Iskanje po katalogu digitalne knjižnice

Opis: Artificial intelligence methods, especially artificial neural networks (ANNs), have increasingly been utilized for the mathematical description of physical phenomena in (metallic) material processing. Traditional methods often fall short in explaining the complex, real-world data observed in production. While ANN models, typically functioning as “black boxes”, improve production efficiency, a deeper understanding of the phenomena, akin to that provided by explicit mathematical formulas, could enhance this efficiency further. This article proposes a general framework that leverages ANNs (i.e., Conditional Average Estimator—CAE) to explain predicted results alongside their graphical presentation, marking a significant improvement over previous approaches and those relying on expert assessments. Unlike existing Explainable AI (XAI) methods, the proposed framework mimics the standard scientific methodology, utilizing minimal parameters for the mathematical representation of physical phenomena and their derivatives. Additionally, it analyzes the reliability and accuracy of the predictions using well-known statistical metrics, transitioning from deterministic to probabilistic descriptions for better handling of real-world phenomena. The proposed approach addresses both aleatory and epistemic uncertainties inherent in the data. The concept is demonstrated through the hot extrusion of aluminum alloy 6082, where CAE ANN models and predicts key parameters, and ChatGPT explains the results, enabling researchers and/or engineers to better understand the phenomena and outcomes obtained by ANNs.
Ključne besede: artificial neural networks, automatic explanation, hot extrusion, aluminum alloy, large language models, ChatGPT
Objavljeno v DKUM: 27.02.2025; Ogledov: 0; Prenosov: 3
Celotno besedilo (3,18 MB)
Gradivo ima več datotek! Več...

Opis: The basis of the master thesis is an in-depth and comprehensive analysis of the scientific literature on damage models of thermo-mechanical fatigue. The aim of the thesis is to investigate and determine the suitability of damage models for their application in numerical simulations of components subjected to thermo-mechanical loading with in-phase, out-of-phase or constant temperature cycles. The theoretical background of material behavior under static and dynamic loads (e.g. low-cycle fatigue, high-cycle fatigue) is presented. The work also includes an overview of damage mechanisms typical of time-temperature varying loading conditions (e.g. cyclic softening and hardening of the material, mean stress relaxation, material creep, visco-plasticity, etc.). This is followed by a structured review of several damage models of thermo-mechanical fatigue (e.g. Neu-Sehitoglu, DTMF, Coffin-Manson, Ostergren, Smith-Watson-Topper, Unified Energy Approach, etc.). An overview of the experimental tests on aluminum alloy and cast iron carried out at temperatures up to 800 °C is given. The idea of processing the raw experimental data including the calibration procedure of the thermo-mechanical fatigue damage models is schematically illustrated and described. The basic mathematical laws of constitutive material models for both material types are given. In the conclusion of the MSc thesis, the correlations of the calibrated damage models are presented, which, together with the constructive opinions, give an important message on the application of the individual damage models depending on the type of material and the loading method.
Ključne besede: thermo-mechanical fatigue, constitutive material model, damage model, aluminum alloy, cast iron alloy, finite element method
Objavljeno v DKUM: 03.01.2022; Ogledov: 1038; Prenosov: 7
Celotno besedilo (4,06 MB)