| Paper |
Title |
Page |
| MOPLT072 |
Effects of Positrons on Relativistic Solitons in Laser-Plasma Interactions
|
719 |
| |
- J.B. Kim, I.S. Ko
POSTECH, Pohang, Kyungbuk
- H. Suk
KERI, Changwon
|
|
| |
An extended 1D kinetic model of relativistic solitons by high power lasers in three species plasmas is suggested and it is applied to analysis on the effects of electron-positron pairs on the solitons. Stability condition of the solitons is derived. The range of parameters for the stable solitons are specified in the frequency-temperature plane. With the creation of electron-positron pairs, relativistic solitons appear stable in wider range of frequencies and temperatures. The regions are expanded toward higher values in overall ranges in the frequency-temperature plane. The stability conditions are affected by the density of positrons. The variation of shapes, peak E-field, and width of the solitons by varying the positron density are analyzed. We discuss the implications of the variation in the soliton on the ion accelerations by it.
|
|
| THPLT085 |
Reengineering and Refactoring Large-scale Scientific Programs with the Unified Process: A Case Study with OSIRIS PIC Program
|
2694 |
| |
- J.B. Kim, I.S. Ko
POSTECH, Pohang, Kyungbuk
- H. Suk
KERI, Changwon
|
|
| |
As science and engineering problems get more complex, programs which help modelling complicated problems larger and more sophisticated. This trend makes us recognize the importance of well-established engineering disciplines not only in designing large-scale scientific programs for special purposes in appropriate development time but also in importing the programs from other research group and refactor it for conveniences and more advanced applications. OSIRIS is a large-scale PIC code which was developed at UCLA for modelling of laser-plasma interactions. OSIRIS was reengineered and documented in UML by our group and ported to Linux cluster machine of 8 nodes. We report our current status of developing the extended version of OSIRIS, which was named as OSIRIS-X, and how a large-scale scientific programs can be enhanced efficiently with the Unified Process. Some guidelines in designing and refactoring large-scale scientific codes are presented and discussed. A common architecture model of numerically intensive programs for large-scale computing is suggested , and it is discussed how we can use it for rapid development and prototyping of scientific programs. We also discuss future challenges and prospects in OSIRIS-X development.
|
|