TY - GEN
T1 - Application of artificial neural networks to the simulation of progressive damage in composite laminates
AU - Deenadayalu, Chaitanya A.
AU - Chattopadhyay, Aditi
AU - Zhou, Xu
PY - 2006
Y1 - 2006
N2 - A procedure has been developed for simulating progressive damage in composite laminates using a combination of an accurate stress analysis technique, a quasicontinuum approximation of the damaged material, and artificial neural networks. A response surface model, based on a multi-layer feed-forward network, is constructed as a surrogate for the computationally intensive higher fidelity finite element models by performing a series of numerical experiments and then observing the responses from those experiments. Two partially recurrent neural network topologies are proposed as part of an adaptive structural health monitoring strategy that emphasizes damage diagnostics as a prelude to predicting potential damage. A finite element model for nonlinear analysis of composite plates coupled with a micromechanics-based mechanistic model for modeling the initiation and evolution of damage and for predicting the effective composite properties is used to construct data for the response surface models.
AB - A procedure has been developed for simulating progressive damage in composite laminates using a combination of an accurate stress analysis technique, a quasicontinuum approximation of the damaged material, and artificial neural networks. A response surface model, based on a multi-layer feed-forward network, is constructed as a surrogate for the computationally intensive higher fidelity finite element models by performing a series of numerical experiments and then observing the responses from those experiments. Two partially recurrent neural network topologies are proposed as part of an adaptive structural health monitoring strategy that emphasizes damage diagnostics as a prelude to predicting potential damage. A finite element model for nonlinear analysis of composite plates coupled with a micromechanics-based mechanistic model for modeling the initiation and evolution of damage and for predicting the effective composite properties is used to construct data for the response surface models.
UR - http://www.scopus.com/inward/record.url?scp=34247130607&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=34247130607&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:34247130607
SN - 1563478080
SN - 9781563478086
T3 - Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
SP - 1998
EP - 2015
BT - Collection of Technical Papers - 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
T2 - 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
Y2 - 1 May 2006 through 4 May 2006
ER -