Title :
Validation of the MEGA finite element code
Author :
Allen, N. ; Coles, P.C. ; Hill-Cottingham, R.J. ; Lai, H.C. ; Leonard, P.J. ; Rodger, D. ; Vong, P.K.
Author_Institution :
Univ. of Bath, UK
Abstract :
It is very important that software is validated and, ideally, benchmark experiments should be designed for this purpose. Finite element software developers tend to live in a hallucinogenic world of mathematics and computer graphics, so they instinctively try to avoid contact with the real world. However, for the purposes of validating software, the real world is indispensable. Ideally, every line of software should be tested, but it is probably more realistic to require that every module should be tested. Some benchmark experiments and comparisons which we have used in the past are described.
Keywords :
benchmark testing; computational electromagnetics; finite element analysis; program verification; benchmark experiments; finite element code validation; software validation;
Conference_Titel :
Validation of Computational Electromagnetics, IEE Seminar on
Print_ISBN :
0-86341-396-X
DOI :
10.1049/ic:20040115