build a printer that is fundamentally more reliable, consistent, and easier to use than what's currently available. Thus I've decided to design my own from scratch. I'm going to try to document my design decisions as I work through the whole design/build process.
A few basic requirements I've determined are critical:
- Large build envelope (e.g. 12" cube).
- Stiff, heated glass build surface
- Dual heads to print dissolvable support material
- Enclosed build envelope for temperature control
- Precise non-contact alignment of the z axis
- Extremely fast rapids (e.g. 300-400 ipm)
Since I am starting the design with no pre-conceived notions, the first step is to decide on the basic configuration of axis for the extruder head and build platform. This drives the basic design decisions for the overall printer. Here are a list of the potential configuration options as well as some comments on each:
|Extruder Head||Build Platform||Comments|
This configuration is uncommon due to the mechanical complexity required. However it has the strong advantage of keeping the soft plastic extremely stable. The Delta robot uses this configuration.
Decoupling one of the axis from the others simplifies the mechanism considerably. This configuration is almost as good as the previous one in that the material gets minimally jostled during the build, yet is considerably simpler to fabricate.
This approach again decouples the axis. This is probably the most common configuration for 3D printers. It is simple and stable, but introduces motion to the build platform that can damage some prints.
This is basically identical to the previous configuration just rotated 90 degrees.
This layout further simplifies/strengthens the print head at the cost of even more motion of the bed. Some current 3D printers use this layout.
Again this is the same as the previous layout just rotated 90 degrees.
This is similar to some small CNC mills (e.g. Tormach). It requires a work envelope at least twice as wide and deep as the build table. It works wells for mills, but is not very efficient or a 3D printer.
This is a classic high capacity mill configuration. It allows for an extremely stiff head, but it requires a work envelope considerably larger than the build table. This would be inefficient layout for a light, fast 3D printer.
The first and last configurations are the most complicated to implement and thus more costly and difficult to achieve for a small, very fast 3D printer. One exception to this is the Delta-style printer, but the limitations in mounting area for the print head and narrow/tall layout makes that a less than ideal solution.
On the other hand decoupling one axis from the other significantly simplifies the mechanical design. But the configurations that move the build platform in X or Y cause motion of the soft material during the build, thus involving the stability of the piece being fabricated. These designs tends towards a larger chance of motion-induced failures. Also since the build envelop will be enclosed, these configurations would require a larger outer box.
Having though through all of these options carefully, I've decided to start with the second configuration, i.e. moving the build table in Z and the head in X/Y. This is relatively easy to implement and fits well in a cube box. In fact the design I have in mind starts with an outer extruded aluminum cube stiffened with corner brackets or side panels and the linear rails mounted within this structure.
I am going to start working up a design approach to the rails and guides within the structure next.