Characterization and Detection of Delamination in Smart Composite Structures

Abstract

A multidisciplinary procedure has been developed for damage diagnosis or interrogation based on the concept of analyzing temporal relations between values of critical variables. Some observable variable of the system is traced through time from a specific initial state. The dynamics of the degradation process can be described by a time-variant mathematical model of this relationship. A novel neural-network-based approach to damage diagnosis and prognosis for nonlinear dynamic systems has been developed. High quality response surface approximations are developed to the progressive damage model. A neural network is trained to model the multi-dimensional response surface that relates the dependent variables to the independent variables. The main advantage of neural networks is that much more complex nonlinear relationships can be modelled, potentially incorporating high order interactions between predictive variables. A distinct feature of the proposed technique for constructing the response surface approximations is that it permits explicit treatment of the dynamics of the process under observation - in this case, structural damage that evolves in time. Assessing and quantifying existing damage may be treated as a static problem and its solution can be summarized by a mapping from parametric descriptions of damage attributes and measured structural response data.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 31, 2006
Accession Number
ADA470797

Entities

People

  • Aditi Chattopadhyay
  • Hsin-piao Chen

Organizations

  • California State University, Long Beach

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Carbon Fibers
  • Composite Materials
  • Computational Science
  • Elastic Properties
  • Finite Element Analysis
  • Kernel Functions
  • Laminates
  • Materials Laboratories
  • Materials Science
  • Materials Testing
  • Mechanical Properties
  • Mechanics
  • Micromechanics
  • Neural Networks
  • Polymer Matrix Composites
  • Reinforced Plastics
  • Stress Strain Relations

Readers

  • Computational Modeling and Simulation
  • Neural Network Machine Learning.
  • Structural Health Monitoring of Composite Structures.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms