Glow Discharge Characterization with Optical Emission Spectroscopy

REU Student: Sophie Anderson, Austin College, TX

Faculty Mentor: Dr. Christopher Limbach

Glow discharge plasma can be generated when a low pressure gas gains sufficient energy to ionize and emit visible light. In the National Aerothermochemistry Laboratory glow discharge plasma will be used in the wind tunnels in order to trip the flow, or generate turbulence, over a wedge model and thus it is important that we know certain characteristics about the plasma being generated. Using optical emission spectroscopy characteristics such as temperature and density can be extracted from determining the electron populations in certain energy levels. During this research we created a permanent setup for our gas discharge tube in the lab including the setup of a spectrometer and camera to take images of emission lines. We were able to obtain a spectrum of our gas discharge tube and identify features typical of O+ and N2 B-A transitions.


Particle Motion for Hall Thruster Kinetic Simulation

REU Student: Michael Fennema, Texas A&M University, College Station, TX

Faculty Mentor: Dr. Kentaro Hara

This project seeks to model particle motion within a Hall thruster chamber. Particularly this model focuses on electron behavior as a response to field data from a SPT-100 thruster, which serves as a basis for a future kinetic simulation of a Hall thruster. Code developed for this simulation models events including electron-neutral collision, excitation, ionization, wall collision, and anode/cathode interaction. The environment designed in this project governs test particles through short, computer-light simulations which show consistent alignment with real phenomenon, as well as with other literature on the subject. The methods used in this project are demonstrated to be accurate and functional as a base for a full intensive kinetic simulation to fully map thruster properties. 


Design and Development of an Experimental Setup for a Novel-Capable Flapping Wing

REU Student: Jakqueline Granillo, Smith College, Northampton, MA

Faculty Mentor: Dr. Moble Benedict

Grad Student Mentor: David Coleman

The current work focuses on establishing a fundamental understanding of the flapping of a rigid wing. This paper talks about the design and development of an experimental setup which measures the inertial and aerodynamic forces of a rigid flapping wing while in pitching and flapping motion. The experiment was designed for a submerged wing in water at a Reynolds number of approximately 50,000, which is a rough estimate of the Reynolds number at which the Hover-Capable Flapping Wing MAV operates. A miniature six-component force load cell was utilized at the root of the wing. The setup was designed with the intention of mimicking the same conditions as the Hover-Capable Flapping Wing MAV. Kinematic viscosity of water is ten times larger than air, therefore, the flapping wing frequency was decreased to 1 Hz and the wing size doubled the MAV’s wing size. Through experimenting, it was discovered that the pitching servo motion depends on the position of the flapping servo. In order for pitching motion to be mimicked, the position of flap motion at real time must be known.


The Design and Testing of a Novel Spectrometer for Plasma Diagnostics

REU Student: Richard Lee Hollenbach III, University of Pittsburgh, PA

Faculty Mentor: Dr. Christopher Limbach

Continuous and pulsed plasma discharges are becoming widely used in the fields of aerodynamics, combustion, and chemistry. Thus, a spectrometer with a high resolution and a wide bandwidth is designed and built to study and characterize the flow of plasma. A model was first programmed using known equations to predict the results of the spectrometer and to optimize parameters. Then the system was constructed upon an optical table using the various parameters calculated from the model. Once built, the spectrometer was tested using a monochromatic laser and the results were compared to the model. The effects of varying two specific parameters were studied and will be useful information for calibrating the spectrometer when more complex light sources are introduced as well as plasma. The effect of a diffraction grating which will prevent cross axis interference was also studied.


Pushing the Limits of Wall Thickness in CNFs

REU Student: Isabel Kalnin, University of Portland, OR

Faculty Mentor: Dr. Mohammed Naraghi

In recent years, there has been a plateau in progress increasing the strength of carbon fibers (CFs) and carbon nanofibers (CNFs). The reason for this is a weakness in the fibers that is an intrinsic part of the creation process. As carbonization is a diffusion driven process, carbonization leads to a distinct skin-core structure that when cooled, creates microcracking, weakening the fibers. While this effect can be somewhat mitigated by decreasing the radius of CFs, another method of decreasing defects is to electrospin a coaxial fiber with a removable core, producing a hollow fiber with a very thin homogenous skin. In this work, we varied solution viscosity, controlled by weight percent composition of solutions, along with the flow rates of the two solutions relative to one another, in order to determine the parameters that best minimize the wall thickness of carbon nanofibers as a means to maximize strength. The purpose of this study is to determine the limits of wall thickness in hollow CNFs.

