Build a Supercomputer
Students will build a supercomputer! Well, almost. Supercomputers typically use thousands of processors running in parallel to solve problems in science, finance, and other areas. They will build a smaller supercomputer to gain insight and understanding in how supercomputers are organized and then how to program them. Students will build a Beowulf cluster using ordinary computers. Students will then write a parallel program, compile the program, and execute that program on the cluster. Areas that will be covered during this project are:
- Computing basics
- Computer networking
- Linux operating system
- Computer programming
Project review and summary
Students will be required to answer the research question: "In what year would the supercomputer we build be considered the world's fastest supercomputer?" Students will be given classroom-style lectures in addition to hands-on assignment to enforce topics discussed.
Joint Institute for Computational Sciences
Mentor: Robert Whitten
Facilitator: Jerry Sherrod
Assistants: Benjamin Taylor, Christian Klenk, Nick Csercsevits, Charlene Watson
Students: Brianna Cummings, Mikenzi Glover, Matthew Herald, Nathan Smith, Elizabeth Sprinkle, Mason VanDyke, Shyderica Young
Calibrating optical diagnostics for use on the prototype Material-Plasma Exposure eXperiment (proto-MPEX)
The prototype Material Plasma Exposure eXperiment (Proto-MPEX) will begin operating at ORNL in August 2014. Proto-MPEX uses magnetic fields in a “linear configuration” to confine a low-temperature, high-density plasma, generated by radio-frequency power. This “beam” of plasma is directed onto a material target to study the material degradation effects that are expected to occur in fusion reactors. This plasma emits visible light as it is generated.
The visible light is collected by diagnostic systems (“Filterscopes” and spectrometers) for analysis. In order to understand the measurements, the diagnostic systems must be calibrated. An important part of this calibration is measuring the transmission of the individual elements (filters, fiber optics, couplers, lenses, windows, etc.), so that the diagnostics can be reconfigured to meet varying experimental conditions. A Labsphere “white” light source will be used as a proxy for the light that will be emitted from Proto-MPEX during normal operation. A detector will be used to make measurements and calculate the transmission properties of the assembled diagnostic systems.
ORNL Division: Fusion & Materials for Nuclear Systems (FMNS)
Mentor: Theodore Biewer, Stephanie Diem, Guinevere Shaw, Holly Ray
Group A (Ocean Optics)
Group B (Filterscopes)
Design of Novel Polymeric Materials Using Computer Simulation
For polymers, interactions between molecules are so complex that one cannot possibly understand the fundamental physics of their motions and stability of the materials they are made of. Macromolecules often self-assemble to form natural materials like protein, DNA and also commercial materials like rubber. We can get knowledge by mimicking nature and produce materials artificially for applications in our daily life. For example, tires, plastics, and medicines are artificially made materials. To design these, expensive experimental techniques are used. These experiments, however, may not produce results and lack the basic understanding of physics. Therefore, it is better to study these material properties on computer first before doing expensive experiments. Our goal is to introduce some of the computer experiments of macromolecules that in turn will help design better materials for future applications.
It is well known that the particles follow Newtonian mechanics at a classical level, i.e., they follow F=ma. Therefore, if we know the force acting on a particle at a given time, we can predict, by using Newtonʼs law, what is going to happen in a future time. For this, we feed the computer with the ʽvirtualʼ macromolecules and instruct the computer to find out the final product following Newtonian mechanics. Hence, the designing of novel polymeric materials on a computer can be achieved. But these large scale calculations require supercomputers.
In this project we will try to understand how the molecules interact using computer simulation. What are the forces that bind them together? Why do they self- assemble in a particular form? What is the temperature and density range that should be used to achieve the best material design?
ORNL Division: Computer Science and Mathematics & Center for Nanophase Material Sciences
Mentors: Monojoy Goswami and Bobby Sumpter
Students: Rachel Brooks, Sadie Harris, Adam Holloway, Amber Hooks, Lucas VanDee
Making a Rube Goldberg Nanomachine
Reuben (Rube) Goldberg was an American cartoonist and inventor. He is best known for a series of cartoons that he made showing very complicated contraptions performing very simple tasks . In this project, we will aim at making such a complicated contraption, but at the nanoscale. The building blocks of this contraption will be: melted nanodroplets, patterned substrates, and computer simulations. The droplets can be made to jump and to move in a particular direction, depending on the substrates upon which they rest. The students will devise a Rube Goldberg Nanomachine using these components, and a video of the result will be made. The overall objective of this work is to show that controlling movement at the nanoscale is not an easy task.
