3D printing

This week I am learning 3D printing, and again in order to efficiently use my time I decided to print something I can use for my thesis. Given that I am very new at this, I opted to print something simple but useful.

Here is a top view of the design ‘fish_chamber_top_view.png’ and the side view ‘fish_chamber_side_view.png’ . I designed this in solidworks and it is basically a fish chamber will allow the fish to be imaged on high-speed voltage imaging light sheet microscopy with light source coming from the side. The two sides will be taped with glass as well as the bottom squared hole. The bottom squared hole allow additional light source to image the fish from below. The chamber has two legs, whose dimensions are determined by the relative position to the laser source.

The entire chamber has a dimension of 57mm wide, 61.42mm long and 39.52mm tall, and because of its small dimensions, I was worried about whether normal thermoplastic 3d printer can give me accurate enough measurements. Because I had access to an AnyCubic Photon Mono M5s resin printer in my lab, I opted to use this to print my chamber.

The next challenge comes with the fact that the two legs of the fish chamber are floating in the air, causing the 3D printer to collapse due to lack of supports. In fact, no particular orientation has no floating pieces, so supports will be needed to print the chamber. On the advice of my lab mates, I opted to flip the chamber so it sits on the build plate like this and I added automatic support:

‘fish_chamber_front_view_support.png’ and ‘fish_chamber_back_view_support.png’

The print was estimated to take about 8hrs. I adjusted many print settings based on advice from labmates, and the print failed a few times, due to too small of a lifting speed intially. Notably, one of the times it only printed the support somehow.

‘3dprinting_failed.jpeg’

Finally, the printing worked.

3D scanning

For the 3D scanning assignment, I just followed the instruction manual in the Harvard REEF shop. It was quite straightfoward. I opted to handheld the scanner while the object is rotating on a ciruclar board so I can get information from all angles stably. Initially, I chose the marker tracking mode but that didn’t work super well. ‘3d_scanning_marked_tracking.jpeg’ .

Then once I went back to feature tracking, I tried to move the scanner to cover both the head and the body, and it came out quite well.

‘3dscanning_feature.jpeg’ .But it was noisy with many unspecific blobs and dots, so I did a combination of “one-click edit” and manual edit, the final results were clean.

‘3dscanning_feature_clean.jpeg’