Applied Research Laboratories

In fused-deposition modelling 3D printing, melted plastic is extruded in lines and deposited successively into a desired shape. This progressive, layer-based process yields parts with nonuniform internal structure. This project studied the effects of print orientation and infill line pattern on acoustic transmission of polylactic acid (PLA). 200x200x3mm test plates were printed with 100% infill at various orientations relative to the print bed, changing the internal layer geometry by extension. The plate was then submerged in an underwater acoustic testing tank between two hydrophones. A chirp containing frequencies from 1Khz to 1MHz was produced behind the plate, and recorded as it passed through the plate and to the other side. Along with frequency being varied, the effect of incidence angle on transmission was also tested by moving the receiving hydrophone. Transmission spectra were produced for each plate, showing acoustic transmission per angle-frequency pair. Despite internal layer geometries, tested plates were shown to be generally isotropic. This result indicates that 3D printed parts for acoustic applications can be printed to optimize time and energy, as print orientations and settings at 100% infill have minimal acoustic effect.
Liberal Arts and Science Academy, Austin, Texas, Advanced Technology Laboratory

Quantum computing promises a future reimagined with its powerful potential speedups over its classical counterpart. However, modern quantum hardware is limited by the number of qubits and depth of the circuits. Striving to accelerate progress in this area and ultimately develop better quantum hardware, we seek to construct a tensor network simulator of random quantum circuits. Working with random quantum circuits allows for the development of a simulator that performs well in almost all use-cases. With this, we can better benchmark quantum hardware while pushing the limits of both classical and quantum computation. Converting a quantum circuit to a tensor network is trivial. We then simplify the resulting network using a variety of techniques including rank simplification, split decomposition, and column reduction. Next, we bi-partition the simplified network into the head and tail according to a cut size target. Fixing the qubits of the head to the zero computational state, and enumerating over all bit string combinations in the tail, we achieve a state-of-the-art contraction scheme that calculates 64 times as many bit strings in less than half the time it takes for the next best simulation. Future expansion of this project extends to further improving contraction path methods, developing better quantum computing benchmarks, and pushing forward quantum hardware.
Cedar Ridge High School, Round Rock, Texas, Signal and Information Sciences Laboratory

A digital zenith camera is an instrument capable of determining its location on Earth by observing the stars overhead, with impressive accuracy and precision. But they are large and heavy; a large SUV is typically required for transporting and deploying the equipment, while also relying on expensive optics. We explore the feasibility of using a smartphone, as a more portable and less expensive instrument, to produce a terrestrial position from the stars. A mobile app is developed to take pictures with configurable exposure duration and light sensitivity, and embeds the local time as well as tilts in the x-axis and y-axis, derived from accelerometer data, at the end of the exposure. Using the images captured with the app, additional software is run to extract and compare the star positions against asterisms formed from a star catalog to determine the celestial position of the image’s center. Subsequent use of the images’ timetags and tilt data helps to determine the user’s position on Earth from the celestial position of the stars in the smartphone’s images. An averaging of precision estimates in optical variation and accelerometer noise over 6 image collections of images taken at 19 second exposure time, 125 ISO, with a 79° field of view, yielded 0.0904° of overall system error, due to optical variations and accelerometer noise. Optical variations were the greatest contributor to overall system error. Additional work on minimizing optical variations should explore if coarser index file scales significantly change celestial position estimates, and explore how different lighting conditions affects solvability as well as celestial position estimates. Future work in addressing accelerometer noise should explore utilizing sensor fusion on modern smartphones, applying advanced noise filtering methods, or pursuing accelerometer specific low-pass filter optimization.
Dripping Springs High School, Dripping Springs, Texas, Space and Geophysics Laboratory