ORNL Division: Center for Nanophase Materials Sciences
Mentors: Miguel Fuentes-Cabrera
Students: Ethan Greist, Kassie Hollabaugh, Zoia Jordan, Kevin Reynolds
Robotic Systems and Engineering Development
Robots are used in the industry to protect humans from hazardous environments or when the work involves highly repetitive and precision tasks. The objectives of this project are to (1) expose students to robotic projects underway at ORNL and (2) provide hands-on experience in designing, constructing and programming a small robot. The students will work in three groups on similar problems at the Remote Systems Group of ORNL's Fusion and Materials for Nuclear Systems Division. The focus of this project is to develop the mechanical and programming skills that are needed to design, build and operate a robot. The student will build a robot that can navigate an obstacle course using various sensors (light, ultrasonic and/or touch). The students will learn which sensors are best suited for which purposes and what logic is appropriate for controlling the robot's trajectory. Students will be using the Lynxmotion Tri-Track Robot and AL5A Robotic Arm for building and testing.
ORNL Division: Fusion and Materials for Nuclear Systems
Mentors: Venugopal Varma, Adam Aaron and Adam Carroll
Facilitators: Carl Mallette and Ken Swayne
Students: Jayda Bettum,
Closed Loop Performance Characterization Of ITER Related Vacuum Pumps, Compressors, And Fast Valves
As part of the ITER project whose goal is to demonstrate power production in a fusion reaction, the management of vacuum and gas flows plays an important role. In order to make sure that the performance of each component can meet the design requirements, baseline and performance testing is required. U.S. ITER and ORNL have been working to build capability to confirm the performance of the Mikuni piston pump, Fluitron compressor, and gas valves in representative conditions.
This project will involve the loop testing of these components to determine the best method for characterization based current standards as well as exploring the sensitivity of the methods to changes in performance parameters with respect to orifice diameter, system pressures, and gas volumes. In addition, diagnostics will be surveyed and analyzed to understand the impact of these on the characterization this equipment.
ORNL Division: Fuel Cycle and Isotope
Mentors: Robert Duckworth
Students: Autumn Griego,
Lithium Titanium Oxide – Synthesis, Characterization and Uses
This research project will be conducted in ORNL’s Chemical Sciences Division (CSD), Materials Chemistry Group and is designed to allow participants to better understand processes required to conduct a research project on materials used to produce batteries. The teachers will experience the multifaceted levels of conducting research. They will be given a research assignment and work with a research scientist to understand the required background, processes, and safety procedures. Along with learning to apply many scientific concepts to a real-world problem, they will learn laboratory skills which will enrich and enhance their teaching when they return to their classrooms. In addition, during the two-week program, the teachers will meet other researchers within the Group and Laboratory community and learn about other ORNL projects.
ORNL Division: Chemical Sciences
Mentor: M. Parans Paranthaman
Facilitator: James R. Davis
Teachers: Lisa Burns,
Physics at SNS
Summer Institute Teacher participants will learn the role of the neutrons in evolution of physics and what contribution the neutron will have in our expanding understanding of physics.
Teachers will learn the impressive history of the neutrons in physics, how we are using neutrons today to learn more about materials, fundamental physics, evolution of universe, and why we need SNS a 1.4 billion dollar Spallation Neutron Source (SNS) to study all this.
They will obtain some hands-on experience and go through phases of solving a scientific problem.
ORNL Division: Physics
Mentor: Seppo Penttila
Teachers: Jonathan Bemid,
Computational Modeling of Protein and Nucleic Acid Three Dimensional
The three dimensional structure of proteins and nucleic acids is of
considerable importance in the determination of the biological activity of these macromolecules. The lock and key hypothesis is one prominent example. Computational methods can assist in the determination of these structures especially when experimental data is not yet available. Experimental methods include x-ray and neutron crystallography, NMR and Mass spectroscopy. Algorithms exist to predict 3D structures based on known sequence data. Our group will examine protein and DNA structures using hand built molecular models and then compare this to the computer generated molecular structures available in the protein and nucleic acid data banks. These large databases have grown to the extent that predictive power has been enhanced.
ORNL Division: : Biosciences
Mentors: Ed Uberbacher
Facilitator: Brian Hingerty
Teachers: Carmelita Everett,
Understanding Energy Efficiency in Buildings Using Computer Simulation
This research project is designed to allow participants gain exposure to sustainable building design and an understanding of the impacts of weather, building design and materials, systems, and operation on building energy use. The teachers will research energy efficiency strategies in residential buildings located in different climates through computer modeling of building energy use. The teachers will be performing building energy simulations using NEAT (National Energy Audit Tool) and MulTEA (Multifamily Tool for Energy Audit) software, developed by ORNL, to evaluate energy savings in residential buildings from a wide range of energy efficient building technologies, such as thermal insulation, air sealing, energy-efficient windows, high-performance systems, and high-efficiency lighting and appliances.
ORNL Division: Energy and Transportation Sciences
Mentor: Mini Malhotra
Teachers: Mackenzie Langman,