Using 3D Printing to produce sub-millimeter sized structures

Hi, my name is Anant Gururaj and I’m going to be a senior at Northwestern. This summer I’ve been working in the Applied Materials Division with Kaizhong Gao.  I’m using 3D printing, specifically Finite Deposition Modelling, to develop a device intended to interact with particulate matter in liquid and gaseous streams. I’m not allowed to discuss specifics or show any prototypes I’ve developed, so I’ll be focusing on the two biggest challenges I’ve faced. These are minimizing printing time, especially for circular geometries, and achieving print thickness of less than 1mm.

I’m using Finite Deposition Modelling (FDM) to create our device. Other techniques are available but FDM is cheap and more widely used than others. Plastic (ABS) filament is fed to heated extruder where it melts. The extruder then deposits plastic onto the (heated) build plate layer-by layer to create a 3-D structure. Temperature of the extruder and plate, and print speed parameters can be controlled.

The design of the device calls for the construction of several rods, with each rod being designed to ensure the maximum surface area to volume ratio. I’m considering only rods with circular and square cross sections, and in either case, minimizing the diameter/side length of the circular/square rod results in maximizing the surface area to volume ratio.

Orientation of the object on the build plate made a massive difference to the print time (below). With circular rods, however, after the first layer is printed, subsequent layers will have overhanging regions that will drip onto the build plate, so orienting circular rods horizontally (which cuts down printing time significantly) on the build plate is not an option, and I only looked at square rods going forward.

 

In terms of minimizing print thickness, the minimum thickness achievable was 0.5mm, due to the printer’s nozzle width being 0.5mm. In fact, the software used by the 3D printer refused to process any structures that had dimensions less than 0.5mm. So I tried printing a rod with a cross section with a 0.6mm side length, and upon measuring it found that its dimensions were 0.75mm x 0.44mm. This deformation is likely due to the ABS ‘sinking’ under its own weight, as it is not completely solid when released from the extruder.

The 0.75mm x 0.44mm rod I printed was a singular rod, so the next logical step was investigating whether I could recreate those dimensions with more complex arrangements of rods, which would be needed in creating the device. Unfortunately, I can’t show any pictures of the prototype I made, but I found that when arranged in more complex configurations, the rods that were printed had dimensions closer to the intended 0.6mm x 0.6mm, but there was still some deformation.

Going forward there are other 3D printing techniques, such as Stereolithography, that could be used to produce rods with smaller dimensions much more precisely. These methods are, however, more expensive. Smaller nozzles could also be used, but these can also be expensive.