COMPOSITES 2023

Numerical and Experimental Methods for the Identification of Interlaminar Properties in LSI Manufactured C/SiC Composite

  • Riva, Marco (DAER - Polytechnic of Milan)
  • Airoldi, Alessandro (DAER - Polytechnic of Milan)
  • Caporale, Antonio Maria (DAER - Polytechnic of Milan)
  • Cavalli, Lorenzo (Petroceramics S.p.A.)
  • De Stefano Fumo, Mario (CIRA, Italian Aerospace Research Center)

In session: - Delamination I

Please login to view abstract download link

Ceramic Matrix Composites (CMC) present good structural properties and resistance to challenging environments. Currently the usage of Silicon Carbide matrix reinforced by carbon fibers (C/SiC) is a solution for hot structures and thermal protections in the aerospace field. To expand the usage of C/SiC composites in structural applications, the knowledge of failure modes and the ability to predict them must be extended. While in plane response of CMC is widely studied, the interlaminar damage and failure modes are less researched. Delamination can occur in CMC structural elements under different circumstances, some mutual to long fiber reinforced polymers, while others peculiar of CMCs and arising from defects due to the manufacturing process. This work aims to characterize the interlaminar properties of a C/SiC composite, hence an experimental campaign was performed by means of Double Cantilever Beam (DCB) tests. The evaluation of different methods to produce the pre-cracks is presented and the potential interaction between in-plane stresses and delamination is assessed using different thickness for the tested specimens. The tests showed that the crack propagation was characterized by deviation, inelastic phenomena and fiber breakage. By the application of Modified Beam Theory (MBT) to evaluate the interlaminar fracture toughness, an R-Curve effect was identified, which could be attributed to the development of damage in the reinforcing fibers of the plies adjacent to the interlaminar crack. The numerical modelling of the delamination is obtained through the implementation of a Cohesive Zone Model based on the tri-linear CZM, using an automatic identification algorithm for the model parameters. The identification procedure uses a regression model based on neural networks and an optimization through a genetic algorithm or a Monte Carlo like method to minimize the discrepancies between the experimental and the numerical response. The application of the method separately on each specimen allows to correlate the scattering of the experimental results with the parameters distribution. The work improves the understanding of the delamination phenomena in CMC produced via LSI technique, by defining a reliable numerical experimental protocol to identify the interlaminar toughness and the parameters of a CZM model capable of predicting the forces required for delamination propagation. The activity is a part of AM3aC2A project funded by Italian Space Agency (ASI).