Fatigue in cast metals: A generalized algorithm for multiaxial loading
Translated title
Udmattelse i støbemetal: En generaliseret algoritme for multiaksiel belastning
Authors
Pedersen, Erik Bouvin ; Nielsen, Morten Eggert
Term
4. term
Education
Publication year
2014
Submitted on
2014-08-04
Pages
110
Abstract
At forudsige, hvordan små revner vokser under gentagne belastninger, er vigtigt for at undgå svigt. Dette speciale præsenterer en algoritme til at simulere træthedsrevnevækst under fleraksial, variabel belastning (belastninger, der ændrer sig og virker i flere retninger). Tre udbredte revnevækstmodeller—Paris’ lov, Forman-ligningen og NASGRO—blev sammenlignet; for de undersøgte casestudier var Paris’ lov at foretrække. Den numeriske beregning af spændingsintensitetsfaktoren for mode I (KI) blev testet og valideret i 2D og 3D med en nøjagtighed inden for 2% af analytiske løsninger. I et plant casestudie blev Richards revnepropagationskriterium vist at være ækvivalent med kriteriet om maksimal tangentialspænding. Forudsigelser af revneretning i blandet belastning gav svingende resultater for mode I+II (åbning + glidning), men gode resultater for mode I+III (åbning + rivning). Et casestudie af revnevækst i mode I+III gav for konservative resultater. De observerede afvigelser i revneretning kan skyldes unøjagtige KII-værdier, og for konservative vækstrater var forventelige, fordi plastisk revnelukning ikke var inkluderet.
Predicting how small cracks grow under repeated loads is important to prevent failures. This thesis presents an algorithm to simulate fatigue crack growth under multiaxial, variable loading (loads that change and act in several directions). Three widely used crack growth models—Paris’ law, the Forman equation, and NASGRO—were compared; for the examined case studies, Paris’ law was preferred. The numerical computation of the mode I stress intensity factor (KI) was tested and validated in 2D and 3D, achieving accuracy within 2% of analytical solutions. In a planar case study, Richard’s crack propagation criterion was shown to be equivalent to the maximum tangential stress criterion. Mixed-mode crack direction predictions gave fluctuating results for mode I+II (opening + sliding) but good results for mode I+III (opening + tearing). A mode I+III crack growth case study yielded overly conservative results. The observed deviations in crack direction may be due to inaccurate KII values, and conservative growth rates were expected because plasticity-induced crack closure was not included.
[This abstract was generated with the help of AI]
Keywords
Documents