Multiscale analysis for predicting the constitutive tensor effective coefficients of layered composites with micro and macro failures

• Published in: Applied Mathematical Modelling Vol. 75; pp. 250 - 266
• Main Authors: Brito-Santana, Humberto, Thiesen, José Luís Medeiros, de Medeiros, Ricardo, Ferreira, Antonio Joaquim Mendes, Rodríguez-Ramos, Reinaldo, Tita, Volnei
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.11.2019; Elsevier BV
• Subjects: Asymptotic homogenization method; Asymptotic methods; Coefficients; Composite materials; Debonding; Delamination; Failure; Failure analysis; Finite element analysis; Finite element method; Homogenization; Laminates; Multilayers; Multiscale analysis; Numerical models; Tensors; Asymptotic homogenization method; Debonding; Finite element analysis; Delamination
• ISSN: 0307-904X; 1088-8691; 0307-904X
• DOI: 10.1016/j.apm.2019.05.031

•Multiscale analysis for layered composites with delamination in macro-scale is considered.•The micro-scale effective properties are calculated using a numerical approach.•Macro-scale effective coefficients are calculated using numerical and analytical models.•The comparison between the numerical and analytical models provides good agreement.•The homogenized numerical model provides excellent approximation to heterogeneous numerical model. The aim of the present work consists of predicting the effective coefficients of constitutive tensor for layered composites with delamination in macro-scale considering the influence of the debonding between fiber and matrix in micro-scale. Thus, a multiscale methodology is proposed to solve this problem. Firstly, the effective coefficients for the micro-scale level are calculated via the Finite Element Method (FEM), considering different degrees of micro failure. By using the micro-scale level homogenization results, the effective coefficients for layered composites are calculated by the Finite Element Method, as well as by Asymptotic Homogenization Method (AHM), considering different extensions of delamination (failures in macro-scale). Results show that the macro-scale analyses are affected by the micro failure, with the most influence of the micro failure in the coefficients C11*, C12* and C66*. In addition, when the thickness of the adhesive between layers increases, the effective coefficients decrease with the macro failure (delamination). Comparisons between homogenized and heterogeneous numerical models show that for almost all effective coefficients, there are excellent convergences. Only for the values of C12*, there are relevant divergences for specific limit cases. Finally, the macro-scale results obtained via FEM and AHM are compared to evaluate the advantageous and disadvantageous of the proposed multiscale methodology.