TY - JOUR
T1 - Multiscale modelling of porous polymers using a combined finite element and D-optimal design of experiment approach
AU - Spaggiari, Andrea
AU - O'Dowd, Noel
AU - Dragoni, Eugenio
PY - 2011/7
Y1 - 2011/7
N2 - This paper presents a general numerical procedure for the analysis of polymeric materials containing spherical voids. The multiscale approach implemented simulates, via three dimensional finite-element analysis, an infinite medium of the material containing discrete voids. A D-optimal design procedure is used to combine the seven normalized variables considered in the problem: the material strength and ductility, the hardening ratio, the void volume fraction, the void arrangement (number of voids), the stress triaxiality and the Lode parameter. The failure criterion considered is based on a critical distance approach, considering a brittle epoxy resin as a reference material. Results are provided for the normalized equivalent stress and strain at failure, the void growth rate and the equivalent failure strain. The influence of the variables on the outputs is estimated and design equation coefficients are calculated.
AB - This paper presents a general numerical procedure for the analysis of polymeric materials containing spherical voids. The multiscale approach implemented simulates, via three dimensional finite-element analysis, an infinite medium of the material containing discrete voids. A D-optimal design procedure is used to combine the seven normalized variables considered in the problem: the material strength and ductility, the hardening ratio, the void volume fraction, the void arrangement (number of voids), the stress triaxiality and the Lode parameter. The failure criterion considered is based on a critical distance approach, considering a brittle epoxy resin as a reference material. Results are provided for the normalized equivalent stress and strain at failure, the void growth rate and the equivalent failure strain. The influence of the variables on the outputs is estimated and design equation coefficients are calculated.
KW - D-optimal
KW - Design of experiment
KW - Multiscale modelling
KW - Polymers with voids
UR - http://www.scopus.com/inward/record.url?scp=79957477594&partnerID=8YFLogxK
U2 - 10.1016/j.commatsci.2011.04.017
DO - 10.1016/j.commatsci.2011.04.017
M3 - Article
AN - SCOPUS:79957477594
SN - 0927-0256
VL - 50
SP - 2671
EP - 2682
JO - Computational Materials Science
JF - Computational Materials Science
IS - 9
ER -