| Résumé : |
In a 2002 paper by Hinton, Kaddour, and Soden (Composites Science and Technology, 62 : 1725-1797) on a summary of the so-called World-Wide Failure Exercise, whereby 14 theories for predicting failure in composite laminates were tested against experimental evidence, the authors stated that "the theories were not sufficiently robust" and "the designer wishing to estimate the stress levels at which ultimate failure might occur in a multi-directional laminate, can, at best, hope for accuracy of +/-50 % in the majority of cases."
To address this lack of "robustness" issue, we propose in this paper a new approach based on the concept of Uncertainly Quantification (UQ), whereby the stress (and/or strain) variables and the material property parameters in each failure criterion are considered to by stochastic. For material property parameters with small coefficient of variation, c, where O
To illustrate this approach, we construct a simple example based on a combination of the well-known failure theories of Tsai and Wu (1971), Hashin (1980), and Christensen (1997). Using a public-domain statistical analysis software named DATAPLOT, we present a numerical example of how to quantify the uncertainty of stress (and/or strain) variables in a composite failure criterion and estimate the A- and B-basis allowables assuming that the uncertainty of all material property parameters is small. Significance and limitations of the UQ methodology developed in this paper are presented and discussed. |
Aller sur edunet
Accueil
Ajouter le résultat dans votre panier Affiner la rechercheComposites failure criteria and estimation of associated 1- and B-basis design allowables in JEC COMPOSITES MAGAZINE, N° 71 (03/2012)
Modélisation et estimation des erreurs de mesure / Paris : Technique & Documentation - Lavoisier (1998)
