This resource requires you to log in before you can proceed with the download.
Multiscale Modeling of Damage Mechanics of Fiber Reinforced Polymer Composites From Micro- to Macroscale: The (Virtual) Reality and Challenges Ahead
Category
Published on
Abstract
Multi-scale modelling of composites is a very active topic in composites science. This is illustrated by the numerous sessions in the recent European and International Conferences on Composite Materials, but also by the fast developments in multi-scale modelling software tools, developed by large industrial players such as Siemens, MSC/e-Xstream, Simulia, HyperSizer, Altair,… Over the past years, the elastic property prediction of composites has become quite established through a variety of multi-scale modelling approaches. When it comes to the nonlinear and damage behaviour of composites, multi-scale modelling becomes much more challenging. Experimental input data on the different scales are not always easy to obtain, in-situ observation of damage mechanisms like fibre/matrix debonding and fibre failure is difficult, anisotropic damage models that take into account effects of temperature and strain-rate are complex and computational efficiency is needed. Based on our recent research, the lecture will cover the three most important scales in multi-scale modelling of composites: (i) micro-scale, (ii) meso-scale and (iii) macro-scale. The nanoscale and related atomistic and molecular modelling approaches are not covered. A strong focus is put on physics-based damage modelling and the need for experimental validation, hence the link in the title to virtual reality and reality. The lecture will not only discuss the finite element based approaches for multi-scale modelling, but also much faster methods, such as Mean Field Homogenization methods and variational methods. Examples will be given for unidirectional composites, textile composites and short fibre-reinforced composites, and advanced experimental testing techniques will be illustrated.
Cite this work
Researchers should cite this work as follows:
Submitter
Purdue University