This Site umd.edu

TREND 2017

TREND Home | 2016 | 2015 | 2014 | 2013 | 2012 | 2011 | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | Publications

 

August 11, 2017

TREND 2017 Projects

TREND2017 group
2017 participants. (l-r): Rebbeca Melkerson, Nathan Super, Julia Huddy (Transportation Electrification REU), Charlotte Slaughter, Joseph Betz, Hana Warner, Rebeckah Fussell, Emma Thackray, Skylar Eiswokitz, Dara Storer, Alex Wikner.
 

Winners of TREND Fair 2017

Best overall project: Skylar Eiskowitz (Dan Lathrop)
Best overall project runner-up: Joseph Betz (Brian Beaudoin)
Best presentation: Charlotte Slaughter and Dara Storer (Wolfgang Losert and Derek Richardson)
Best multimedia project: Rebeckah Fussell and Alex Wikner (Tom Antonsen, Michelle Girvan, and Ed Ott)

“Betz

X-Band Microwave Accelerator Cavity

(Presentation)

Joseph Betz, Widener University

(Mentors: Research Prof. Brian Beaudoin)
 
Inspired by research conducted at Stanford’s Linear Accelerator Center (SLAC), this research focused on the designing and manufacturing of an X-band (8 - 12 GHz) microwave accelerating cavity. Conventional RF accelerators are manufactured by molding sheet metal to the desired dimensions of the cavity and brazing the pieces together. Advances in manufacturing technology such as CNC mills, lathes, and 3D printers have allowed smaller and more precise cavities to be created. The accelerating cavity presented here, was designed using CAD software and optimized using finite element analysis (FEA) simulations. An aluminum test cavity was then manufactured for experimental testing. Various parameters were calculated using simulation software and compared to the experimental results. The completed accelerator would consist of many of these cavities arranged along a beam tube. Future work on this project will consist of creating a multi-cavity designs which include a series of waveguides and apertures to deliver power to each cavity.
 
Multimedia Project

mmpbetz


“Eiskowitz

First Observation of a Hall Effect in a Dusty Plasma: A Charged Granular Flow with Relevance to Planetary Rings

(Presentation)

Skylar Eiskowitz, Cooper Union

(Mentor: Prof. Dan Lathrop)
 
The particles in Saturn’s rings exhibit complex dynamical behavior as they orbit Saturn. They experience solar radiation pressure, electromagnetic forces, and granular collisions. To investigate the possibility of the Hall Effect in the dusty plasma that comprise Saturn’s rings, we have built an experiment that demonstrates the Hall Effect in granular matter. We focus on the Hall Effect because the rings’ grains become collisionally charged as they orbit transverse through Saturn’s dipolar magnetic field. The experimental setup includes a closed ring-like track where granular matter is forced to circulate driven by a continuous force of compressed air. The structure sits between two electromagnets so that a portion of the track experiences up to a 0.2 T uniform magnetic field. We vary the strength of the magnetic field and the speed of the particles. Here we report the voltage differences between two conducting plates on opposite sides of the track at various magnetic field strengths and various speeds of the particles. If Saturn’s rings do indeed experience the Hall Effect, the inside and outside of rings will have developed a charge separation that can lead to various phenomena including orbital effects influencing their orbits. Small rings may be led to solid body rotation based on the resulting electric fields.
 
Multimedia Project

 “mmpeiskowitz


“FusselandWikner

Good Prediction with a Bad Model: A Machine Learning Approach to Forecasting Chaos

(Presentation)

Rebeckah Fussell, Haverford College

Alex Wikner, Rice University

(Mentors: Profs. Tom Antonsen, Michelle Girvan, and Ed Ott)
 
The prediction of chaotic dynamical systems poses an important problem because of the ubiquity of these systems and a difficult problem due to the sensitivity of these systems to small deviations in the system's state. Where knowledge of the underlying mechanisms exists, the typical approach to prediction is to use this knowledge to build an approximate mathematical model. On the other hand, recent results in the neural-network-based machine learning technique known as “reservoir computing” have demonstrated successful model-free prediction of dynamical systems based solely on previous system measurements. We propose a hybrid technique that combines model-based prediction and reservoir computing for use on systems where the model is insufficiently accurate to perform good prediction. We test this hybrid prediction technique by applying it to a low-dimensional chaotic system and a high-dimensional spatiotemporal chaotic system. We find that when compared to model-only and reservoir-only approaches, our hybrid technique is able to accurately predict for a significantly longer period of time. In addition, our hybrid prediction technique is successful even when prediction via the component model and reservoir fail.
 
Multimedia Project


“Melkerson

Computational Studies of Particle Dynamics During Magnetic Reconnection

(Presentation)

Rebbeca Melkerson, Washington and Lee University

(Mentors: Prof. James Drake and Dr. Marc Swisdak)
 
During magnetic reconnection, magnetic field lines change topology and convert magnetic energy to kinetic and thermal particle energies. Reconnection occurs in many space and astrophysical plasmas, and the resultant charged particles have potentially hazardous consequences at Earth. We use particle-in-cell simulations on massively parallel supercomputers to learn how reconnection accelerates charged particles under different parameters. We first examine one recent claim of power-law (non-thermal) electron energy distribution production. While in-situ observations have shown that reconnection produces power-laws, simulations have had difficulty doing so. Next, we consider the effects of beta (the ratio of thermal to magnetic pressure) on the production of energetic electrons. Using a guiding-center model, we examine the principle energization terms and how they vary with the parameters. We find that the direct acceleration by electric fields is generally less significant than the process known as Fermi acceleration. 
 
Multimedia Project

mmpmelkerson

(Interactive diagram)


“SlaughterandStorer

Simulation and Observation of Periodic Forcing on a Granular System

(Presentation)

Charlotte Slaughter, Cornell University

Dara Storer, Brown University

(Mentors: Profs. Wolfgang Losert and Derek Richardson)
 
Properties of granular systems change when subject to various external conditions and forces. Many of these properties have been thoroughly observed and quantified in two-dimensional systems, but require further study in three-dimensional systems. With an index-matching technique, we are able to track the translational and rotational movements of particles within a granular system when they are subject to an oscillating compression wall. We simultaneously use pkdgrav, a soft-sphere simulation code, to model the experimental setup. These models allow us to easily change system parameters and thus study system properties that are impossible or impractical to observe experimentally. With a combination of experimental and simulated results, we hope to discover a threshold at which forcing on a granular system becomes irreversible, which has applications in many fields.
Multimedia Project

 “mmpslasto


“Stauffer

Design and Test of Auger Spectroscopy System for Photocathode Surface Studies

(Presentation)

Eileen Stauffer, University of Maryland

(Mentor: Dr. Eric Montgomery)
 
Photocathodes are devices specially designed to emit high-current, fast-response-time electron beams through the use of the photoelectric effect. This capability makes photocathodes ideal for photoinjector applications such as particle accelerators or free-electron lasers. The efficiency of electron emission in a photocathode is dependent on the work function of its surface, a factor which is in turn dependent on the chemical composition of the cathode surface. This chemical surface environment can be probed using Auger electron spectroscopy. We have designed a cathode holder and vacuum system compatible with the use of a compact Auger analyzer. Photocathode surface scans with this instrument will be presented. 
 
Multimedia Project

 “mmpstauffer


“Super

Avalanches and Entropy Generation in Sand Timers

(Presentation)

Nathan Super, College of William and Mary

(Mentors: Profs. Tom Murphy and Raj Roy)
 
We investigate the dynamics of granular flow and avalanching patterns in an hourglass. Previous experiments on similar systems have found power law statistics in the size and duration distributions of falling sand avalanches and have examined the power spectra of time-series data. We continue this line of inquiry and examine entropy generation in the system as a function of spatial resolution. We look to establish a quantifiable relationship between avalanche distributions, observation resolution and entropy generation. We send light through the sand timer and into a detector and observe sand flow via fluctuations in the transmitted light. The spatial resolution is regulated by employing optical apertures to constrain the illuminated area. Avalanche duration is calculated as the time from when the signal drops below a determined threshold to when it returns; size is computed as the integrated voltage over that interval. Simple histogramming is used to examine distributions for these variables. We use the Cohen-Procaccia algorithm to determine entropy generation rates from time series data. This quantity is the average information, in bits/sec, in the measured signal, and is a function of the binning resolution and time resolution of the algorithm.
 
Multimedia Project


Density Evolution in Dynamical Systems

Emma Thackray, Mount Holyoke College

(Mentor: Prof. Derek Paley)
 
Multimedia Project


“Warner

Design and Implementation of PID Controller for Interferometer Stabilization

(Presentation)

Hana Warner, College of William and Mary

(Mentor: Prof. Mohammad Hafezi)
 
Interferometric measurements are highly sensitive to environmental fluctuations. This study explores the implementation of a Proportional-Integral-Derivative (PID) control loop to stabilize a Michaelson interferometer used for single photon interference. Error signals are generated by photodiode measurements and processed in LabVIEW using a C++ DLL. Output signals are then sent to a mirror mounted on a piezoelectric controller to compensate for changes in arm length. This stabilizes the interferometer’s power output. The linear controller was adapted to the nonlinear position/power system by introducing directionality to error and phase measurements. The control loop succeeded in correcting for a long-time drift in our laser and stabilized interferometer power within 5% of set point relative to the laser’s power range.
 
Multimedia Project

mmpwarner

(Interactive diagram)


“KnollmanandSitaram

Detecting Torsional Modes in Tapered Optical Nanofibers, Felix Knollmann and Meghna Sitaram - JQI

(Presentation)

Felix Knollmann, Williams College - Participation through JQI

Meghna Sitaram, University of Maryland - Participation through JQI

(Mentor: Prof. Paul Lett)
 
This summer we explored optomechanical effects—interactions between light and the mechanical modes of macroscopic objects—in tapered optical nanofibers and their potential use for generating quantum light. Tapered optical nanofibers are created from standard optical fibers through controlled heating and stretching. A large part of our work involved reconfiguring the fiber pulling apparatus and pulling new nanofibers. Beyond that, we studied the optomechanical interactions of the torsional (twisting) modes with light transmitted through the fiber using heterodyne detection. This technique mixes a high-powered, frequency shifted local oscillator with a smaller signal. The frequency shift appears on the spectrum analyzer as a beat note and has side bands of ~200 kHz if the light has coupled with the torsional modes of the fiber. The frequency splitting of the torsional modes is reminiscent of the atomic levels used in the process of four-wave mixing, which is known to generate squeezed light. Squeezed light allows for more precise measurements than coherent light because of its modified noise properties. The long-term goal of this experiment is to use the torsional modes of the nanofiber to generate squeezed light of virtually arbitrary wavelength.