Title :
Parallel computing for sheet metal strain analysis
Author :
Chan, Hau-L ; Spence, Allan D. ; Sklad, Matt P.
Author_Institution :
McMaster Univ., Hamilton, Ont., Canada
Abstract :
This paper describes an application of high performance computing for industrial sheet metal strain analysis. A grid pattern of small circles is first electrochemically etched on the surface of a sheet metal blank. The blank is then formed into the desired shape. After forming, a laser digitizer scans on the surface of the sheet metal part and collects (X, Y, Z) coordinate points with the laser beam reflected intensity data. The scan resolution is 10,000 points per square centimetre. Using parallel data processing, each deformed circle is extracted, grouped and fitted. The strains are computed by comparing the fitting result to the original grid element dimensions.
Keywords :
forming processes; grid computing; mechanical engineering computing; optical scanners; parallel processing; sheet metal processing; strain measurement; grid pattern; high performance computing; industrial sheet metal strain analysis; laser digitizer; parallel computing; parallel data processing; Capacitive sensors; Computer industry; Etching; High performance computing; Metals industry; Parallel processing; Performance analysis; Shape; Surface emitting lasers; Surface fitting;
Conference_Titel :
High Performance Computing Systems and Applications, 2005. HPCS 2005. 19th International Symposium on
Print_ISBN :
0-7695-2343-9
DOI :
10.1109/HPCS.2005.39
