Digitalizacija cikličnih sigma-epsilon krivulj

Benedičič, Matjaž (author), Nagode, Marko (mentor), Šeruga, Domen (comentor)

Z linearno interpolacijo cikličnih sigma-epsilon krivulj naredimo numerično napako. Preverili smo obstoječe načine zmanjševanja napake in ugotovili, da niso dovolj učinkoviti pri uporabi majhnega števila deformacijskih ravnin. Razvili smo tri metode, s katerimi smo to napako poizkusili zmanjšati. S programskih jezikom Python smo metode pretvorili v programe in jih preizkusili na realnem materialnem primeru. S pomočjo statistične analize smo ugotovili, da ena od razvitih metod ponuja izboljšavo linearne interpolacije in je smiselna za vgradnjo v algoritme preverjanja termo-mehanskega utrujanja.

• Language: Slovenian
• Keywords: termo-mehansko utrujanje, ciklične sigma-epsilon krivulje, linearna interpolacija, algoritmi, nelder-mead metoda, optimizacija
• Work type: Bachelor thesis/paper
• Organization: FS - Faculty of Mechanical Engineering
• Year: 2024
• Publication date in RUL: 14.08.2024

Secondary language

• Language: English
• Title: Digitalization of cyclic sigma-epsilon curves

Abstract

When we linearly interpolated the cyclic sigma-epsilon curves, we made a numerical error. After testing the existing means of reducing this error, we concluded that they are not effective enough when applied to only a few deformative plains. We’ve developed three methods which aim to address this problem. By using the Python Programming Language, we’ve transformed these methods into programmes and tested them out on a genuine material case. Through statistical analysis, we've determined that one of the developed methods offers a clear improvement in linear interpolation. It would be sensible to incorporate it into the algorithms that monitor thermo-mechanical fatigue.

• Keywords: thermo-mechanical fatigue, cyclic sigma-epsilon curves, linear interpolation, algorithms, nelder-mead method, optimization