The Advanced Software Development and Commercialization Project

Technical Report · Sat Sep 01 00:00:00 EDT 1990

DOI:https://doi.org/10.2172/6357026· OSTI ID:6357026

Gallopoulos, E ^[1]; Canfield, T R; Minkoff, M; Mueller, C; Plaskacz, E; Weber, D P; Anderson, D M; Therios, I U ^[2]; Aslam, S; Bramley, R; Chen, H -C; Cybenko, G; Gallopoulos, E; Gao, H; Malony, A; Sameh, A ^[3]

---

1. ed.; Illinois Univ., Urbana, IL (USA). Center for Supercomputing Research and Development
2. Argonne National Lab., IL (USA)
3. Illinois Univ., Urbana, IL (USA). Center for Supercomputing Research

This is the first of a series of reports pertaining to progress in the Advanced Software Development and Commercialization Project, a joint collaborative effort between the Center for Supercomputing Research and Development of the University of Illinois and the Computing and Telecommunications Division of Argonne National Laboratory. The purpose of this work is to apply techniques of parallel computing that were pioneered by University of Illinois researchers to mature computational fluid dynamics (CFD) and structural dynamics (SD) computer codes developed at Argonne. The collaboration in this project will bring this unique combination of expertise to bear, for the first time, on industrially important problems. By so doing, it will expose the strengths and weaknesses of existing techniques for parallelizing programs and will identify those problems that need to be solved in order to enable wide spread production use of parallel computers. Secondly, the increased efficiency of the CFD and SD codes themselves will enable the simulation of larger, more accurate engineering models that involve fluid and structural dynamics. In order to realize the above two goals, we are considering two production codes that have been developed at ANL and are widely used by both industry and Universities. These are COMMIX and WHAMS-3D. The first is a computational fluid dynamics code that is used for both nuclear reactor design and safety and as a design tool for the casting industry. The second is a three-dimensional structural dynamics code used in nuclear reactor safety as well as crashworthiness studies. These codes are currently available for both sequential and vector computers only. Our main goal is to port and optimize these two codes on shared memory multiprocessors. In so doing, we shall establish a process that can be followed in optimizing other sequential or vector engineering codes for parallel processors.

• Research Organization: Argonne National Lab. (ANL), Argonne, IL (United States)
• Sponsoring Organization: DOE/ER; National Science Foundation (NSF)
• DOE Contract Number: W-31109-ENG-38
• OSTI ID: 6357026
• Report Number(s): ANL/TM-484; ON: DE91002956; CNN: 90-82144; CCR900000N
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE
42 ENGINEERING
75 CONDENSED MATTER PHYSICS
SUPERCONDUCTIVITY AND SUPERFLUIDITY
FLUID MECHANICS
C CODES
MECHANICAL STRUCTURES
W CODES
PARALLEL PROCESSING
COMPUTER ARCHITECTURE
ARRAY PROCESSORS
HEAT TRANSFER
HYDRAULICS
PERFORMANCE TESTING
PROGRESS REPORT
COMPUTER CODES
DOCUMENT TYPES
ENERGY TRANSFER
MECHANICS
PROGRAMMING
TESTING
990200* - Mathematics & Computers
420200 - Engineering- Facilities
Equipment
& Techniques
640410 - Fluid Physics- General Fluid Dynamics