In this study, we used coaxial electrospinning to create a hollow fiber. The core of the fiber is composed of polymethyl methacrylate (PMMA), while the skin of the fiber is composed of polyacrylonitrile (PAN). Solutions of PAN and PMMA, both in dimethylformamide (DMF), were pumped through a coaxial needle on to a rotary target disk with an applied voltage of 12 kV. Solutions of PAN were kept constant, at 10 wt % throughout the study, while two concentrations of PMMA were used, 10% and 16%. Solutions were pumped at ratios of 0.75:1 PAN: PMMA and 1.5:1. Electrospun fibers were then hot drawn to a ratio of 2, stabilized in an oxidizing environment at 275 degrees C, and carbonized in an inert environment at 1200 degrees C. Following carbonization, the diameter and wall thickness of fibers were then measured using a SEM. According to theoretical calculations, the thinnest fiber fabricated in this study should have an area ratio of 2.13 PMMA: PAN, meaning that the majority of the cross-section of a fiber should become hollow. This was calculated for the 10% PAN : 16% PMMA set of samples. From preliminary images, fibers spun from 10% PAN : 10% PMMA solutions were not closed, and showed much smaller fiber diameters, on the range of 200 – 300 nanometers, whereas fibers spun from 10% : 16% solutions were closed, and showed diameters on the range of 500 – 600 nanometers.


Developing Carbon Nanofibers on a Spool via Near-Field Electrospinning (NFES)

REU Student: Christopher Kuo-LeBlanc, University of Massachusetts at Amherst, MA

Faculty Mentor: Dr. Mohammed Naraghi

Grad Student Mentor: Jizhe Cai

Carbon fibers (CFs) have held the position for the automotive and aerospace industries’ best for lightweight structural materials. However, in the past two decades CFs have reached a plateau in strength. Recent improvements have downsized CF technology to structures known as carbon nanofibers (CNFs) to increase strengths by over 100%. Manufacturing CNFs entails processing polyacrylonitrile (PAN) nanofibers into CNFs. Electrospinning is the most popular method of producing these precursor CNFs. This process, however, is batch and provides a distribution of PAN nanofiber diameters, and thus CNF properties, that is too large for commercial use. Our primary objective is to develop a method of continuous CNF precursor production with a narrower distribution of diameters via near- field electrospinning (NFES). Future developments could yield continuous development and processing of CNF precursors analogous to industrial CF manufacturing, which spin CF precursors onto a spool while simultaneously pulling the fiber to later processing spools. We will show in this study that NFES is a viable method of CNF precursor development with diameters comparable to traditional electrospinning, but with a consistent nanofiber diameter. 


Linearized Model of UAV Flight in Extreme Temperature Environments

REU Student: Samuel Lemma, Bristol Community College

Faculty Mentor: Dr. James G. Boyd

Carbon fibers (CFs) have held the position for the automotive and aerospace industries’ best for lightweight structural materials. However, in the past two decades CFs have reached a plateau in strength. Recent improvements have downsized CF technology to structures known as carbon nanofibers (CNFs) to increase strengths by over 100%. Manufacturing CNFs entails processing polyacrylonitrile (PAN) nanofibers into CNFs. Electrospinning is the most popular method of producing these precursor CNFs. This process, however, is batch and provides a distribution of PAN nanofiber diameters, and thus CNF properties, that is too large for commercial use. Our primary objective is to develop a method of continuous CNF precursor production with a narrower distribution of diameters via near- field electrospinning (NFES). Future developments could yield continuous development and processing of CNF precursors analogous to industrial CF manufacturing, which spin CF precursors onto a spool while simultaneously pulling the fiber to later processing spools. We will show in this study that NFES is a viable method of CNF precursor development with diameters comparable to traditional electrospinning, but with a consistent nanofiber diameter. 


Ion Acoustic Instabilities in Collisionless Plume Plasma

REU Student: Brian Puckett, Hastings College

Faculty Mentor: Dr. Kentaro Hara

This work presents grid-based simulations of collisionless plasma in the plume region of a hollow cathode. When an electron current travels through current carrying plasma, particle trapping as well as nonlinear instabilities and oscillations, or “ion acoustic turbulence”, occurs. This disturbance is suggested to be the source of observed anomalous electron transport, which can affect performance and lifetime of hollow cathodes. Previous research suggests that instabilities do not occur when electron Mach number is below unity. The effect of electron Mach number (ratio of bulk velocity to thermal velocity) on the ion acoustic turbulence is investigated in this work. Resulting ion velocity distribution functions from these simulations show a direct relationship between electron Mach number and phase velocity of present ion acoustic waves.


