A numerical approach to the analysis of failure modes in anisotropic plates
β Scribed by A. Pandolfi; K. Weinberg
- Publisher
- Elsevier Science
- Year
- 2011
- Tongue
- English
- Weight
- 860 KB
- Volume
- 78
- Category
- Article
- ISSN
- 0013-7944
No coin nor oath required. For personal study only.
β¦ Synopsis
This study is motivated by the attempt to characterize failure modes of silicon chips commonly used in electronic industries. Previous experimental investigations provided the failure probability of dies made of a single-crystal and produced a large variety of crack patterns, but were not able to elucidate the link between defect distributions and crack initiation and propagation. To get some insight in the fracture activation and propagation mechanisms, we resort to finite element analyses and adopt an explicit methodology for crack tracking, based on the self-adaptive insertion of cohesive elements into a coherent mesh of solid elements. Finite kinematics material models with anisotropic features for both bulk and cohesive surfaces are employed to describe the behavior of single-crystal silicon plates undergoing a particular bending test up to failure. The cohesive model adopted in the calculation is fully anisotropic and newly formulated to accomplish the present study. Numerical simulations considering different material properties were able to ascertain the effects of particular flaws on failure modes of brittle silicon plates.
π SIMILAR VOLUMES
## Abstract Some new aspects of the continuum strong discontinuity approach (CSDA) to model material failure in geomaterials are addressed. A new global algorithm, for tracking multiple crack lines/surfaces in 2D/3D cases is proposed. It is based on solving a simple heat conductionβlike problem acc
The extended kinematic minimum principle is used to formulate a quadratic pro-gram~ng solution to the rate problem in elastoplasticity for plane continua. A simple algorithm is presented which involves the minimisation of a continuous functional subject to inequality constraints. The algorithm has b