Laser Propagation Modeling. Making Large Scale Ultrashort Pulse Laser Simulations Possible. Sponsor: Dr. William Roach Air Force Research Labs. Student: Mr. Noah Wolfe UNM/AFRL. Mentor: Dr. Mohammad Zunoubi State University of New York. Date: 11/30/2011. UNM Senior Design Project
Making Large Scale Ultrashort Pulse Laser Simulations Possible
Dr. William Roach
Air Force Research Labs
Mr. Noah Wolfe
Dr. Mohammad Zunoubi
State University of New York
UNM Senior Design Project
Fall Semester 2011
End of Semester Presentation
Challenges & Concerns
Senior Design Project Relevance
Ultra Short Pulse Laser Propagation
(But Animated )
Dr. Roach (Sponsor)
Laser Physics Support
Simulation and Modeling:
Develop unique 3D laser propagation/interaction code
Develop CUDA implementation of 3D code
Converted Propagation Code (December 23, 2011)
Previous Fortran subroutines have been converted and coupled into one object oriented C program effectively modeling all Non-Linear Schrodinger terms.
GPU Cluster Implementation (April 15, 2011)
Parallel version of the 3D code has been implemented with CUDA to allow for meaningful large scale testing.
Numerical Testing and Comparison (April 30, 2011)
Real experimental laser characteristics are input and simulated to compare with that of the experimental results.
GUI Development (May 7, 2011)
Output data is compiled and converted into a 3D animation that portrays the laser propagation and interaction with matter.
3D USPL Propagation Code
Need for reliable computational tools for the analysis of High-Power USPL propagation in atmosphere
GUI Program & GUI User Manual
Visual representation of the laser propagation in both space and time. Most likely implemented in Matlab.
Manual documents the process of creating a new animation from output data.
GPU Cluster Interface
Any and all software required to interconnect the multiple standalone computers into one functional cluster for high performance computations.
GPU Cluster User Manual
Complete write-up documenting the process of adding new nodes and running new codes.
Capable of computing over 1 Billion unknowns in Minutes
GPU compute cluster has been setup and confirmed operational.
Any GPU compatable program can be run to perform large-scale, computationally heavy simulations.
Finishing the Fortran to C conversion.
60% of the code converted/debugged
Also working on Matlab code to plot the output of the program.
Gaussian Pulse Dispersion
Collaborating with a team of engineers
development of effective project management, time management communication, and technical skills
Real world application of classroom material
application of the basic science, mathematics, engineering and design skills taught in earlier courses
3D propagation Code C programing ECE 131/231
GPU Cluster Computer Architecture ECE 337
Code Optimization Software Design ECE 330
Hands on experience of Engineering Process
Research, Development, Implementation, Testing, Maintenance/Improvement