1D Hybrid Simulation of Hall Thruster Plasma with Oscillating Discharge Voltage

REU Student: Kareem Ramadan, Ohio State University

Faculty Mentor: Dr. Kentaro Hara

A Hall thruster is an electric propulsion device that uses magnetic and electric fields to generate plasma and produce thrust. They are used globally and can be found on deep-space satellites due to their relatively high thrust and simple construction. In order to further develop the thruster, the effect that discharge oscillations have on performance should be studied. To this end, AC modulations are added to the otherwise DC discharge voltage to enhance the oscillations. A hybrid simulation which employs direct kinetic modeling for ions and neutral atoms and continuum modeling for electrons is used to obtain the discharge plasma properties. Current experimental results from the Princeton Plasma Physics Laboratory (PPPL) indicate that the mean thrust and fluctuation increase with stronger modulations. However, the simulation results portray a static mean thrust with increasing fluctuations. The discrepancy could be attributed to the difference in thruster geometry, inaccurate anomalous electron modeling (plasma turbulence) or the exclusion of cathode effects. Future work will involve adapting PPPL’s hall thruster geometry and the different plasma physics that accompany it into the simulation.


Hypersonic Laminar-Turbulent Transition Prediction using Linear Stability Theory

REU Student: Andrew K. Riha, Texas A&M University, College Station, TX

Faculty Mentor: Dr. Helen Reed

Grad Student Mentors: Alexander J. Moyes, Travis S. Kocian

Linear Stability Theory (LST) is a technique for analyzing the stability of different mechanisms present in a given flow field. An LST model is derived and coded in Fortran for use on supersonic and hypersonic vehicles with simplistic yet representative geometries. Results from this code are verified against the stability code EPIC using a flat plate at Mach 2 and the Purdue Compression Cone (PCC) at Mach 6. The PCC is further analyzed to find that 282 kHz is the most amplified disturbance, reaching N factors of over 13 at the rear of the vehicle. These results compare well against previous experiments on this cone by Wheaton et al (2009). 


Asynchronous Computation for Multi-Dimensional Problems

REU Student: Gabriella Sallai, Franklin and Marshall College, Lancaster, PA

Faculty Mentor: Dr. Sharath S. Girimaji

Grad Student Mentors: Ankita Mittal

Massively parallel simulations currently use synchronization between processing elements (PEs) in perform their tasks. This method reduces computation time, but fails to reduce communication time between PEs, actually increasing communication time. Relaxing synchronization between PEs (implementing asynchronization) effectively lowers both computation and communication time. However, it also affects the accuracy of the simulation by reducing the average error to first-order. Currently, a variety of approaches are being developed to improve the order of accuracy in asynchronous computation. The approach I work with modifies the original governing equation to develop a proxy-equation. When solved asynchronously, this proxy-equation recovers the order of accuracy, previously missing in the average error, of the original numerical scheme. This equation has previously been implemented on one-dimensional simulations. I propose that, for 1D simulations using the proxy-equation method, there is a critical wavenumber at which asynchronous error becomes dominant over synchronous error and the method is no longer successful in effectively in correcting for asynchronous error. I also propose the proxy-equation method recovers order of accuracy for higher dimension (2D and 3D) simulations and lowers the average error displayed in asynchronous computations in these dimensions.


Micromechanical Modeling of High Temperature Shape Memory Alloys

REU Student: Aliyah Smith, University of Maryland, Baltimore County (UMBC)

Faculty Mentor: Dr. Dimitris Lagoudas

Post-doctoral Mentor: Alexandros Solomou

Shape Memory Alloys (SMAs) are active materials with unique thermomechanical properties that can transform a non-mechanical stimulus into a mechanical response. The material properties of SMAs can be manipulated through a metalworking process known as heat-treatment, which involves the heating of alloys under specific temperatures and lengths of time with an aim to alter their properties due to changes that induce in their microstructural level. However, a purely experimental, intuition-based approach is not sufficient due to the high cost and preparation time required to synthesize and characterize the material. Thus, a micromechanical model is needed to predict the material properties of heat treated SMAs. Based on an existing code suitable for NiTi SMAs, an updated code has been developed that uses the geometry of ellipsoids to fabricate the particles by calculating their dimensions, checking for any intersections, and rotating them within the microstructure. This algorithm will be integrated with other codes to a produce a model that will be used to accelerate the characterization of heat treated SMAs to be implemented in various aerospace applications.