Boris Stok
Univerza v Ljubljani - Fakulteta za strojnistvo
Laboratorij za numericno modeliranje in
simulacijo v mehaniki
1000 Ljubljana, Askerceva 6
P. Koc, FS Ljubljana
Danasnje stanje razvoja matematicnega in numericnega modeliranja
omogoca obravnavo vecine tehnicnih problemov na osnovi numericne
analize in racunalni{ke simulacije. Verodostojnost taksnih
simulacij je pogojena v veliki meri s stopnjo poznavanja snovnih
lastnosti, ki obravnavani problem karakterizirajo. Se posebej je
pomembna temperaturna odvisnost, pa naj gre za toplotne, reoloske
ali elektromagnetne lastnosti snovi. V prispevku obravnavamo
racunalnisko podprto identifikacijo temperaturne odvisnosti
snovnih lastnosti, zasnovani na metodologiji resevanja inverznih
problemov.
Kljucne besede: matematicno modeliranje, inverzni problem, snovna
identifikacija
The actual state of the art in mathematical and numerical
modelling has enabled most of technical problems to be treated by
means of numerical analysis and computer simulation. The
reliability of such simulations is affected considerably by the
degree to which the material properties characterizing the
considered problem are specified. In technical applications it is
the temperature dependence of thermal, rheologic and
electromagnetic material properties which is of particular
interest. The computer aided identification of the temperature
dependent material properties, which is based on the inverse
problem solution methodology, is presented in the paper.
Key words: mathematical modelling, inverse problem, material
characterization