详情

原文传递 3-D Multi-Scale Modeling Combined with Machine Learning for a Novel Structural-Prognosis Framework.

题名：	3-D Multi-Scale Modeling Combined with Machine Learning for a Novel Structural-Prognosis Framework.
作者：	Spear, A.
关键词：	Crack propagation, Machine learning, Microstructure, Materials, Aluminum alloys, X-ray computed tomography, Fatigue (mechanics)
摘要：	The goal of the research is to enhance structural-prognosis capabilities for the USAF by discovering a quantitative model capable of predicting the morphological evolution of three-dimensional (3-D)microstructurally small fatigue cracks (MSFCs) based on local, microstructure-sensitive fields. Three of the most significant hindrances to predicting the MSFC life for an arbitrary material microstructure under arbitrary far-field loading include: (1) uncertainty in the 'rules' (i.e. quantitative, parametric representations of the crack-driving mechanisms) that are used to evolve a 3-D crack at the scale of the microstructure; (2)missing or incomplete information in the cracks' surroundings and applied boundary conditions; and (3)inadequate representation of cracks as evolving discontinuities and their corresponding fluctuations. During the three-year period of this AFOSR Young Investigator Program (YIP) award, the PI and her graduate students have made significant research advancements toward improving structural-prognosis tools for the USAF by addressing each of these challenges.
报告类型：	科技报告