Fast Radio Bursts are millisecond-long radio flashes of unknown physical origin. One such, FRB121102 (named based on the day it was first detected), has been observed to repeat: from time to time there will be an hour or two in which many bursts are detected. These bursts are not simple flashes: they appear to consist of sub-bursts that drift downwards in frequency. This thing is a collection of 3D prints of the burst structures, visualizing this structure by printing it out in plastic.
Each of these models is to the same physical scale: the height of each point is proportional to the flux density, and the time axis is marked in milliseconds, while the frequency axis is marked in hundreds of megahertz. The time of the first tick is marked on each object, as is the frequency range, the date, and the instrument used: PUPPI (Puerto-rican Ultimate Pulsar Processing Instrument), at the 300-m dish at the Arecibo Observatory.
This thing includes models for a number of individual bursts. They are named according to the date and time of their detection, but I have also added a word to the filename to briefly describe what the burst looks like.
For more information on this object and this data, see the note at ASTRON's web page or the paper Michili et al. 2018 and Hessels et al. 2018.
The only overhangs are the slightly indented text; this should pose no difficulty to the printer. The STL files are huge, and the gcode files are also large, but the shapes are not very difficult to print. The main challenge I had was getting them off the print bed when they were done.
The tallest of these is actually just barely short enough to be printed on a Prusa i3 Mk2s - in fact PrusaControl will refuse to slice it, and though slic3r will slice it fine, there is a risk the print head will collide with the spool holder.
Note that the "huge" and the "long" FRBs do not appear in Thingiview; they're in binary STL format, and while they print fine, Thingiview can't cope. The "huge" one is pictured against the Whirlpool Galaxy in one of the photos.
We have raw data files containing burst flux as a function of time and frequency. I load these into an ipython notebook and use a Savitzky-Golay filter to smooth them. I then use SolidPython to add annotations and create OpenSCAD files. I render these files to STL on a compute server as OpenSCAD requires quite a lot of RAM for this computation.