Efficient structural computations with parameters uncertainty for composite applications Export

@article{citeulike:6080814, abstract = {The aim of this paper is to propose a methodology in order to take into account the influence of uncertain data in structural calculations. To attain this goal, an approximation of the responses as a function of the uncertain data response called surface method (RSM) is proposed. In order to decrease the number of identification points necessary for the RSM, a progressive strategy is proposed. Error indicators are also used in order to increase the confidence. This strategy is applied to a laminate plate subjected to bending tests. The results are compared to Monte-Carlo simulations (considered as a reference). The method proposed in the present work permits to estimate correctly the whole response and is very simple to use (pre- and post-processing). This method is applied to asses the robustness of the point-stress method used to predict the rupture of perforated plates.}, author = {Carrere, N. and Rollet, Y. and Leroy, F. H. and Maire, J. F.}, citeulike-article-id = {6080814}, citeulike-linkout-0 = {http://dx.doi.org/10.1016/j.compscitech.2008.09.035}, citeulike-linkout-1 = {http://linkinghub.elsevier.com/retrieve/pii/S0266353808003734}, doi = {10.1016/j.compscitech.2008.09.035}, issn = {02663538}, journal = {Composites Science and Technology}, keywords = {kriging, mean-variance, metamodel, monte\_carlo, polynomial, robust, uncertainty}, month = {July}, number = {9}, pages = {1328--1333}, posted-at = {2009-11-06 13:14:10}, priority = {2}, title = {Efficient structural computations with parameters uncertainty for composite applications}, url = {http://dx.doi.org/10.1016/j.compscitech.2008.09.035}, volume = {69}, year = {2009} }

TY - JOUR ID - citeulike:6080814 L3 - citeulike-article-id:6080814 N2 - The aim of this paper is to propose a methodology in order to take into account the influence of uncertain data in structural calculations. To attain this goal, an approximation of the responses as a function of the uncertain data response called surface method (RSM) is proposed. In order to decrease the number of identification points necessary for the RSM, a progressive strategy is proposed. Error indicators are also used in order to increase the confidence. This strategy is applied to a laminate plate subjected to bending tests. The results are compared to Monte-Carlo simulations (considered as a reference). The method proposed in the present work permits to estimate correctly the whole response and is very simple to use (pre- and post-processing). This method is applied to asses the robustness of the point-stress method used to predict the rupture of perforated plates. IS - 9 JF - Composites Science and Technology SN - 02663538 EP - 1333 TI - Efficient structural computations with parameters uncertainty for composite applications VL - 69 SP - 1328 KW - kriging KW - mean-variance KW - metamodel KW - monte_carlo KW - polynomial KW - robust KW - uncertainty AU - Carrere, N AU - Rollet, Y AU - Leroy, FH AU - Maire, JF PY - 2009/07// UR - http://dx.doi.org/10.1016/j.compscitech.2008.09.035 ER -